Sei sulla pagina 1di 370


KllG'S llDIAI VOlUlf 1

First published in 2010 by Gloucester Publishers plc (formerly Everyman Publishers
pk), North burgh House, 10 Northburgh Street, London EC1 V OAT

Copyright© 2010 David Vigorito

The right of David Vigorito to be identified as the author of this work has been as­
serted in accordance with the Copyrights, Designs and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a

retrieval system or transmitted in any form or by any means, electronic,
electrostatic, magnetic tape, photocopying, recording or otherwise, without prior
permission of the publisher.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

ISBN: 978 1 85744 645 6

Distributed in North America by The Globe Pequot Press, P.O Box 480,
246 Goose Lane, Guilford, CT 06437-0480.

All other sales enquiries should be directed to Everyman Chess, North burgh House,
10 Northburgh Street, London EC1 V OAT
tel: 020 7253 7887 fax: 020 7490 3708
email:; website:

Everyman is the registered trade mark of Random House Inc. and is used in this
work under licence from Random House Inc.

This book is dedicated to two people: for my darling wife Heather, for all her love
and support; and for my dear friend Joe Fang.for getting me well on my way in the
King's Indian Defence with many long nights of speed chess back in the olden days...

Everyman Chess Series

Chief advisor: Byron Jacobs
Commissioning editor: John Emms
Assistant editor: Richard Palliser

Typeset and edited by First Rank Publishing, Brighton.

Cover design by Horatio Monteverde.
Printed and bound in Great Britain by Clays, Bungay, Suffolk.

Bibliography 5
Introduction 7

Part I: The Classical Variation

1 d4 lt:Jf6 l c4 g6 3 lt:Jc3 il.g7 4 e4 d6 S lt:Jf3 o-o
6 ii.el es 7 0-0 lt:Jc6 10

1 Mar del Plata Variation: 9 tt:Je1 lt:Jd7 10 tt:Jd3 12

2 Mar del Plata Variation: 9 tt:Je1 lt:Jd7 10 f3 38
3 Mar del Plata Variation: 9 tt:Je1 lt:Jd7 10 .ie3 47
4 Mar del Plata Variation: 9 tt:Jd2 76
s Mar del Plata Variation: 9 b4 tt:Jhs without 10 .:tel 92
6 Mar del Plata Variation: 9 b4 tt:Jhs 10 .!:tel 105
7 Mar del Plata Variation: White's Other 9th Moves 144
8 White's Eighth Move Deviations 156

Part II: The Classical Variation

1 d4 lt:Jf6 l c4 g6 3 lt:Jc3 il.g7 4 e4 d6 S lt:Jf3 o-o
6 ii.el es without 7 o-o 167

9 Gligoric Variation: 7 .ie3 lt:Jg4 8 f6 9 tt:Jc6, 9 ilh4 tt:Jc6 168
10 Gligoric Variation: 7 .ie3 lt:Jg4 8 f6 9 il.h4 gs 184
11 Petrosian Variation: 7 dS as without 8 202
12 Petrosian Variation: 7 dS as 8 222
13 Exchange Variation : 7 dxes dxes 8 '1i'xd8 .l:txd8 238

Part Ill: The Samisch Variation

1 d4 lt:Jf6 2 c4 g6 3 lt:Jc3 .ig7 4 e4 d6 S f3 o-o 6 .ie3 tt:Jc6
7 tt:Jge2 a6 8 '1i'd2 :tbs 264

14 Panno Variation: 9 h4 without 9 ... h s 266

1S Panno Variation: 9 h4 h S 10 o-o-o 2 73
16 Panno Variation: 9 h4 hS 10 tt:J c1 293
17 Panno Variation: 9 tt:J c1 302
18 Panno Variation: 9 0-0-0, 9 .ih6 and 9 g4 320
19 Panno Variation: 9 a3, 9 !:!.bl and 9 .:tel 331
20 Panno Variation: Other lines 344

Index of Variations 3 64

6 ... liJc6 in the Sii.[nisch Variation, King's Indian Defense, John Watson (Chess Enter­
prises 1982)
Attack ing Manual I, Jacob Aagaard (Quality Chess 2008)
Beat the KID, Jan Markos (Quality Chess 2009)
Beating the Anti-King's Indians, Joe Gallagher (Batsford 1996)
Beating the Fianchetto Defences, Efstratios Grivas (Gambit 2006)
Dangerous Weapons: The King's Indian, Richard Palliser, Glenn Flear & Yelena
Dembo (Everyman Chess 2009)
King's Indian Defence: 4 e4, Efim Geller (Batsford 1980)
King's Indian Defence: Ma r del Plata Variation, Svetozar Gligoric (Batsford 2003)
My Great Predecessors Part Ill, Garry Kasparov (Everyman Chess 2004)
My Great Predecessors Pa rt IV, Garry Kasparov (Everyman Chess 2005)
Play 1 d4!, Richard Palliser (Batsford 2003)
Play the King's Indian, Joe Gallagher (Everyman Chess 2004)
Starting Out: 1 d4!, John Cox (Everyman Chess 2006)
Starting Out: The King's Indian, Joe Gallagher (Everyman Chess 2002)
The Art of the King's Indian, Eduard Gufeld (Batsford 2002)
The Classical King's Indian Uncovered, Krzysztof Panczyk and Jacek llczuk (Every­
man Chess 2009)
The Controversial Sii.misch King's Indian, Chris Ward (Batsford 2004)
The King's Indian: A Complete Black Repertoire, Victor Bologan (Chess Stars 2009)
The King's Indianfor the Attacking Player, Graham Burgess (Batsford 1993)
The Main Line King's Indian, John Nunn & Graham Burgess (Batsford 1996)
The New Classical King's Indian, John Nunn & Graham Burgess (Batsford 1998)
The Sii.misch King's Indian, Joe Gallagher (Batsford 1995)
The Sii.misch King's Indian Uncovered, Alexander Cherniaev and Eduard Prokuronov
(Everyman Chess 2008)
Understanding The King's Indian, Mikhail Golubev (Gambit 2005)

A t ta c king C h e s s : The King 's In dian, V o l u m e 1

Winning with the King's Indian, Eduard Gufeld (Macmillan 1991)

Winning with the King's Indian, Andrew Martin (Caissa 1989)

Chess Informant through Volume 105
New In Chess Magazine through issue 2010/4
New In Chess Yearbook through Volume 94

Electronic Resources
Mega Database 2009 (ChessBase)
The King's Indian, Viktor Bologan (ChessBase 2009)
Chess Publishing .com
The Week in Chess through issue 815


The King's Indian Defence is one of the richest openings in all of chess theory.
Black does not play to equalize as he does in the classical defences. Rather he seeks
to unbalance the game from the outset. The last decade has seen a revitalization
of the King's Indian, as even top players are often trying to win with the black
pieces. Compared to the classical openings, the price of each move is quite high
and a mistake by either side can easily lead to disaster.
The King's Indian has always been considered a somewhat risky opening, but
despite that common sentiment, the King's Indian has an impressive pedigree.
While this dynamic system was pioneered in the 1950s by Russian and Yugoslav
players such as David Bronstein, Efim Geller and Svetozar Gligoric, the two big
names that are often attached to the King's Indian are those of its World Cham­
pion practitioners, Robert Fischer and Garry Kasparov. Whereas Fischer's retire­
ment signalled the end of his King's Indian era, Kasparov gave up our favourite
opening while he was still an active player, which 'indicated' its unsoundness. At
least that was the general feeling after he lost a well-known game in 1997 to
Kramnik in the then dreaded 'Bayonet' system.
In fact Kasparov stated something to the effect that the Sicilian and King's In­
dian were too much to keep up with at the level he was playing at, and so he stuck
with the Sicilian while heading for more solid systems in the closed openings.
Nowadays young players are not so worried about this; with advances in technol­
ogy many modem talents play both the Sicilian and the King's Indian, as well as
other sharp defences.
Opening fashions come and go. The beginning of the new millennium brought
forward a great new champion of the King's Indian Defence in Teimour Radjabov.
Like Kasparov, Radjabov hails from the city of Baku in Azerbaijan. Radjabov really
took over where Kasparov left off, even scoring well in the aforementioned Bayo­
net (see Chapters 5 and 6). Radjabov's success influenced the younger generation
as well as the old guard, and nowadays most of the top players have been found at
one time or another on the black side of the King's Indian.

A ttac king Ch ess: Th e King 's In dian, Vol u m e 1

The King's Indian Defence has always been an opening I've felt greatly attached
to. Despite the fact that I have written extensively on the Slav Defences, the King's
Indian was my first real defence to 1 d4. While the King's Indian is considered to
be a 'tactical' opening, I have always considered it to be very strategic in nature. It
is an opening where a feeling for piece placement and pawn structure is very im­
portant. There are many thematic ideas and although the opening lends itself to
frequent complications, the tactics have always seemed 'logical' to me. So, while it
is true that when I 'grew up' I began to rely more on the solid Slav systems, it is
always useful to have a sharp weapon available, especially when one really wants
to try to win with Black.
Even though the King's Indian is a complicated opening, I do not think it is so
difficult to learn. For one thing, it is relatively 'move order proof'. That is, the King's
Indian set-up can be employed against 1 d4, 1 c4, or 1 tt:Jf3. Also, the King's Indian
lends itself to just a handful of pawn structures, so the ideas are easier to assimi­
In this book, as well as the second volume, I will generally focus on the main
lines. The reason for this is that I think the best way to learn an opening is to study
the main lines. It is easy to add other secondary systems later. The biggest exam­
ple of this is in the Classical Variation, 1 d4 tt:Jf6 2 c4 g6 3 tt:J c3 .ig7 4 e4 d6 5 lt:Jf3
o-o 6 ile2 es 7 o-o. Here I have gone for what is no doubt the main line, 7
This is the most uncompromising approach and it is also the most difficult. The
main alternative is 7... tt:Ja6. This line is safer, easier to learn and may well be just
as good as 7... tt:Jc6. In fact, I have pl ayed 7 ... tt:Ja6 myself on several occasions. Nev­
ertheless, I think it is best for the aspiring King's Indian player to learn the main
lines. One great thing about the King's Indian is its flexibility - if you learn the
main systems, it is easy to expand your repertoire by adding additional lines with­
out having to learn a whole new opening.
In the Sa mi sch Variation, 1 d4 tt:Jf6 2 c4 g6 3 tt:Jc3 .ig7 4 e4 d6 5 f3 o-o 6 .ie3,
the main line nowadays is 6... cs. This was not always the case, and 6 c6 and
6 ... es used to be considered Black's two main systems. I have elected not to go
with 6 ... cs even though it may well be the best move. While White often steers the
game into a Benoni structure, it is also possible for White to simply grab a pawn
while exchanging queens as well. Modem practice has clearly shown that Black
gets sufficient compen sation for the pawn, but some White players are rather well
prepared in these endings. If Black is well prepared too and has a good under­
standing of these positions, he should certainly be able to draw, but I believe that
heading into a pawn-down endgame straight out of the opening is probably not
to everyone's taste.

In tro d u c tion

Moreover, 6...cs against the Samisch has been very well covered in modem
King's Indian literature and I did not have a lot to add to what is already out there,
especially as the lack of popularity of the Samisch at high level has not produced
much fresh material to examine. The classical 6 ... es must still be playable, but I
always thought it was easier to play White in these lines and so decided to go for
the Panno Variation with 6 ... tt:Jc6. This is an uncompromising system that still has
a lot of fresh territory to explore.
Note that the second volume will cover the Fianchetto Variation, the Four
Pawns Attack, the Averbakh and all of White's other tries.
I would like to thank a few people for their help with this book: John Emms, for
suggesting the topic, allowing me to split the book into two parts and for his pa­
tience; Richard Palliser, for listening to me rant and rave about various lines that
may or may not have found their way into this book; Joe Fang, for the use of his
impressive library; and Vasik Rajlich, for keeping me well supplied with Rybka 3
and 4. Thanks to you all!

David Vigorito
Somerville, Massachusetts,
October 2010

Part 1
The Classical Variation
1 d4 lbf6 2 c4 g6 3 ttJc3 i..g7 4 e4 d6
s tt:Jf3 o-o 6 il.e2 es 7 o-o lbc6 and 6, which aim to

avoid certain lines in Part II (most nota­
bly the Exchange Variation), but in my
opinion we should dive headfirst into
the most critical lines. After 7 0-0 (other
moves are considered in Part I I ) we will
play 7 ... tt:Jc6!. Here White usually plays 8
ds (other moves are discussed in Chap­
ter Eight), and after 8 .. lt:Je7 we have the

Mar del Plata Variation.

The Classical Variation 1 d4 tt:Jf6 2 c4

g6 3 ltJC3 ilg7 4 e4 d6 5 tt:Jf3 0-0 6 .ie2
is one of White's most principled sys­
tems against the King's Indian Defence.
White develops in a logical manner by
simply developing his kingside and
preparing to castle. Despite this appar­
ently modest approach, the Classical is
a very complex set of systems and sub­
systems which can lead to many differ-
ent types of position. Most of the world's top players have
We will focus on the main move, been found on one side or another of
6 ... es!. There are other moves, such as this position (some players like both

The Classic a l Va ria tion

sides), because the play is very rich knight by playing ...'it>h8 and
both strategically and tactically. In This looks funny at first, but the knight
general White will play on the queen­ can re-emerge on f6 or even h6. There
side and Black on the kingside, but are also cases where the knight goes to
there are many instances where one c8 or even to c6.
side can take the initiative on their White has four basic ways to pro­
'weak' side. ceed in this position. He can play 9
One feature that is specific to the tt:Je1, which can be considered the
Mar del Plata Variation is the position 'main line'. This move is considered in
of Black's e7-knight. In many King's Chapters 1-3. A different knight rede­
Indian lines with a blocked centre, ployment is 9 tt:Jd2, which is covered in
Black's queen knight would head for Chapter 4. The dangerous 'Bayonet', 9
the natural cs-square by b4, has been White's main weapon for
or ... ltJb8-a6-cS (see, for example, the the last couple of decades. The play
Petrosian Variation of Chapters 11 and tends to be quite different here, as cen­
12). In the Mar del Plata, Black has tral play is more prominent than it is
played 7 c6 in order to entice White after White's knight moves. 9 b4 is cov­
to clarify the structure in the centre ered in Chapters 5 and 6. Other 9th
with 8 dS. In general the knight on e7 is moves are less common but they are
not well placed and Black's success is not altogether harmless. These lines
often connected in some way with ac­ will be examined in Chapter 7. Some­
tivating this piece. In many of the lines times too White chooses to avoid the
with 9 tt:Je1, Black brings the knight to discussion of the Mar del Plata Varia­
g6. The knight is not necessarily well tion and avoids the critical 8 dS. These
placed there, but it may participate in a sidelines are not so dangerous, but
kingside attack. Alternatively, the Black should not ignore them, as they
knight may emerge on fS (after an ex­ can be tricky to meet for the unpre­
change of pawns on fS or e4), and pared. White's 8th-move deviations are
sometimes Black will redeploy the covered in Chapter 8.

Chapt�r1· . ·. . . .

The Mar def Plata Variation

1 d4 lt:Jf6 2 c4 g 6 3 tt:Jc3 .ig7 4 e4 d6 5 f4 The knight will help force through

tt:Jf3 0-0 6 ii.el es 7 0-0 tt:Jc6 8 ds lt:Je7 9 White's important c4-cS advance and it
tt:Je1 may also drop back to the f2-square,
where it will help prevent Black's ... gs­
g4 pawn break.
One thing we must always be aware
of is the value of Black's c8-bishop.
Without this piece not only is it diffi­
cult to break with ... gS-g4, but the
bishop is often needed to sacrifice itself
on h3 if the kingside becomes blocked.
Sometimes the bishop does not even
move until it can deliver the decisive
blow. So valuable is this bishop to
This move quickly became White's Black's attack that we will often see
main try when the Mar del Plata Varia­ White spending a number of tempi to
tion first became popular and it re­ hunt down the bishop on its original
mains topical to this day. White pre­ square.
vents Black's knight manoeuvre, 9 lt:Jd7

prepares to bolster the centre with f3 if Black clears the way for .. .f7-fS and
necessary and avoids blocking in his c1- covers the cs-square at the same time.
bishop. Although the knight is passively Another plan for Black is 9 This
placed at the moment, it can quickly looks less logical, but as we shall see in
come to d3, from where it surveys the the main lines following 9 ... lt:Jd7 10
important central squares cs, es and tt:Jd3 fS 11 .id2 tt:Jf6 12 f3 f4, Black of-

Th e M a r de/ Plata Varia tio n : 9 tlJe1 tlJ d 7 10 tiJd3

ten drops the knight back to e8 after all The next chapter examines the radical
in order to protect the sensitive c7- plan 10 f3 fS 11 g4! ?, while Chapter 3
square. By playing 9...t2Je8 straight­ covers White's most aggressive con­
away, Black hopes to gain a couple of tinuation, 10 .ie3.
tempi. Of course White does not need 10 t2Jd3
to be so cooperative though. For one White improves the position of his
White can continue with 10 t2Jd3 fS, knight and prepares the c4-c5 advance.
but instead of 11 .id2, as played in this This is a very straightforward continua­
chapter, White may change the charac­ tion and some of the theory goes very
ter of the position with 11 f4, exploit­ deep in the main lines. The fashion of
ing the fact that Black's knights are far such long lines comes and goes, and
away from the es-square. Another eventually White turned to other sys­
dangerous plan is to continue along tems in the Classical (most notably 10
the lines of Chapter 2 by playing 10 f3 .ie3 and the Bayonet), while Black also
fS 11 g4, when the knight on e8 is less found ways to deviate. Recently 10 t2Jd3
flexible than it would be on d7. Perhaps has become rather fashionable again.
the greatest danger, though, comes We will generally study the main lines
from 10 .ie3. In this case White will because that will help to develop a
have no trouble playing c4-c5 and good general King's Indian under­
plans involving a4-a5 become very standing. There are also some interest­
dangerous. ing sidelines that will be mentioned
While 9 ... t2Je8 may be playable, the that may be worthy of further research.
main move has always been 9 ...l2Jd7 10 f s 11 .id2

and most experts consider it best.

This is the main line. White simply

From here White has three main develops his bishop and prepares !:!.cl.
schemes to choose between. In this There are a couple of alternatives.
chapter we look at the flexible 10 t2Jd3. a) First, let us take a look at 11 f3.

A ttacking Ch ess: The King 's In dian, Vo l u m e 1

will have to spend an additional tempo

to break with c4-cS. This move does
have some historical significance, how­
ever. Now 12 ... gs 13 b4

This is quite a natural reaction to

Black's pawn advance, but sometimes
this move is considered inaccurate.
This is because Black can begin his
kingside play while leaving his knight 13 14 cs tt:Jg6 1S .l:tc1 .l:tf7
on d7 to hold up White's cs-break. compares favourably to the main line
Nevertheless, this move is not bad and for Black because White has played b4
some even consider it to be the most (instead of tt:Jbs or cxd6, for example)
accurate move because it avoids the which may or may not tum out to be a
line 11 .id2 fxe4!?, which we discuss in useful move.
the notes to Black's 11th move, below. Black's last few moves seem rather
White may also steer the game into obvious today, but this was not always
independent channels by avoiding .id2 so. An old plan for Black in such posi­
altogether. Here Black can play 11 .. .f4 tions is to play ....l:tf6 and ... .rf.h6, with
right away, because there is no .ie2-g4 the idea of directly attacking down the
to worry about. 11 . is possible too, h-file with ...'1i° e8 and ...'ifh s. If White
but Black can try to take advantage of plays h2-h3, Black will try to engineer a
White's move order by leaving the ....ixh3 sacrifice. Although this plan
knight at d7 to hinder White's cs­ can indeed be quite dangerous to the
advance. Note that 11 ... 'it>h8?! is not so first player, to me it seems a bit too
good because White can switch gears simplistic to think that White's position
with 12 .ie3!, aiming for positions akin can just be taken by storm like this.
to those in Chapter 3 where 11 ...'it>h8 is Often Black will have to worry about
likely to be a wasted tempo. the possibility of White playing tt:Jbs
After 11 .. .f4 White has: after the queen has gone to e8. He can
al) 12 .id2 looks like an inferior deal with this by throwing caution to
version of the main line because White the wind and allowing tt:Jbsxe7. This

Th e M a r de/ Plata Va ria tion : 9 li.Je1 li.J d 7 10 li.Jd3

sounds like fun, but if White can play 1s ... :tf7 above) 16 li.Jf2 h S 17 cxd6 cxd6
li.Jf2 and h3, the h-file attack will not 18 a4 .if8! 19 as .l:tg7 20 h3 li.Jh8!.
work. In this case, Black can change the
direction of the attack by then playing
... .l:tg6 and ... hs.
Another way for White to defend is
by playing 'it>h1 and .ig1. Black does
not really have any way to increase the
pressure on h2, but if he could get a
knight to g 3, it would be mate. This can
be achieved either by moving the black
queen from h S to threaten or
by a sacrifice such as ... li.Jf6xdS; exds
li.Je7-fS. This all sounds very nice, but if As we shall see, Black's 18th and
White is alert to the possibility it is not 19th moves are extremely common­
so easy to achieve these tricks. place, but the last one requires a few
White may also defend against words of explanation. The g6-knight
mate on h2 in some positions with his often heads to h4 to both unleash the
bishop on f2 by playing the ugly h4. g7-rook and scare the white king. The
This may look fatally weakening, but it knight manoeuvre to h8 is not so un­
is not always easy to get through and common, though, and the knight will
Black's queenside will certainly disap­ then come back to f7, from where it
pear. may hop to gs or even to h6, to further
Black can also throw in ... a6 before support the ... g4-break. We will even
moving the queen to safeguard against see positions where Black plays ... 'it>h8
the knight raid to bS, but this costs a and ... li.Je7-g8 in order to come to h6.
tempo and also leaves Black vulnerable The game continued 21 li.Jbs g4 22 fxg4
to li.Ja4-b6 ideas after an exchange of hxg4 23 hxg4 a6 24 li.Ja3 .id7 2S li.Jc4
pawns on d6. In general I have avoided .l:tc8 26 li.Jb6 .l:txc1 27 .ixc1 .ie8 28 .ia3
lines where Black seeks to attack in this li.Jf7 29 '1i'c2 li.Jh6 30 gs .l:txgs 31 !:i.c1
way, but there are exceptions as we .l:tg3 32 .ib2 li.Jfg4 33 li.Jxg4 li.Jxg4 34
shall see, most notably in Line C of .ixg4 .l:txg4 3S '1i'f2 .ig6 36 .l:tc4 '1i'e7 37
Chapter 3. .ic3 '1i'h7 38 '1i'e2 .l:Ih4 39 'it>f2 f3 40 '1i'e3
Another attacking plan, which .l:tf4 41 gxf3 '1i'h2+ 42 'it>e1 '1i'h1+ 43
seems more natural to me, is to play 'it>e2 .ihs 44 'it>d2 .:txf3 4S 'iY'gs+ .ig7
... li.Jf6 and ... hs. Gligoric was one of the 46 'it>c2 !:i.f2+ 47 .id2 '1i'd1+ 48 'it>c3
first players to develop this concept. '1i'a1+ 0-1, M.Najdorf-5.Gligoric, Mar del
One of his games continued (after Plata 19S3. Najdorf learnt a lesson

A tt a c king C h e s s : Th e King 's In dian, Vo l u m e 1

from this game as we shall see in Line B 1 4...'it>g 7!?) 1 S ii.el .ixel 16 tt:Jxel 'it>g7
of Chapter 3. 17 h4?! gs (or 17 ... hxg4 18 fxg4 .l:th8 19
Going back, White probably can re­ tt:Jg2 gs 20 hs tt:Jf6) 18 tt:Jg2 hxg4 19
frain from 13 b4 and play 13 !:!.cl tt:J g6 fxg4 gxh4 20 tt:Jxh4 .l:th8 Black had a
14 cs!? when instead of grabbing the strong attack brewing in S.Karp­
pawn, Black should really just play A.Kuzmin, Ostend 1991.
14 to tran spose to the main lines, a3) 12 b4 (in the main line we will
while 13 ... hs is also possible. see that this preparatory move usually
Another idea is 13 g4 when Black is not necessary; the only way this
can pl ay 13 ... hs 14 h3 !:i.f6!? (Black tries move makes sense for White is if he
to turn the delay in going to his intends to deploy the bishop on a3 or
benefit) 1S b4 (White should probably b2) 12 ... gs (Black has also tried to take
prefer a quick ii.el and 'it>g 2) 1S ... !:i.h 6 advantage of White's move order with
16 'it>g2 tt:Jg6 17 !:i.hl lt:Jh4+ 18 'it>f2 .if8 12 ... as, but I think Black should just
19 !:!.cl as 20 a3 c6, which gave him leave the queenside alone and get on
good play in R.Shabtai-D.Komljenovic, with it) 13 cs tt:Jf6 14 a4 and now:
Biel 1989.
a2) 12 g4 is quite a common
counter in general - see, for instance,
Line D of this chapter. Here, however, it
does make a difference that Black has
not played either or ... gs yet.

a31) 14...'it>h8 ! ? (here we see an­

other typical plan; Black gears up for a
quick ...g4, rather than play ....l:If7-g 7,
which makes some sense because
White has refrained from .id2 and
!:!.cl, so the c7-square does not really
White's thrust is well met by need additional protection) 1S .ia3
12 ....if6 !. Black's delay in playing ... tt:Jf6 .l:tg8 16 as .if8 17 cxd6 cxd6 18 bS (or
allows this move, seizing control of the 18 h3 hS 19 tt:Jf2 when Black could try
h4-square. After 13 .id2 hS 14 h3 .ih4 19 ....l:I g7 20 !:!.cl tt:J eg8, with the idea
(alternatives are 14... �7 and h6 and ...g4) 18 ... g4 and Black had

Th e M a r de/ Plata Va riatio n : 9 li.J e1 li.Jdl 10 li.J d3

good play in S.Lagrotteria-B.Socko, .igs 17 .id3 gave White a slight edge

Saint Vincent 2001. in F.Doettling-A.Hunt, Patras 1999) 1S
a32) The direct 14...h s is also possi­ li.Jxe4 'ii'b6+ (1S ...cxds 16 'iY'xds+) 16
ble and after 1s li.Jf2 li.Jg6 16 as .l::i. f7 17 cs!? dxcs 17 .ic4 gives White danger­
c6 .if8 18 cxb7 .ixb7 19 !:i.a3 .l:tg7 20 h3 ous compensation for the pawn .
.ic8 21 bS li.Jh8 22 li.Ja2 li.Jf7 23 li.Jb4 b3) 13 ... cs 14li.Jfe4 li.Jd4 1 s .igs .ifs
li.Jh6 24 li.Jc6 '1i'e8 2S '1i'c2 g4 Black had looked fine for Black in T.Overbeck­
good counterplay in G.Vescovi­ V.Rojicek, Pardubice 2009.
M.Krylov, Moscow 2010. b4) 13 ... li.Jd4 is simple enough: 14
b) The old move 11 exfs is not seen li.Jfe4 c6 1S .ie3 li.Jxe4 16 li.Jxe4 cxds 17
too often these days. The simplest solu­ cxds .ifs was equal in R.Markus­
tion is 11 Usually in the King's R.Polzin, Austrian League 2006.
Indian Black prefers to take with the f­ Now we finally return to the main
pawn, but here Black immediately line, 10 .id2.
solves the problem of the e7-knight 11 li.Jf6

and has good piece play. That said, At the cost of 'letting go' of the cs­
11 ... gxfs 12 f4 li.Jg6 is also possible. Af­ square, Black induces White to play f3
ter 11 12 f3 li.Jf6 13 li.Jf2 Black before advancing on the kingside. Note
has: that the immediate 11 .. .f4? is a big
strategic mistake because after 12 .ig4
Black will be hard pressed to avoid the
exchange of the light-squared bishops.
Black does have a couple of other ideas
here, though:
a) One possibility is 11 ...'it>h8!?.

bl) 13 ....ih6 14 .ixh6 li.Jxh6 and

now 1S g4 was suggested by Hort, to
prevent the knight from coming to d4,
but after 1s Black looks to be fine
in any case.
b2) 13 ... c6 was Nunn's recommen­
dation. However, 14 li.Jfe4 li.Jxe4 (in- This is Black's main alternative to
stead 14... cxds 1s li.Jxf6+ .ixf6 16 li.Jxds the main lines. If Black wants to avoid

A ttac king Ch ess: The King 's In dian, Vo l u m e 1

the long main lines of 'A' and 'B', the looks too simplistic, but it i s not easy
complications of '(' and the blocked for White to prove an advantage. Mar­
positions of 'D', this is a decent alterna­ kos, for one, thinks enough of this reac­
tive. Instead of fixing the kingside, tion to avoid playing 10 tt:Jd3 alto­
Black makes a useful, semi-waiting gether. After 12 tt:Jxe4 tt:Jfs (12 ... tt:Jf6
move. Playing 11 ... 'it>h8 does a few looks less accurate because Black
things. The king gets off of the a2-g8 would like to recapture on f6 with the
diagonal, which may seem rather queen) it seems as though Black has
vague, but in the main lines we will see lost time compared to the line 11 exfs
a few situations where this could be tt:Jxfs, but there White generally prefers
useful. Black also moves his king off of to play f3, tt:Jd3-f2 and tt:Jfe4. White's
the g-file. This could be good or bad - knight moves have also left him with
we will come across positions where less control of the dS-square, so a ...c6-
after multiple exchanges on g4, a white break may be appealing.
rook or queen ends up sliding over to
h3. Perhaps the biggest advantage to
11 ... 'it>h8 is that the g8-square is
cleared for Black's knight. Playing
... lt:Je7-g8 creates the possibility of play­
ing ....�h6 and clears the h4-d8 diago­
nal for both Black's queen and g7-
bishop. This can be useful if White
adopts a plan involving g4. And, lastly,
by leaving the knight on d7, Black
hopes to make White spend an extra
move preparing the cs-advance. Now 13 f3 tt:Jf6 14 tt:Jdf2 c6!
Perhaps the most natural continua­ (14... lt:Jd4 is also okay) looks fine for
tion is 12 !:!.cl. White hopes to get a Black. One example: 1S .id3 cxds 16
favourable version of the main lines by cxds 'it>h8 17 tt:Jxf6 .ixf6 18 lt:Je4 '1i°b6+
opening the c-file. Now 12 ... tt:Jf6 13 f3 19 .l:tf2 .ih4 20 g3 .�e7 with a good
f4? just allows 14 cs and thus is a game for Black in J.Sofrevski-L.Portisch,
worse version of the main lines, be­ Skopje 1968.
cause Black has played ... 'it>h8 rather A better idea is 13 .ic3, although
than ... gs. In stead Black usually plays this is not too terrifying either: 13 ... tt:Jf6
13 ... cs to try and make the move .l:tc1 14 .if3 lt:Jh4 ( 15 .ixe4 '1i°h4
look irrelevant, while 13 eg8 and 16 .:tel b6 17 g3 looks a little better for
13 ... c6 are altern atives. White) 1S tt:Jxf6+ '1i"xf6 16 .ie4 .�fs 17
b) Another idea is 11 .. .fxe4!?. This '1i°e2 .ixe4 18 '1i"xe4 'iY"fs 19 .l:tael was

Th e M a r d e/ Plata Va riation: 9 li.J e 1 li.Jd7 10 li.Jd 3

agreed drawn here in P.Lukacs­ Usually White just proceeds with

T.Radjabov, Budapest 2000. After his own play, and in Lines A and B we
19 ... .l:tae8 20 f3 'iY"gs 21 '1i"e2 li.Jf5 22 li.Jf2 look at White's traditional continua­
ii.. h6 Black should be able to hold the tions where we have a typical King's
balance. Indian 'race', with both sides trying to
Overall, 11 .. .fxe4 looks very solid, al­ blow the other away. Line C examines a
though it is a little dull. radical idea that has become very
These lines are interesting and look popular of late. Finally, Line D looks at a
quite playable. I think that after learn­ completely different plan where White
ing the main lines, these alternatives looks to slow Black down on the king­
would make a useful addition to a side before pursuing his queenside
player's repertoire, but we will chiefly ambitions. The position becomes
stick to the main line, 11 blocked and the play greatly slows
12 f3 f4 down.
Here 12 ... 'it>h8?! would make little
sense, as White could proceed with 13 A) 13 cs
cs immediately.

This is the main line. White imme­

With the centre closed, the battle diately initiates queenside play, trust­
lines are clear - White will play on the ing in his chances in the ensuing race.
queenside, whereas Black will try to He will decimate the black queenside
attack the white king. Here the way while Black tries to mate the white
divides: king. Although White's play is usually
faster, Black is playing for higher
A: 13 cs gS 14 cxd6 cxd6 1S l:c1 stakes.
B: 13 cs g s 14 cxd6 cxd6 1 s tDf2 13 gs

C: 13 cs g5 14 :C1 liJg6 15 %51? There is no need for Black to be sub­

D:Hg41? tle at this point. White will now bring a

A ttacking Chess: Th e King 's I n dian, Vo l u m e 1

rook to the c-file, but he has a choice as We will see this multi-purpose
to how to go about it. move again and again. Black defends
14 cxd6 the c7-square in advance, and prepares
White immediately opens the c-file. ... 3'.f8 and ... l:tg7. These moves will help
The preparatory 14 l:tcl will be seen in him prepare ... g4. The bishop on f8 is
Line C. often well placed - from there it pro­
14...cxd6 tects the d6-pawn and sometimes the
bishop will come to e7 from where it
could join a kingside attack (with
... 3'.h4 or ...3'.gs), or emerge on the
queenside ( ... 3'.d8).
Note that 16 ... 'ii'b 6+?! doesn't win a
piece; rather it just wastes time be­
cause White is happy to play 17 tllf2
when the bS-knight is suddenly pro­
tected by the e2-bishop.
17 'ikc2

15 !tel
White's main choice, taking the c­
file immediately. Instead 15 tllf2 pre­
pares to bring the king's rook to cl. This
is Line B.

White again threatens tll c 7-e6, so

Black has to retreat his knight, making
the ... g4 advance more difficult to
achieve. Also the tactical shot tllxa7 is
sometimes possible, exploiting the
loose bishop on c8.
1s ...tllg6 16 tllbs 17 ...tlle8 18 a4
The knight threatens to invade on This move is always useful for
c7, from where it will go to e6. White to include. It gives the bS-knight
16...l:tf7 extra support and should Black kick the

Th e M a r def Plata Varia tio n : 9 t'De1 t'Dd 7 1 0 t'Dd3

knight back with ... a6, then t'Da3 will be may also improve the bishop by playing
played when is prevented, and ... 3'.e7. For a while 19 ... 3'.d7 supplanted
White will often thr�aten l'Dc4 and 3'.as 19 ...3'.f8, but I think the two moves are
(or a4-a5), with a grip on the b6-square. of approximately equal value.
If White can play t'Db6, he will be able
to eliminate Black's important light­
squared bishop. Black will often feel
compelled to play ... a6 to drive the
knight away anyway, but usually he
makes some useful moves first
1s hs

Black threatens ...g4 again.

19 C'Df2

20 h3
This prophylactic move is the most
common. Instead 20 b4 is not terribly
useful, but it does set a positional trap.
After 20 ... l:tg7 (Black could consider
20 ... a6) 21 t'Dxa7 Black should avoid
21 ... l:tc7? because of 22 t'Dc6! bxc6 23
dxc6 when White had more than
enough for the piece in N.Rashkovsky­
White prevents ... g4 and increases E.Gufeld, Daugavpils 1978. Instead
the scope of his light-squared bishop. It Black can play 21 ... 3'.d7! with a good
is too early for ...a6 because of t'Da3-c4, version of the pawn sacrifice discussed
followed by as and t'Db6. The e8-knight below.
has to defend c7, so Black has to make 20 'ikb3 is a tricky move order:
a choice between two bishop moves. 20...l:tg7?! (20 ... 3'.d7 transposes to Line
We have: A2, but Black should consider 20...l'Df6
too) 21 l:tc2! when White can omit the
A1:19..:..lf8 prophylactic h2-h3 and double on the
A2t 19.- .t.d:J'rf
• c-file immediately.
White does have an important al­
Al) 19 3'.fS
... ternative in 20 t'Dxa7.
This is the traditional move. Black This grabs a pawn, but Black has a
prepares ... l:tg7, overprotects d6 and counter available which regains the

A ttacking C h e s s : Th e King 's I n dian, V o l u m e 1

pawn if he is willing to go into an end- tion of an eventual 'Llc7-e6.

game. One would think that this would
favour White, but actually Black equal­
izes pretty easily. Still, most black play­
ers have preferred to sacrifice a pawn
here to keep playing for an attack.
Black has:

The importance of this detail can be

seen after 23 ... g4?! 24 fxg4 hxg4 2S
hxg4 'Llf6 26 'Lle7! (much better than
26 .Jte1 'Llh S! which gave Black coun­
terplay in L.Ftacnik-A.Sznapik, Baile
Herculane 1982) 26 ... 'Llxg4 27 'Lle6
a) 20.. .lk7 forces White's reply and 'Llxf2 28 l;Ixf2 .Jtxe6 29 dxe6 'iii> h 8 30
mass exchanges follow: 21 .Jtas l::t xc2 l:tc3 .lte7 31 l::t h 3 when White was
22 .Jtxd8 l::t x e2 23 'Llxc8 l:txa4 (worse is much better in L.Ftacnik-M.Vokac,
23 .. Axb2 24 as! with hopes of an edge) Czech Championship, Frenstat 1982.
24 'Lld3 (24 .Jtxgs l:txb2) 24...g4 2s l:tf2 After 23 ...'ith8 White has tried both 24
l::t e3! 26 'Llel l::t a8 27 l:tfc2 l:tb3 28 'iii> f2 l::t c 3 and 24 as. In both cases Black
l:ta2 29 l::t b1 'iii> f7 30 'ite2 .lte7 31 'Llxe7 should sacrifice a pawn with 24... g4,
'Llxe7 32 'Lld3 l::t a8 33 .Jtxe7 'itxe7 was when in practice he has been rather
completely equal in l.Novikov-1.Glek, successful. This is all very nice, but I
USSR 198S. suspect that 20 ... l::t c7 is sounder.
b) 20....Jtd7!? leads to a more full­ 20 l:tg7

blooded struggle, but it is of course Instead 20 ... .Jtd7 21 'ikb3 l:tg7 will
riskier: 21 'Libs l:tg7 22 h3 'Llh4 23 'ii b 3 likely transpose to Line A2.
'iii> h 8 sees Black patiently make one Also possible is 20...'Llh4. Then 21
more little preparation for ...g4. 'Llxa7 .ltd7 (note that 21...l:tc7? is not
good here, because after 22 .Jtas l:txc2
(seefollowing diagram) 23 .Jtxd8 l::t x e2 24 'Llxc8 l:txa4 2S .Jtxg s !
White wins a pawn and a tempo) 2 2
The main point of this move is to 'Li b s l:tg7 23 'ii b 3 is the pawn sacrifice
get off the a2-g8 diagonal in anticipa- in variation 'b' above.

Th e M a r def P lata Varia tio n : 9 CDe 1 CDd 7 1 0 CDd3

21 'ikb3 ltJc7 f 3 30 3'.xh4 l:txh4 31 'ike6+ 'iii> h8 32

White prepares to double rooks on l:txf3 gives White a big advantage ac­
the c-file. The queen is also useful here cording to Mikhalevski.
for defence - after mass exchanges on b) 22 ... lLif6 ! ? is possible, though: 23
g4, she may slide along the third rank. l:tfcl g4 24 fxg4 hxg4 2 S hxg4 3'.xg4 26
Sometimes White plays l:tc3 (or l:ta3 in 3'.xg4 ttJxg4 27 ttJxg4 l:txg4 has been
certain positions) to utilize a rook in play a few times and seems okay for
this fashion too. Black. Then 28 3'.el f3 ! 29 3'.xh4 'ikxh4
White can also change his mind and 30 'ikxf3 l::t xe4 31 l:tc3 (31 g3 'ikg4 32
take the pawn after all with 21 ttJxa7, 'ikxg4+ l:txg4 was at least equal for
when 21 ... 3'.d7 22 lLibS is again varia­ Black in K.Langeweg-J.Donner, Amster­
tion 'b' above. Sounder is 21 .. Jk7 22 dam 1971) 31 ... l::t el+ 32 l:txel 'ikxel+ 33 which leads to the same endgame 'ikfl 'ikxfl+ 34 'itxfl a6 3 S lLic7 l::t c 8 36
as before with White having the extra tLie6 l:txc3 37 bxc3 b6 38 lLic7 as 39
move h3. This does not necessarily help 'ite2 3'.e7 1/2-1/2 was J.Knap-V.Tasic, cor­
him: for example, 22 ...l::t x c2 23 3'.xd8 respondence 2006.
:xe2 24 ttJxc8 l:txa4 2S ttJd3 g4 26 l::t f2
:e3 27 lLi el g3 was fine for Black in
O.Averkin-G.Kasparov, Moscow 1979.
Another idea is 21 as CDh4
(21 ... 3'.d7!? 22 'ikb3 lLih4 is probably
better) 22 ttJxa7 ! ? l:tc7 (22 ...3'.d7 ! ?) 23
.'Llc6! bxc6 24 dxc6 l:tg7 2 S b4 lLic7 26
bS lLixbs ! ? 27 3'.xbs g4 and now in­
stead of 28 hxg4 CDxf3+! 29 gxf3 't!Vh4,
which was unclear in A.Jakubiec­
K.Chojnacki, Lublin 2009, I suspect that
28 fxg4 hxg4 29 hxg4 3'.xg4 30 ltJxg4 23 ttJa3 lLif6
l:txg4 31 l:tf2 would test Black's play. Another possibility is 23 ...3'.d7,
21...ttJh4 22 l:tc2 a6 which actually transposes to the main
Black plays this now that he will be line of Line A2.
able to force through ... g4. Alterna­ 24 3'.e1
tively: This move sizes up the h4-knight
a) 22 ... g4?! (the immediate advance and clears the second rank for defence.
is premature) 23 fxg4 CDf6 24 3'.el hxg4 24...g4 25 hxg4
2S hxg4 ctJhS was played in V.lvanchuk- Instead 2S fxg4 hxg4 26 hxg4 (26
1.Cheparinov, Sofia 2008. Here 26 CDdl! ttJxg4?! ttJxe4) 26 ...ttJxg4 should trans­
3'.xg4 27 3'.xg4 l:txg4 28 'ikh3 'ikg s 29 pose, but White should certainly avoid

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

the temptation of 27 l:txc8?. 32 l::tf3 3'.e7

He hopes to win two pieces for a This position has been reached a
rook, but Black has the trick 27 ...'Llxf2! few of times in practice and Black has
28 3'.xf2 (28 l:txd8 l:txg2 mate) 28 ... l:txc8 enough play. White's knight is far away
29 't!Vh3 'Llxg2 30 3'.g4 'Lle3 31 3'.xe3 and his pawn structure is a little loose.
'ikd7 and 0-1 in V.Talla-M.Szelag, Us­ One game continued 33 l:th3 l:tg8 34
tron 2008. l:txh4+ 'ikxh4 35 l:tc3 'ikel+ 36 'ith2
2s ...hxg4 26 'Llxg4 'Llxg4 l:txg2+ 1/2-1/2, l.Schuett-V.Conti, corre­
The alternative 26 ...'Llhs may be spondence 1999.
playable, but it looks sketchy to me.
Instead of 27 as 'Llg3 28 3'.xg3 fxg3 29 A2) 19...3'.d7
'ikb6 'it e7 30 l:tfc1 3'.xg4 31 fxg4 'Llg6 32
ilf3 'ikh4 33 �1 'Llf4, when Black had
plenty of counterplay in L.Ftacnik­
Z.Polgar, Trencianske Teplice 1985,
White should play 27 'Llc4!. Now
27 ... 'Llg3 28 3'.xg3 fxg3 29 'ikb6 'ike7 30
'Llce3 l:th7 31 l:tfc1 has been played in a
couple of correspondence games. The
position is messy, but I think Black is
better off with the text move.
27 fxg4 3'.xg4 28 3'.xg4 l:txg4 29 'ikh3
This is a good square for defensive This is a more modem move. Black
reasons as well as for a possible coun­ avoids any 'Llxa7 tricks and sometimes
terattack. However, Black can hold the even plays on the queenside himself. Of
balance. course, the main plan is still a kingside
29...'ikgs 30 3'.xh4 l:txh4 31 'ike6+ 'ith8 attack

Th e M a r def Plata Varia t i o n : 9 tDe1 lDd 7 1 0 CDd3

20 'ikb3 and 'ika3, laying siege to the d6-pawn)

White can also play 20 h3 3'.f8 21 23 ... a6 24 CDa3 bS gave Black counter­
'Wb3 which will transpose. play in S.Zilka-R.Ramondino, Vienna
20 3'.fS
... 2009.
This is very natural. White doubles
rooks on the c-file. Others:
a) 21 h3 and now 21 ... l:tg7 will gen­
erally transpose elsewhere, depending
where White moves his rook, but Black
also has the option of making some
manoeuvres on the queenside:
al) 21 ... 'ikb8!? 22 as l:tg7 23 l:tc3?!
(this leaves the as-pawn without sup­
port; 23 l:tc2 is more natural and could
Black has also tried 20 ...3'.f6 ! ?, which transpose to the next note) 23 ... 3'.e7 24
is not without its logic. Black often l:tal (not a nice move to make) 24 ... 3'.d8
plays ... 3'.f8-e7 and then sometimes 2S l:tccl lLih8 ! ? (a typical idea; Black
either ... 3'.h4 or ...3'.d8, so it makes routes the knight to h6 to support the
some sense to put the bishop on the ... g4-break) 26 ttJa3 a6 27 CDc4 3'.bs 28
h4-d8 diagonal in one move instead of 'iii> fl CDf7 29 3'.el ltJh6 30 lLib6 3'.xb6 31
two. White has: axb6 3'.xe2+ 32 'itxe2 'ikd8 33 CDd3 g4
a) With 21 3'.b4 White decides to gave Black the initiative in M.Roeder­
take advantage of Black's last move to M.Hebden, Bern 1992.
attack d6 immediately, but he should a2) 21 ... 3'.e7!? 22 l:tc2 'ikb8 23 l:tfcl
probably wait on this, as it runs into 3'.d8 24 as a6 2 s ltJa3 bS 26 3'.e1 (26
some tactical problems: 21 ... 3'.e7 22 3'.b4 l:tg7 27 'iii> fl lLif6 28 'ikd3 lLih8 29
:te2 a6 23 CDa3 g4! 24 fxg4 hxg4 2S lLibl CDf7 was similar in M.Vukic­
3'.xg4 3'.xg4 26 ttJxg4 'ikb6+ 27 'iii> h l as D.Sahovic, Tuzla 1983) 26 ...lLif6 27 'ikd3
28 ctJh6+ 'iii> g7 29 ttJxf7 'ikxb4 with l:tg7 28 3'.d1 ctJh8! left ... g4 imminent in
complications that were not unfavour­ A.Olcayoz-E.Grivas, Mangalia 1992.
able to Black in V.Neverov-M.Vokac, b) 21 l:tc3 was played in S.Conquest­
Bled Olympiad 2002. P.Thipsay, British Championship,
b) 21 l:tc2 'ikb8!? (consistent; Black Southampton 1986. It guards the third
wants to play ... 3'.d8 quickly) 22 as 3'.d8 rank (in the event of exchanges follow­
23 l:tfcl (the prophylactic 23 ltJa3! ? is ing ... g4), and leaves the possibility of
another idea when 3'.e2-bs is a possi­ 'ikb3-d1 open, but compared to 21 l::t c 2,
bility and there are ideas like CDc4, 3'.b4 White limits his d2-bishop - an a4-aS

Attacking C h e s s : The K i n g 's I n dian, Vol u m e 1

advance will not be supported and .Jtb4 Without White having spent a
is not possible. tempo on h3, 21 ...'ii b 8?! looks too slow:
c) 21 J::t c4 is a bit strange looking be­ 22 l::t fcl a6 23 'Lla3 .lte7 24 as .Jtd8 2S
cause there is no 'Lla3-c4 possibility: 'Llc4 .Jtbs 26 .ltb4 b6? was G.Kacheish­
21...a6 22 'Lla3 (again, there is no 'Llc4; vili-D.Sharavdorj, Lubbock 2009, and
perhaps White was hoping to use the here White could have played 27 'Llxd6!
rook on the fourth rank with l:tb4 to 'Llxd6 28 .Jtxd6 'ikxd6 29 .Jtxbs axbs 30
pressure the b7-pawn) 22 ...l:tg7 23 as l:tc6, winning the knight on g6.
'Llf6 24 'ikb6 'ike8 2s h3 (2s 'ikxb7? .Jtbs) 22 'Lla3 l:tg7
2S ...g4 26 fxg4 hxg4 27 hxg4 'Llh4 28 22 ...'Llf6 23 h3 l::t g7 is the same.
l::t c7 'ikg6! 29 .Jte1 (or 29 'ikxb7 l::t d 8 30 23 h3 'Llh4
�xa6 ii.xg4 31 l::t x g7+ .Jtxg7 32 'Llxg4 Black has to time his pawn break
'Llxg4 33 .Jtxg4 'ikxg4 34 l:tf2 f3 3S �d3 carefully: 23 ...'Llf6 24 l:tfc1 g4?!
l::t f8 with a decisive attack according to (24...'Llh4) 2S fxg4 hxg4 26 hxg4 'Llh4
Gallagher) 29 ...l::th7 30 'ii b3 (30 'ikxb7 27 l::t c 7! .lte7 28 .Jte1 was much better
'ii h6! 31 'Llh3 'Llxg2 wins) 30 ... 'Llxe4 31 for White in L.Ftacnik-G.Ligterink, Am­
'ikd3? (after 31 'ikxb7 'Llg3! 32 .Jtd3 .Jtfs! sterdam 1977. If 28 ...'Llxg4 then 29
33 l::t xh7 'ikxh7 34 'ikxh7+ .Jtxh7 Black .Jtxg4 .Jtxg4 30 'Llxg4 l::t xg4 31 't!fh3
wins the exchange, but this was proba­ l::t x g2+ 32 'iii> h l wins for White.
bly the best chance) 31 ...'Llg3 32 l::t xb7 24J::tfc1
was Z.Kozul-T. Radjabov, Sarajevo 2003.

24 'Llf6

Here Gallagher points out that Again 24... g4?! is premature: 2S

32 ... .Jtfs! 33 l:txh7 (33 gxfs 'Llxe2+ 34 fxg4 hxg4 26 hxg4 l:tc8 (or 26...'Llf6 27
'ikxe2 'ikxg2 mate) 33 ... .Jtxd3 34 .Jtxd3 l:tc7, transposing to 23 ...'Llf6, above) 27
'itxh7 3S .Jtxg6+ 'Llxg6 gives Black a l::t xc8 .Jtxc8 28 'Llc4 'Llf6 29 'ikh3 with a
clear advantage. clear advantage in N.Rashkovsky­
21 a6
... A.Vitolinsh, Daugavpils 1978.

Th e M a r def Plata Va ria tio n : 9 tDe1 lDd 7 1 0 lDd3

Black also does not need to commit B) 13 cs gs 14 cxd6 cxd6 1S ttJf2

to 24 ... l:tb8 yet. After 2S CDC4 b6 26 as!
g4 27 fxg4 lLif6 28 ttJxb6 hxg4 29 hxg4
.!Lixg2, as in V.lvanchuk-J.Timman,
Hilversum 1991, White could play 30
'itixg2! ltJxg4 31 3'.xg4 3'.xg4 32 �1
3'.d7 33 'ite2 when he should win.
2S 3'.e1
Alternatively, 2S 'ikxb7? runs into
2s ... ttJxg2!, while 2S l;Ic7 can be met by
2s ... l:tb8 26 as ltJe8 27 l:t7c2 lLif6 28 l:tc7
!lJe8 29 l:t7c2 lLif6 when H.Gruenberg­
P.Hesse, Eilenburg 1984, ended in a Before playing 'ikc2 and �fcl, White
draw. shores up his defence of the g4-square.
2s ...g4 1s ...ttJg6
Black has made the necessary Black could also begin with 1s ...h s
preparations and now breaks. t o try t o force White to play h3. I f 16
26 fxg4 hxg4 27 hxg4 3'.xg4 28 3'.xg4 'ikc2 (16 h3 ttJg6 17 'ikc2 �f7 18 l:tfcl
!Llxg4 29 ttJxg4 l:txg4 30 'ikh3 'ikgs transposes the main line), then
16 ... g4! ? (16 ...l:tf7 11 ttJbs ttJg6 18 l:tfc1
ltJe8 19 a4 allows White to delay h3) 17
lLibS g3 is an untried possibility.
16 'ikc2 l:tf7 17 l:tfc1 hs
After 17...ltJe8 18 a4 hS White went
backwards with 19 CDcdl ! ? and won a
long struggle after 19 ... 3'.f8 20 l::t a 3 in
L.Aronian-H.Nakamura, World Team
Championship, Bursa 2010. The text
move avoids this possibility because
then 18 ttJcdl could be met with
This position was agreed drawn in 18 ... g4!.
F.Quiroga-E.Maggiolo, Buenos Aires Black has also tried 17 ...a6, but after
2000. White was quite a bit higher­ 18 a4 hS 19 h3 3'.f8 20 as we transpose
rated here, which shows the health of to 18 ... a6 19 a4 3'.f8 in the note to
Black's position. The position is very Black's next move, which looks unsatis­
similar to the one we saw in Line Al, factory for him.
but here Black has an even better ver­ 18 h3
sion. It is important for White to hold up

A ttacking C h e s s : The King 's I n dian, Vo lu m e 1

... g4 as long as he can. He can, though, 'ika7 2S 3'.as l:tb8, and now White has a
also play the immediate 18 'Libs here. very pleasant choice between 26 3'.c7
After 18 ...'Lle8 19 a4 3'.f8 or 19 ...3'.d7 and 26 g s ! 'Llg4 27 3'.xg4 3'.xg4 28 'Li es!
the play is similar to the main line and (but less good here is 28 'ikc6 3'.e2 !).
may even transpose. b) 18 ...g4!? is a better try: 19 fxg4
hxg4 20 hxg4

18 3'.fS!?

Black stays calm. Practice has also 20...'Lle8! (Black intends ...3'.f6-h4; a
seen him take immediate measures on useful idea to know) 21 a4 3'.f6 22 l:ta3
either flank: 3'.h4 23 'Llcdl 3'.g3 24 l:tc3 (24 'Llh3
a) 18 ... a6 stops 'Li bS, but it does 't!fh4 2s 'Lldf2 'Llf6 26 'ikd1 3'.d7 27 as
weaken the queenside and White has l:taf8 28 3'.el f3 29 l:txf3 'Llf4 30 l:txf4
other ways into the black position after and Yi-1/2 was G.Sosonko-F.Hellers, Wijk
19 a4!. Now 19 ...b6 is considered bad aan Zee 1986) 24 ... 3'.d7 2s 3'.bs?! 3'.xbs
after 20 l:ta3 (with the idea of 'Lla2 and 26 axbs 'ikh4 27 'Llh3 'Llf6 28 l:tc8+? (28
l:tb3 or l:tac3, while 20 'Lla2 is also pos­ gs 'Llg4 29 l:tc8+ l:txc8 30 'ikxc8+ 'Llf8! is
sible) 20...3'.f8 21 'Lla2 ! . The alternatives given by Gallagher) 28 ...l:.xc8 29 'ikxc8+
are supposed to be better, but it all just 'iii> g7 3o'Lldf2 3'.xf2+ 3 1 'Llxf2 f3 ! 32 'ikfs
seems rather shaky to me: 'Llxe4 33 'ikxf7+ (33 'ikxe4 'ikg3)
al) After 19 ...'Llh4 20 'ikdl g4 21 33...'itxf7 34 'Llxe4 'ikxg4 and Black had
fxg4 hxg4 22 hxg4 'Llh7 23 as 'Ll gs 24 a winning attack in G.Andruet­
.i.f1 bS 2S axb6 'ikxb6 26 'Lla4 'ika7 27 V.Spasov, Sofia 1990.
3'.as ! l:tb7 28 l:tc6 l:tab8 29 3'.b6! White 19 'Libs 'Li es 20 a4
was winning in B.Lalic-L.McShane, The thematic 20 'Llxa7?! seems to
Southend 2000. work tactically because the c8-bishop is
a2) Instead 19 ...3'. f8 runs into 20 as loose, but it appears too slow: 20 ... 3'.d7
(with the idea 'Lla4-b6) 20 ... g4 21 fxg4 (the counter 20 ...l:tc7? does not work
hxg4 22 hxg4 bS 23 axb6 'ikxb6 24 'Lla4 well here after 21 3'.as l:txc2 22 .i.xd8

Th e M a r def Plata Va ria tio n : 9 tDe1 lDd 7 1 0 lDd3

.:.Xe2 23 ttJxc8 l:txb2 24 3'.xgs l:taxa2 2 S

.:.Xa2 l:txa2 2 6 l:tb1 and White i s much
better) 21 ttJbs g4! 22 fxg4 hxg4 23
hxg4 f3 ! 24 gxf3 3'.xbs 2s 3'.xbs l:txf3
gives Black excellent play: for example,
26 'ikdl l:txf2 27 'itxf2 'ikb6+ 28 3'.e3
'ikxbs M.Stean-J.Hjartarson, Lucerne
Olympiad 1982, or 26 3'.xe8 't!fh4 27
�gs 'ikg3+ 28 'itf1 l:txe8 29 'ike2 (29
'Wd2 't!Vh 3+ 30 'ite2 'ikg2 31 l:tf1 lLif4+ 32
.i.xf4 exf4 is also winning for Black)
29 ... 3'.e7 0-1, J.Barkhagen-M.Tumer, Sas Black intends ... lLih4, ... a6 and ...lLif6
van Gent 1992. to force through ... g4. White should
20 3'.d7
... play 22 as or double rooks with 22 l:ta3
With the bS-square covered, Black or 22 l:tC3. Instead 22 'itfl?! 'takes ad­
should not ignore White's idea: for ex­ vantage' of the free fl-square, but this
ample, 20...lLih4 21 ttJxa7! 3'.d7 22 ttJbs is not a very good place for the king.
g4 23 fxg4 hxg4 24 hxg4 l:tg7 2S 'ikdl After 22 ...lLih4 23 3'.el 3'.e7 24 l:tc3
.i. e7 26 l:ta3 was much better for White 'ikb8 ! ? (another idea is 24... a6 2 S ttJa3
in Z.Kozul-A.Sznapik, Tbilisi 1988. lLif6 26 l:tacl g4 2 7 hxg4 hxg4 28 fxg4
20...l:tg7 is playable: 21 l:ta3 a6 22 'ike8!?, hitting the a4-pawn and prepar­
:C3 3'.d7 23 ttJa3 lLih4 24 3'.el l:tb8 2 S ing ... 'ikg6) 2S as 3'.d8! (here we see one
'Wdl lLif6 gives Black counterplay. of the points of 24...'ikb8; from d8 the
21 'ikd1 bishop can help defend the queenside
This move tries to clamp down on and it may even become active itself)
the g4-square. Probably White should 26 l:tb3 a6 27 CDc3 (27 ttJa3 3'.a4!)
look at the alternatives, though. 21 l:ta3 27 ...CDC7!? (Black intends to use the bS­
a6! forces the knight back to c3, while square, but a more thematic plan is
after 21 as!? Black should play 21 ...l:tg7, 27 ...lLif6, trying to force through ...g4)
which is the same as the Line A2, ex­ 28 l:tb4 'ith8 29 l:ta3 'ikc8 30 l:tc4 'ikb8
cept that White's rook is on al instead 31 l:tb4 'ikc8 32 l:tab3 ttJbs 33 ttJxbs
of fl. This probably does not change axbs 34 3'.xbs 3'.xas 3S 3'.xd7 l:txd7 36
much, especially as White usually dou­ l:ta4 l:tc7! Black suddenly took over the
bles rooks on the c-file over the next queenside in A.Gavrilov-A.Shomoev,
couple of moves, but White will have Krasnoyarsk 2007.
the option of playing a quick 'ikdl or
�el without locking his rook in on fl. C) 14 l:tc1
21...l:tg7 White clearly will need a rook on the

A t ta c k i ng Ch ess: Th e King 's Indian, Vo l u m e 1

c-file, so he plays this immediately in­ 2 1 l:te1? ! (better i s 2 1 'Lias 'ikd7 2 2 'Llf2
stead of exchanging on d6. gxf2+ 23 l:txf2, although Black still has
14 'Llg6 1s 'Libs!?
... the initiative after 23 ...'LlhS) 21 ...bxc4
This is a very sharp move which has 22 l:txc4 'Llh4 was clearly better for
become popular recently. Of course lS Black in H.Hoeksema-L.Riemersma,
cxd6 cxd6 brin gs us back to Line A. Dutch League 1987.
By delaying the exchange on d6 a2) 19 3'.as gxh2+ (worse is 19 ...'ike7
White is able to create some unusual 20 'Llb6 gxh 2+ 21 'itxh2 l:tb8 22 'ikc2! )
problems. Black can now force White to 20 'itxh2 'ike7 2 1 l:th 1 'Llhs 22 'Llb6 'Llg3
make a positional piece sacrifice or he gave Black good play in M.Yudovich ­
can play normally, although this allows E.Arlind, correspondence 1974.
White to play an original manoeuvre to b) 16 cxd6! (this piece sacrifice is
weaken the black queen side. critical) 16 ... axbs 17 dxe7 'it d 7 (worse is
1s l:tf7
... 17...'ike8?! 18 'ikb3 g4 19 'Lies 'Llh4 20
Instead both 1S ... g4? 16 cxd6 cxd6 fxg4 and White was much better in the
17 'Llc7 and 1S ...'Lle8? 16 3'.as are game F.Berkes-M.Pavlovic, Kragujevac
clearly better for White, but practice 2009) reaches quite a complicated po­
has also seen the critical 1S ... a6, trying sition.
to refute White's play.

White's play certainly seems easier,

Here White has: but Black is up a piece:
a) 16 'Lla3 intends 'Llc4 and 3'.as. bl) 18 'ikb3 'Lle8 with a further di­
However, this move abandon s control vide:
of e4 and Black can play 16 ... g4 imme­ b11) 19 3'.b4 'Lle8 19 'ikb3 'Llxc7 20
diately. After 17 cxd6 cxd6 18 'Llc4 g3 3'.xf8 Lf8 looks unclear.
White has: b12) 19 'Li es 'ikd6 20 'Lle6 (after 20
al) 19 h3 3'.xh3 20 gxh3 bS (better 3'.b4 Black can just play 20 ...'ikb6)
than 20 ... 'ikd7 21 'Llf2! gxf2+ 22 l:txf2) 20 ... 3'.xe6! (much better than 20 ...'ikb6+

Th e M a r def Plata Va riatio n : 9 tDe1 lDd 7 1 0 lDd3

21 'iii> h l 3'.xe6 22 dxe6 ttJxc7 23 e7+ l:tf7

24 3'.c4 lLih8 25 l:tfdl when White had a
big advantage in O.Biriukov-S.Soloviov,
Saint Petersburg 1999) 21 dxe6 ttJxc7
22 3'.b4, and now both 22 ... 'ikb6+ and
22 ...'ikd4+ 23 'iii> h l l:tfc8 are possible,
while earlier 21 3'.b4 'ikb6+ 22 3'.c5 'ika5
23 dxe6 ttJxc7 24 3'.xf8 'itxf8 25 l:tfdl
'ikb6+ 26 'iii> h l ttJxe6 27 'ikb4+ 'itg8 is
unclear according to Soloviov.
b2) 18 3'.b4 was Giri's recent try.
Black has: 16 3'.as!
b21) 18 ...ltJe8 19 3'.xf8 (19 'ikb3 is This is the point of White's play - he
'b11' above) 19 ...3'.xf8 with the idea of forces Black to weaken his queenside .
...3'.d6 was given by Nunn, although This is somewhat annoying, but the
Hoeksma points out that White has manoeuvre costs White some time and
some initiative after 20 'ikb3, intending Black is not without practical chances.
l:tc3 and l:tfcl. Instead 16 cxd6 cxd6 17 'ikc2 would
b22) 18...g4?! 19 CDc5 (19 d6! gxf3 20 again take play back into Line A.
gxf3 b6 21 tLif2 'it h8 22 a3 l:tg8 23 'iii> h l
'ike8 24 l:tgl gives White more than
enough compensation according to
Giri) 19 ... 'ikxc7 20 CDe6 and here, in­
stead of 20...'ikf7? 21 3'.xb5! when
White was much better in A.Giri­
F.Nijboer, Haaksbergen 2009, Black
should have tried 20 ... 'ikb6+ 21 3'.c5
b23) 18...l:tf7 19 'ikb3 3'.f8 20 3'.xf8
'itxf8 21 'ikb4+ 'itg8 22 CDf2 CDe8 23
3'.xb5 'ikd6 24 'ikxd6 ttJxd6 25 a4 'itf8 is 16 ... b6
a line given by Giri, who thinks that This weakens the c6-square, but
Black is okay here. there is no choice.
1s ...l::tf1 17 cxd6!
This is a more normal continuation. This is more accurate than 17 cxb6
Black hopes to transpose to the main when 17 ...axb6 may be a playable al­
lines, but White has an independent ternativefor Black.
course. 11...cxd6

A tt a c k i ng C h e s s : Th e K i ng 's In dian, Vo l u m e 1

Instead 17 ... bxas? is just bad after cause the e4-pawn is not well pro­
18 dxc7 'ikf8 19 'Lies (also good is 19 tected. Another possibility is 18 ... a6 19
'ika4 g4 20 'Lies gxf3 21 3'.xf3 'Llh4 22 'Llc3 (after 19 'Lla3 !? Black should
'ikxas 'Lle8 23 'Lle6 and Black was probably play 19 ... g4, again exploiting
mauled in V.Bhat-D.Aldama, Boca the fact that the e4-pawn lacks protec­
Raton 2008) 19 ... a6 20 'Lle6 'ikb4 (or 19 tion), and here:
'Li es a6 20 'Lle6 'ike7 21 'Llxg7 'itxg7 22 a) 19 ... 3'.d7?! 20 'Llb4 bS 21 'Llc6
d6 'ike8 23 'Lla3 with a crushing posi­ 3'.xc6 22 dxc6 was much better for
tion in M.Mchedlishvili-A.Rustamov, White in D.Pergericht-W.Winterstein,
Tashkent 2010) 21 a3 'ikxb2 22 l:tbl Luxembourg 1987.
'ika2, as in D.Fridman-M.Szelag, War­ b) 19 ...hs?! compares poorly with
saw (rapid) 2009, and now 23 'Lld6 the main line: 20 'Llb4 g4 21 'Lla4 l:tb7?
wins quickly. 22 3'.xa6 l:txa6 23 'Llxa6 bS 24 3'.as !
18 3'.e1 'ikxas 2s l:txc8+ 'iii>h 7 26 'Ll4cS l:ta7 27
This is better than 18 3'.b4 3'.f8 19 'ikel 'ikxa2 28 'Llb4 'ikc4 29 b3! 'ikd4+ 30
'Llf2 ?! (Bhat suggests 19 l:tc6 or 19 'ikc2) 'iii> hl and 1-0 was G.Meier­
19 ... a6 20 'Lla3 hS 21 CDC4 as 22 3'.el S.Bromberger, Zurich 2009.
3'.d7 23 a4 l:tg7 24 h3 'Llh8!?, which c) 19 ... as! is the best move. Black
gives Black good play. After 2S 'Lla3 'Llf7 must cover the b4-square to prevent a
26 'Libs l:tc8 2 7 l:txc8 'ikxc8 28 'Lld3 g4 knight invasion.
29 3'.h4 'Llgs he scored a notable upset
in V.Bhat-K.Sai, New Delhi 2009.

Here White has tried:

cl) 20 'Libs g4 21 l:tc6 3'.f8 22 3'.f2
18 g4!?
... l:tb8 23 'ikc2, although Black had coun­
This is a very principled reply and a terplay after 23 ... g3 24 3'.xb6 gxh 2+ 2S
typical reaction to the white knight's 'itxh2 l:txb6 26 l:txc8 'ike7 27 l:tc7 'Lld7
jaunt to bS. Black takes the oppqrtunity 28 'Llf2 'Llh4 in D.Cummings-B.Sam­
to play this advance immediately be- buev, Toronto 2010.

Th e M a r def P lata Varia t i o n : 9 tb e 1 tll d 7 1 0 tll d3

c2) 20 tllf2 3'.f8 21 tllb 5 h5 22 l:tc6

l:tg7 23 a4 3'.d7 24 'ith1! ? tllh 8 25 l:tc3
g4 26 l:tgl (winning the exchange with
26 tllc7 leads to an unclear position
after 26 ... 3'.xa4 27 'ikcl 3'.d7 28 tllxa8
'ikxa8) 26 ... g3 27 tllh 3 l;Ic8 28 'ikc2 gxh2
29 'itxh2 tll g 4+! 30 fxg4 l:txc3 31 'ikxc3
hxg4 32 g3 gxh3 33 tllc7 was G.Meier­
J.Lopez Martinez, Pamplona 2009. Now
Black's best looks like 33 ...tll g6, with the
idea 34 tlle 6 3'.xe6 35 dxe6 d5! when
Black has the initiative in a very com­ Black is a pawn down and White
plicated position. has made some inroads into his queen­
c3) 20 3'.f2 l:tb8 21 tllb 5 3'.f8 22 l:tc6 side, but he has very good counterplay.
tll e8 23 a4 3'.d7 24 l:tc3 l:tb7 25 b4 h5 26 2 3 l:tc3
bxa5 bxa5 27 tlla7 tllf6 28 tllc 6 'ika8 29 To stop ...3'.e3+.
'ikc2 g4 30 tllb 2 l:tg7 31 tllc4 g3 32 hxg3 2 3 ...3'.f4
fxg3 33 3'.xg3 h4 34 3'.h2 tllh 5 was There are other tempting possibili­
G.Meier-R.Polzin, Austrian League ties as well, such as 23 .. 11g7 ! ? and
2010. White is probably quite a bit bet­ 23 ... tllh 5 ! ?.
ter here, but as usual there is counter­ 24 3'.xf4
play and Black went on to score a big Not 24 3'.f2? 'ikh6! 25 g3 (25 l:te1 'ikh2+
upset against one of the main propo­ 26 'iii> f l 'ikhl+ 27 3'.gl tllh 4 mates)
nents of White's set-up. 25 ...'ikh3, which is winning for Black
19 tll b4 24 ...'Llxf4 25 g3
This is critical. White pounces on
the weak c6-square. Instead 19 fxg4
tll x e4 is unclear. 19 l:tc6 is also worth
investigation, but after 19 ... 3'.f8 at least
there is no knight coming into c6.
A thematic and strong pawn sacri­
fice. We will now follow the game
R.Rusev-A.Diamant, Malakoff 2010:
20 hxg3
20 tllc 6 just gives Black the extra
option of 20...gxh2+. 2 s...'Llh3+
20...fxg3 21 tll c 6 'ikf8 22 3'.xg3 3'.h6! Also good is 25 ...l;Ig7!? 26 'iii> f2 'Ll4h5

A ttacking C h e s s : The King 's I n dian, Vol u m e 1

(or 26 ... 'Llh 3+) 27 'Itel (27 g4? 'Llxe4+) proach. Instead of engaging i n a
27 ... 'Llxg3 28 l:tgl 'Llgxe4!? 29 l::t xg7+ straight race, White takes a stance on
'ikxg7 30 fxe4 'Llxe4 31 'ikc2 'ikgl+ 32 the kingside first. If White can com­
3'.fl 'Llxc3 33 'Llxc3 3'.h3 (33...'ike3+ i s pletely block the kingside, he will have
also interesting) 3 4 'ikd3 3'.g4!? which a free hand on the other side of the
seems promising for Black. board, where he holds a spatial advan­
26 'iii>g2 l:tg7 27 l:th1?? tage. This plan was very popular for
Under pressure White blunders, but White in the late 1980s and prompted
after 27 'ikc2 'Llh s Black has strong Black to look at flexible alternatives to
pressure. 11 ... 'Llf6 12 f3 f4, such as 11 ... 'ith8.
27 ...'Llxe4! 28 'Lle7+ Nowadays this plan is not considered
As 28 fxe4 'ikf2 is mate, White's po­ to be so alarming to Black, although
sition falls apart. the play can become a bit stodgy.
28 ... l::txe7 29 l:txc8 'Llf4+!
Black does not relax, even in a win­
ning position. 29 ...l:txc8 30 l:txh3 a6 is
still much better for him, but the text is
much stronger.
30 gxf4 l:tg7+ 31 '>W1 l:txc8 32 fxe4
'ikxf4+ 33 'Itel l:tcl 34 'Llxd6 l:txdl+ 35
'itxdl l:tc7 36 l:tgl+ 'ith8 3 7 3'.c4 'ikf3+
38 'Itel l;Ixc4 0-1
In conclusion, lS 'Libs is rather dan­
gerous, but White must take risks too
and not everyone wants to commit to 13 ... gs
sacrificing a piece with the white After 13 .. .fxg3?! 14 hxg3 Black does
pieces right out of the opening. If prac­ not even have an advantage on the
tice does eventually show this line to be kingside, so he leaves the position
too dangerous for Black, it will negate closed and plans to open the h-file. The
several decades of opening theory! immediate 13 ... hs?! is bad because af­
That said, although we must be pre­ ter 14 g s 'Llh7 lS h4 Black cannot open
pared for this line, we will not worry the kingside unless he makes an un­
too much just yet. The important thing sound piece sacrifice on gs.
is to be aware of the possibility of l S By playing 13 ...g s Black prevents h4
'Li b s and t o have something i n mind. and will open the kingside himself with
... h s. He also prepares ... 'Llg6, aiming
D) 13 g4!? for the h4-square. White usually plays
This is a completely different ap- 'it g2, so he can contest the h-file, and a

Th e M a r def Plata Varia tio n : 9 tDe 1 CDd l 1 0 lDd3

quick 3'.e1-f2, when the bishop covers Black's idea) 20 ... ttJg3 21 3'.xg3 fxg3 22
the h4-square and also takes aim at the 'ikd2 l:th4 23 'itxg3 3'.d7 24 l:txh4 gxh4+
black queenside. Admittedly, Black has 2S 'iii> h 2 3'.f6 gave Black dark-squared
a narrow field to play with, so he needs compensation for the pawn in
to hold things together on the queen­ B.Gelfand-G.Kasparov, Reggio Emilia
side while gradually building up on the 1991. Kasparov's sacrifice may not have
kingside, looking for tactical opportu­ been completely sound (it is easier to
nities. This often involves sacrifices on say this almost 20 years later!), but the
g4 or e4. idea is still noteworthy.
14... hs 1s h3 ttJg6

14 3'.e1
Both sides have a lot of flexibility 16 cs
here. Instead 14 'itg2 hS lS h3 ttJg6 16 White begins his queenside play.
3'.el transposes to note 'b' to White's Some alternatives:
16th move, below. White can also play a) 16 3'.f2 l:tf7 17 a4 3'.f8 18 'itg2
on the queenside immediately: 3'.e7 19 as 'ikf8! 20 ttJbs 3'.d8 21 ttJxa7
a) 14 cs hs 1s h3 'iii> f7 (1s ... ttJg6 16 hxg4 22 hxg4 3'.xg4! was a typical shot
3'.el leads back to the main line) 16 in H.Fioramonti-0.Cvitan, Geneva 199S .
l:tc1 l:th8 17 ttJbs hxg4 18 hxg4 l:th3 19 b) 16 'itg2 l:tf7 (I prefer this regroup­
'itf2 3'.xg4! 20 cxd6 cxd6 21 fxg4 ltJxe4+ ing to 16 ...'itf7 because Black keeps the
22 'Itel 'ikb6 gave Black excellent play option of a quick ...'ikf8 available) 17
in G.Tallaksen-E.Lie, Gausdal 2007. 3'.f2 3'.f8 and here:
b) 14 b4 hS lS h3 'iii>f7 (1S ...ltJg6 16 bl) 18 b4 l:th7 19 l:th1 3'.e7 20 'ikb3
3'.el l:tf7 is another way) 16 3'.el l:th8 'itg7 21 cs 3'.d7 22 l:tacl a6 (Black
17 'it g2 ltJg6 18 cs hxg4 19 hxg4 lLihS?! avoids 22 ... 'ikh8 23 c6! bxc6 24 dxc6
(an amazing offer, which Gelfand de­ 3'.xc6 2S 'ikc4) 23 a4 hxg4 24 hxg4
clines) 20 l:th1 (20 gxh s l:txhs 21 l:th1 lLih4+ 2S 3'.xh4 l:txh4 26 l:txh4 gxh4 27
l:txh l 22 'itxhl 'ilih8+ 23 'itg1 tests 'ikc4 h3+ 28 'iii> hl (28 'itxh3 't!fh8+ 29

A ttacking C h e s s : Th e King 's I n dian, Vol u m e 1

'itg2 'ikh4 gives Black a strong attack) the possibility of Black recapturing on
28 ... 't!fh8 29 c6 bxc6 30 dxc6 3'.c8 31 d6 with his bishop. The downside to
l:tg1 'Llh7 32 'ikb3 'Llgs 33 'Lids 3'.e6 34 this early exchange is that White gives
bs axbs 3S axbs 3'.xds 36 'ikxds 'ikb8 up some of his queenside space. He
was good for Black in P.Lukacs-E.Grivas, also loses flexibility, since, for example,
Budapest 1993. cS-c6 is no longer an option. After
b2) 18 l:thl l:th7 19 'ikh3 (19 cs 'Llh4+ 17 ... cxd6 18 l:tcl 3'.f8 19 3'.f2 l:th7 20
20 3'.xh4 gxh4 21 b4 'ike8 22 'ith2 hxg4 'ikb3 hxg4 21 hxg4 'Llh4 22 3'.xh4 l:txh4
23 fxg4!? l:tg7 24 3'.f3 'Llh7 2S l:tcl 3'.d7 23 'Llf2 a6 24 a4 l:th7 2S 'itg2 3'.d7 26
26 l:tfl as gave Black good counterplay 'Llb1 bS! 27 axbs axbs 28 'Lla3 'ikb6 29
in S.Knott-M.Hebden, British Champion­ 'LlxbS (after 29 'Llc2 Black would play
ship, Scarborough 2004) 19 ...'Llh4+ 29 ... 'iii> g7, with the idea of ... 3'.e7 and
(Hebden is fond of this knight hop; the ... l:tah8 with an attack) 29 ... 3'.xbs 30
kingside closes up a bit, but White may 'ikxbs 'ikxbs 31 3'.xbs l;Ib8 32 3'.c6 l:txb2
miss his dark-squared bishop) 20 3'.xh4 33 l:tal, as in A.Khalifman-V.Spasov,
gxh4 21 l:tagl hxg4 22 hxg4 h3+ 23 Manila lnterzonal 1990, and now
l:txh3 l;Ixh3 24 'itxh3 'Llhs! ? 2s 3'.d1 3'.e7 33 ... 'Lld7 34 l:tfb1 l:txb1 3 s l:txb1 'Li es 36
26 'itg2 'Llg3 27 'Lle2 'Llxe2 28 3'.xe2 3'.h4 l:tal 'Llb3 with the idea ... 'Lld4 equalizes
29 l:th1 'ikg s 30 'Llf2 'itf8 31 l:th3 'iii> e7 32 according to Khalifman.
'ikdl 3'.d7 33 'ikhl l:th8 gave Black 11 ...3'.fS 18 'itg2 3'.e7 19 3'.f2 'ikf8!
enough compensation to draw in
S.Knott-M.Hebden, British League 2009.

This is a nice set-up for Black. The

d6-pawn remains well protected, from
f8 the queen can go to the h-file and,
17 b4 more importantly, there is latent pres­
White maintains the tension on the sure along the f-file, which may create
queenside. He may also exchange on opportunities for a sacrificial break­
d6 immediately with 17 cxd6 to avoid through on g4 or e4.

Th e M a r def Plata Varia tio n : 9 t'De 1 CDd 7 1 0 CDd3

20 t'D bs .ilds! 24 t'Dc3 repeats moves, but Black chose

This is another typical idea. The to play on with 24... a6 in Z.Gyimesi-
bishop efficiently defends the c7- 0.Cvitan, Ticino 1994.
square and an exchange of pawns on c) A good example of Black's possi­
d6 will also open a route for Black's bilities is 22 l:th l as 23 bxas l:txas 24
bishop to the queenside. cxd6 cxd6 25 t'D b2? (White wants to
21 a4 3'.d7 play l'D c4, but he overlooks a tactic)
2S ... hxg4 26 hxg4 t'Dxg4! (White
thought that everything was covered,
but it turns out that the bS-knight is
not sufficiently supported) 27 fxg4 f3+
28 3'.xf3 3'.xbs 29 3'.g3 (the white
queen is overloaded, since 29 axbs
l:txal 30 'ikxal l:txf3 wins for Black)
29 ...l'D h4+! (now this check is well
timed; the position has opened up and
Black comes in on the dark squares) 30
3'.xh4 gxh4 31 l:th3 3'.d7 (31 ... 3'.a6! pre­
Black has a harmonious position : vents l'D c4 and looks even stronger) 32
a) Grabbing a pawn with 22 cxd6 l'Dc4 l:tcs 33 t'De3 l:tg 7 34 'ith1 'ikf4
cxd6 23 t'Dxa7 fails to the shot 23 ... hxg4 when Black was still better and went
24 hxg4 3'.xg4!. on to win in l.Bedgarini-P.Popovic,
b) Instead 22 t'Dc3 3'.e7 23 t'D bs 3'.d8 Moscow Olympiad 1994.

Chapter 2
The Mar del Plata Variation
9 lbe1 ltJd7 10 f3

1 d4 'Llf6 2 c4 g6 3 'Llc3 3'.g7 4 e4 d6 5 ter 1, White fights for space on the

'Llf3 0-0 6 3'.e2 es 7 0-0 'Llc6 8 ds 'Lle7 9 king side.
'Lle1 'Lld7 10 f3

There i s a big difference here, how­

This move can just be used for ever, because Black has not resolved
transpositional purposes, but in this the tension with ...f4. Although in gen­
chapter we will look at an independent eral Black should be happy about this,
idea. he must not forget that White may be
10...fs 11 g4 able to play on the kingside too. The
Instead 11 'Lld3 transposes to Chap­ famous game Pinter-Nunn did a lot to
ter 1, while 11 3'.e3 transposes to Chap­ dampen enthusiasm for this line from
ter 3. White's point of view, but what is old is
This idea was pioneered by the new and recently there has been some
Hungarian Grandmasters Benko, Pinter revived interest in this variation.
and Lukacs. As in Line D seen in Chap- 11...'ithS

Th e M a r def Plata Variatio n : 9 tb e 1 tll d 7 1 0 f3

seen too as preparation for an eventual

...f4. This advance will make some
sense if White can be prevented from
blocking the kingside. 11 ...'ith8 helps
because the retreat ...tllg 8 will allow
Black to fight for the h4-d8 diagonal
with both his queen and dark-squared
bishop. Note that the other knight re­
mains on d7 for the moment, not only
to leave the f6-square free for Black's
bishop in some positions, but because
This flexible move is now consid­ Black may want to play ... as and ... tll c s,
ered the main line and will be the only increasing the pressure on the e4-
continuation we examine in detail. pawn.
11 .. .f4 is considered to be a mistake Because 11 g4 is becoming popular
because White can keep the kingside again, rather than just examine the
closed with 12 h4, although after aforementioned model game Pinter­
12 ... cs Black's position is very solid. Nunn, we will look at all of White's pos­
Still, because Black cannot open the sibilities in some detail.
kingside, he is basically just trying to
hold the queenside and this way of A: 12 .te3
playing is too passive. 11 ...tllf6 is the B: 12 liJd3
other ma in continuation. After 12 tlld3 C: 12 tllg2
c6 Black increases the tension across D: 12 h4
the board. Then 13 .te3 'iii> h8 is similar
to the main lines, except that both Instead 12 .td2 tllg 8 13 'ikcl pre­
sides have committed themselves - vents any ideas of ... .th6, but is a bit
White with tlld3 and .te3, and Black artificial. After 13 ...f4 Black is ready to
with ... tllf6. begin his kingside play. Meanwhile 12
It is logical for Black to keep the ten­ gs tll g 8 13 h4 transposes to Line D, as
sion and the move 11 ... 'ith8 is almost does 12 'itg2 tll g 8 13 gs f4 14 h4. Fi­
always useful to him. The main idea is nally, 12 'iii> h l tllg 8 13 l:tgl has been
to improve the placement of Black's played several times by the Hungarian
worst piece - the knight on e7. It will IM Kiss. Black can play 13 ...f4, when 14
go to g8 from where it can re-emerge tlld3 gs transposes to Line B, or he can
on f6 to pressure the e4- and g4-pawns. try 13 ... as!?.
Moving the knight also opens up the
h4-d8 diagonal and 11 ...'ith8 can be A) 12 .te3

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

llld 3 l:Ih7 18 .l:thl .Jth4 19 .Jtd1

Statistically this has been the most

popular move, but it is not seen much 19 ...lllfB
nowadays. The bishop goes to an active Often in the Mar del Plata Variation
square, but now ...f4 will come with we see Black play ... llld7-f6 and ... llle7-
tempo. g6. Here Black has already played
12 ...lllgB 13 'ii'd 2 ... lllg 8, so he finds another way to bring
Instead 13 lll d 3 is considered in the a knight to g6.
notes to White's 13th move in Line B. 20 cs .Jtxf2 21 'ii'xf2 gs 22 cxd6 cxd6 23
There is an alternative in 13 lll g 2 f4 l:Ic1 .Jid7 24 .Jta4 .Jtxa4 2s lll xa4 lllg6
14 .Jtf2 hS (Black could also try 14 ... g s Black certainly had no problems
with the idea 1 s h4 h S!) 1s gs (after 1 s here in D.Jacimovic-M.Vukic, Kastel
h3 .Jtf6 16 b4 l:.f7 1 7 cs l:Ih7 18 'ii'd3 Stari 1988.
lllh 6 19 cxd6 cxd6 20 lllb s lllf7 21 h4 a6
22 lllc3 gs Black tore open the kingside B) 12 llld 3
in S.Mirovshchikov-S. Soloviov, St Pe­
tersburg 2002) 1s ...'ii'x gs 16 .Jih4 'ii'h6
11 lllbs gs 18 .Jtf2 llldf6 19 lllxa .Jth3
20 lllxa8 .i::tx a8 was A.Lesiege­
J.Fedorowicz, New York 1993. White
has won the exchange, but Black's
kingside play is still very dangerous.
13 ...f4!
This advance is usually called for
when White cannot play h4. Here Black
has not committed to ... gs yet, so he
can still use the h4-d8 diagonal. 12 ...lllgB
14 .Jtf2 hS 1S h3 l:.f7 16 @g2 .Jtf6! 17 Black continues with the standard

The M a r def Plata Varia t i o n : 9 t:D e 1 t:D d 7 1 0 f3

plan. Another idea is to block the (13 .. .f4 14 j_f2 hS is also possible and
queenside and activate the d7-knight then 1S h3 j_f6 is similar to Line A)
by going to cs. After 12 ... as 13 j_e3 b6 Black has scored quite well. Compared
14 'ii'd 2 (14 a3 t:Dcs 1s b4? walks into to 12 j_e3 t:Dg8 13 t:Dg2, White does not
1s ... axb4 16 axb4 l:Ixa1 17 'ii'xa1 t:Dxd3 control the h4-square and after 14 j_f2
18 j_xd3 fxg4) 14 ... t:Dcs 1s t:Dxcs bxcs j_gS Black has improved the scope of
16 @h1 j_d7 17 .i:Ig1 f4 18 j_f2 t:Dc8 19 his bishop.
h4 j_f6 20 'ii'e 1 t:Db6 Black was very 13 ...f4
solid in E.L'Ami-S. Shyam, Dieren 2009. Again the plan with 13 ... as is a solid
Black can also play the immediate alternative. After 14 l:Ig1 f 4 1S j_d2
12 .. .f4. Because White's 12th move did j_f6 16 a3 b6 17 b4 j_h4 the game
nothing to influence the h4-square, K. S akaev-V.Bologan, European Club
Black immediately closes the kingside, Cup, Ohrid 2009, was soon agreed
after which the e7-knight will often go drawn.
to g6. Now 13 h4 t:Dg8 14 'ii'e 1 j_f6 1 S 14 �g1
g s j_xg s (Black could also try 1s ...j_e7!?
with the idea of ...h6) 16 hxgs 'ii'x gS+ is
a draw. More interesting is 13 l:If2!?
j_f6 14 l:Ig2 j_h4 1S b4, which was
played in P.Eljanov-1.Cheparinov, Jer­
muk 2009, and here Golubev suggests
1s ... gs 16 cs t:Df6, with the idea of ... h s
and ...t:Dg6.
13 @h1!?
White's kingside manoeuvres indi­
cate that Black should switch plans.
After 14...j_f6 1S b4 j_h4 16 cs t:Ddf6 17
j_b2 Black's pieces do not coordinate
well and White can play on all parts of
the board. Then 11 ... j_d7 18 a4 h S?! 19
t:Dxf4! exf4 20 g s was V.Belov­
A.Grischuk, Russian Team Champion­
White's next couple of moves indi­ ship 2009, where White won back the
cate that he may have ambitions on piece and kept the initiative.
the kingside himself. After 13 j_e3 j_h6 15 j_d2 hs 16 h3 l:If6 17 I:tc1 l:Ih6 18

A ttacking C h e s s : Th e King 's I n d i an, Vo l u m e 1

@g2 .Jtf8 19 b4 tfJe7 20 cs tfJg6 21 cxd6 ered in the notes to White's 13th move
.Jtxd6!? in Line A.
Perhaps Black was concerned about 13 ... as
21 ... cxd6 22 tfJbs, although this sortie Because White is making prophylac­
does not achieve much after 22 ... hxg4 tic kingside measures, Black goes for
23 hxg4 CiJh4+ 24 @f1 tfJf6 2S CiJC7 .U.b8 the queenside plan. It will not be easy
because White's remaining pieces are for White to keep control of Black's
not very active. Still, the capture with counterplay across the whole board.
the bishop is quite viable. 14 .Jte3 tfJcs
22 tfJbs tfJf6 23 tfJf2 .Jtd7 24 a4
This was 1.Cheparinov-A.Fedorov,
Khanty Mansiysk 20os, and in this bal­
anced yet dynamic position a draw was

C) 12 tfJg2

15 .U.b1
This does not give White anything,
but other moves have also failed to
trouble Black:
a) 1s gs f4 16 .Jtf2 h6 opens the
kingside before White is ready for it.
b) 1S Itel 'ii e7 16 a3 fxg4 17 fxg4
White shores up his kingside, but I:txf1+ 18 @xf1 (it is unappealing to
this move is somewhat passive. capture this way, but it is usually nec­
12 ...tfJgB essary, as here, to avoid ...tfJf6 forking
Black can also play 12 ... as 13 h4 tfJcs White's e4- and g4-pawns) 18...tfJf6 19
14 .Jte3 tfJg8 and in fact this was the .Jtf3 .Jtd7 20 b4 axb4 21 axb4 tfJa4 was
actual move order of Pinter-Nunn. pleasant for Black in T.Markowski­
However 12 ...tfJg8 is both more com­ S .Dolmatov, Polanica Zdroj 1993. White
mon and more flexible, so we will con­ is beginning to look overextended.
sider it the main line. c) 1s 'ii d2 b6 16 exfs gxf s 17 gs f4
13 h4 (this gives up the e4-square, but Black
The alternative 13 .Jte3 was consid- will get many dark squares in return)

Th e M a r def Plata Va riati o n : 9 lb e 1 lb d 7 1 0 f3

18 .Jtxcs bxcs 19 lbe4 lbe7 20 .Jtd3 lbfs 17 a3?!

gave good play in A.Chernuschevich­ This is a mistake, but it is not clear
J.Rowson, Bratislava 1993. how White should continue. For exam­
d) 1S 'ii' c 2 .Jtd7 16 .U.ad1 b6 17 .Jtf2 ple, after 17 'ii'd2 fxg4 (the immediate
'ii'c8 18 g s h6 19 @h2 'ii'd8 (Black's little 17 ... lbf6!? may be even stronger) 18
queen moves annoy White; first the g4- fxg4 lbf6 White is forced to part with
pawn is attacked and then when it ad­ his dark-squared bishop with 19 .Jtxcs,
vances, the black queen takes aim at it when Black can play 19 ...bxcs or even
again) 20 exfs .Jtxfs 21 'ii'd2 @h7 22 19 ... dxcs with the idea ... lbe8-d6.
.Jte3 'ii'd7 23 .U.g1 hS was very comfort­ 11 ... a4!
able for Black in B.Rumiancevas- This is a useful device to be ac­
1. Schutt, correspondence 1996. quainted with - not only in this varia­
e) 1S a3 fxg4 (Black could also try tion, but in the King's Indian in general.
1s ... a4 16 .Jtxcs dxcs 17 lbxa4 fxg4 18 White's knight on g2 is a long way from
fxg4 I:txf1+ 19 @xf1 lbh6!? 20 lbc3 lbf7 controlling d4 and Black takes full ad­
21 @g1 lbd6 when White's position vantage.
looks a bit airy) 16 fxg4 l:txf1+ 17 @xf1 18 b4 lbb3 19 lbbs
lt:lf6 18 .Jtf3 h S!? (a notable idea; Black This prevents ... lbd4, but it leaves
grabs control of g4 for his pieces) 19 White's pawn structure looking vul­
gxh s gxh s 20 b4 (20 .Jtxhs lbcxe4) nerable.
20... axb4 21 axb4 I:txa1 22 'ii'xa1 lba6 19...lbf6
23 bS lbcs 24 'ii'd1 lbg4 1/2-1/2 A.Gipslis­ Also good is 19 _.fxe4 20 fxe4 I:txf1+
D.Lapienis, Parnu 1982. White was 21 @xf1 (else 21 ...lbf6) 21 ...lbf6 22 .Jtf3
much higher-rated here and he surely 'ii' c 8! 23 g s lbQ4 24 .Jt g 1 (to cover h2)
realized his position was beginning to 24. ..h s and Black's firmly entrenched
look rather overextended. knights give him the advantage.
1s... .Jtd1 16 b3 b6 20 exfs gxfs

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

This is exactly what Black wants. rest of the moves of the classic game
The position begins to open up and J.Pinter-J.Nunn, Thessaloniki Olympiad
White's kingside pawn advances just 1988, were: 28 llle2 lllxe4 29 j_xe4 f3
look weakening. 30 lllef4 fxg2 31 lllx g2 'ii'h s 32 'ii'd 3
21 lllc 3 j_g4 33 l:Ie3 'ii'f7 34 'ii'd2 'ii'g 7 35 .U.d3
Instead 21 gs f4! 22 gxf6 j_xf6 23 I:tf7 36 l:Ie1 I:taf8 37 llle3 I:tf4 38 lllg 2
j_f2 l:Ig8 24 @h2 l:Ixg2+! 2 s @xg2 j_xh4 l:I4f7 39 llle3 j_h5 40 I:tf1 I:txf1+ 41
26 @h1 'ii'f 6! gives Black a winning at­ lllxf1 I:tf4 42 'ii'e 1 j_d4+ 43 @g2 'ii'e 5 44
tack. lllg 3 j_g4 45 bS j_f2 46 'ii'xf2 j_h3+ 0-1.
This is another typical King's Indian D) 12 h4
idea. Black sacrifices a pawn for control
of the dark squares.
22 gs
22 fxe4 lllxg4 is much better for
22 ...lllh s
This unleashes an attack on the c3-
23 fxe4 f4
This further pawn sacrifice gives
Black a good position, but 23 ...lllg 3 24
I:tf3 fxe4 may have been even stronger. This committal continuation has re­
24 j_d2 lllxd2 25 'ii'xd2 'ii'e 8 26 j_f3 cently become popular. White is not
lllg 3 27 I:tfel j_e5 simply trying to block the kingside;
rather he has his own ambitions on
that side of the board.
12 ... lllgB
Black continues his plan. Again try­
ing to hold up the queenside with
12 ... as is also possible. After 13 llld3 b6
14 @g2 lllcs 15 I:th1, rather than the
1s ... j_d7 16 j_e3 lllg 8 of M.Rodshtein­
V.Bologan, Moscow 2008, when 17 h S!
gives White a big advantage, Markos
suggests 1s ... h S!?.
For the pawn Black has the bishop­ With White making pawn moves on
pair and a grip on the dark squares. The the kingside, creating more tension

Th e M a r def Plata Va ri a t i o n : 9 t:D e 1 t:D d 7 1 0 f3

with 12 ... c6!? is also logical. 13 @g2 (13 the queen may go t o the h-file and she
.!Z)g2 t:Df6 14 a4 stops any ideas, may support the gs-pawn. With this
but the position became rather sterile move White can also assist his queen­
after 14 ... as 1s l::t f2 j_d7 16 j_e3 l::t c 8 17 side play - the queen supports a possi­
l:Ia3 cs in V.Laznicka-H.Nakamura, ble cs-advance and even attacks the
European Club Cup, Ohrid 2009) a7-pawn. White has done well from
13 ...t:Df6 14 t:Dd3 bS gave rise to im­ this position in practice, but that is
mense complications in P.Eljanov­ mainly because Black has been ill­
T.Radjabov, Elista 2008. prepared for a thematic sacrifice. De­
13 gs spite the closed nature of the position,
White initiates kingside play him­ Black should play concretely, paying
self. Instead 13 t:Dg2 transposes to Line attention to White's possibilities.
c. 11 ...l::t h 7
13.. .f4 Black readies himself for the open­
This cuts off White's kingside ing of the h-file. The downside to this
pawns. 13 ...h6 has also been tried, but I move is that the f4-pawn lacks support.
do not like the look of 14 exfs gxfs 1s It is indeed protected already, but often
f4 when Black's minor pieces are White will sacrifice a piece on f4, when
cramped. his pawn chain can become rather
14 @g2 ominous and smother the cramped
black minor pieces. There are some al­
ternatives, but they do not change the
essence of the position. Black must be
alert to White's sacrificial ideas.
A typical example of what to avoid
went 17 ...hxgs 18 hxgs+ l:Ih7 19 l::t x h7+
@xh7 20 �2 j_e7 21 t:Dxf4! (Black must
always be wary of this type of sacri­
fice!) 2 1 ...exf4 22 j_xf4 .!Z)es?! 23 j_e3
t:Df7 24 f4 with a winning position for
White in L.Polugaevsky-V.Arbakov,
14...h6 U SS R 1986.
Black nibbles away at White's far­ Instead after 17...@g7 18 @fl j_e7
advanced pawns. Instead 14...l::t f7 1S 19 j_d2 White is willing to sacrifice a
.!Z)d3 (or 1S l::t h 1 j_f8 16 t:Dd3) 1s ... j_f8 pawn because he will control the open
16 l:Ih1 h6 transposes to the main line. lines on the kingside and be able to
15 l::t h l l::tf7 16 t:Dd3 j_f8 17 'ii'gll play on the queenside as well. Here
This is a strong manoeuvre. From g1 19 ... hxgs 20 hxgs j_xgs 21 .!Z)bs and

A ttacking C h e s s : Th e King 's I n d i an, Vo l u m e 1

then both 21 ...lbf8 22 cs a6 23 lba3 the f4-pawn as well. After 19 gxh6

�6? 24 lbc4 Ith7 2S .Jtc3 .Jih3+ 26 @e1 I:txh6 the manoeuvre ... ltJgf6-hS is in
.U.c8 27 @d2 lbd7 28 lbf2 and 1-0 in the air and Black can pick up the h4-
J.Carron-J.Rosenthal, Winterthur 2008, pawn at his leisure.
and 21 ... .Jth6 22 cs a6 23 lba3 lbdf6 24 19...exf4 20 .Jtxf4 Itf7!
lbc4 lbe8 2s lbcxes! dxes 26 lbxes I:tf6 20...lbes? 21 .Jte3 hxgs 22 hxgs
27 .Jtc3, 1 .Cheparinov-D. Stellwagen, .Jih3+ 23 @e1 was another disaster for
Amsterdam 20os, are additional ex­ Black in P.Kiss-M.Ricci, Eger 1994.
amples of what Black should avoid. 21 .Jte3 hs!
18 @f1 This move not only keeps the king­
side closed, it secures the g4-square,
which in tum helps Black secure the
es-square. Similar is the immediate
21 ... lbes, as played in F.Khairallah­
T.Cali stri, Cannes 2007. Black has the
same idea of playing ... hs.
22 @e1
22 f4 can still be met by 22 ... lbes
(Black could also try 22 ....Jtf8 23 @e1
.Jt g7) 23 @e1 (2 3 .Jtd4? I:txf4+) 23 ... lbg4.
22 ...lbes 23 'ii'g 3 .JtfB 24 lbbs
This is actually a critical position.
Black must be careful or a piece sacri­
fice on f4 could blow him away.
18 ....Jte7!
Black should keep the tension
rather than open lines towards his own
king. 18 ...hxg s 19 hxgs I:txh1 20 'ii'x h1+
@g7 21 'ii'h4 .Jte7 22 lbxf4! exf4 23
.Jtxf4 lbes 24 .Jte3 has been seen sev­
eral times with terrible results for
Black The position is very similar to
Polugaevsky-Arbakov, above. This was P.Van Hoolandt-J.Cabrera
19 lbxf4 Trujillo, Cannes 2007. Here Black
If White does not play this, then should play 24... c6! to open the posi­
Black will just take on gs, which will tion when the extra piece should count
not only gain material, but will protect for more than White's two pawns.

Chapter 3
The Mar del Plata Variation
9 lbe1 tt:Jd7 10 ..te3

1 d4 lllf6 2 c4 g6 3 lllc 3 .Jtg7 4 e4 d6 S bishop, not only with .. .f4, but likely too
..'Lif3 o-o 6 .Jte2 es 7 o-o lllc6 8 ds llle7 9 from a ... g4-g3 advance in the form of a
lllel llld7 10 .Jte3 pawn sacrifice. In this variation White's
queenside attack is very straightfor­
ward, but the defence of his king can
be a bit more complicated.
Some consider this line to be a bit of
a suicidal variation for White. Indeed
Black's kingside attack can be very
strong, but both players must develop
a feel for attack and defence to be able
to enjoy success from either side of the
board in these lines.
10...fs 11 f3 f4 12 .Jtf2 gs
Although the 10 llle1 fS 11 llld3 and
10 f3 fS 11 g4 lines of Chapters 1 and 2
respectively have had their own kind of
resurgence in recent years, there is no
doubt that the aggressive 10 .Jte3 is the
main line of the 9 lll e 1 complex. From
e3 the white bishop points directly at
Black's queenside, not only supporting
c4-c5 but also targeting the a7-pawn.
This disadvantage is that Black's pawn­
storm will gain time by attacking the

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

The last few moves have been self­ a) 1 S 'ii'd2 h S 1 6 lllg 2 (16 j_e3 j_f6
evident, but now White faces a broad with the idea of ... h4 is given by Nunn),
choice. and now rather than 16 ... as 17 llla 4!?
b6 18 a3 h4 19 g4 lllf4 20 @h2 lllf6 21
A� 13 ti4;
: � '· " " '
llle3 with an edge for White in
9tl.3.l£dJ . S . S avchenko-A.Fedorov, Nikolaev 1993,
C;,t.J. 'llbs Black could play 16 ...h4 17 g4 lllf4 18
Q::ti @h2 lllx g2 19 @xg2 l:.f7 20 b4 lllf8,
·E:U U. which looked okay for him in M.Socko­
K. Kachiani Gersinska, Wuppertal 1998.
Instead 13 g4 is rare, but not neces­ Here Bologan's suggestion of 16 ...j_h6
sarily bad. White tries to play alo.n g the 17 j_e3 @h7 18 b4 as 19 a3 h4 with
lines of Line D of Chapter 1 and all of counterplay also looks sufficient.
Chapter 2. Now should Black play b) With 1S lllg 2 White would like to
13 ... hS 14 h3, his play is slower than in bring the knight to fs. Now:
those similar lines we have examined bl) 1s ... hs 16 llle3 as (or 16 ... h4 17
because White has got his bishop to g4 lllf4 18 @h2 as 19 lllfs j_f6 20 'ii'd2
the f2-square very quickly, where it is b6 21 b3 lllcs 22 .i::tab1 j_d7 23 j_d1
useful for both pressuring the queen­ 'ii'c8 24 a3 'ii'a6 with an unclear posi­
side and defending the h4-square. The tion in D.Ruzele-R. Speckner, Regens­
f4-square is weaker than before, how­ burg 1998) 17 @h2 lllc s 18 llla 4 b6 19
ever, and taking on g3 is now quite vi­ lllxcs bxcs 20 'ii'c2 h4 21 g4 j_d7 22
able. Indeed, Black can quickly initiate lllfs j_f6 was V.Dydyshko-L.Van Wely,
kingside play with 13 .. .fxg3! 14 hxg3 Yerevan Olympiad 1996. The position is
lllg 6. messy, but looks satisfactory for Black.
b2) 1s ... as (Black secures the cs­
square for his knight) 16 llle 3 (White
can get the knight to fS, but this is not
fatal; Black has squares too) 16 ... lllcs 17
b3?! (better was 17 'ii'c2 because now
the c3-knight is loose and Black makes
use of this) 17 ... lllf4! 18 j_e1 (instead
18 gxf4 exf4 19 'ii'd2 fxe3 20 j_xe3
j_es ! is nice for Black - if 21 j_xgs?
j_xc3 ! ) 18.. ll.f6 ! ? 19 gxf4 gxf4 20 lllg 2
l:Ig6 21 I:.f2 @h8! and Black had a
Black takes aim at the f4-square. strong attack in N. Stanec-G.Timo­
Now ... hS-h4 is one idea. White has: shenko, Vienna 1998.

Th e M a r def Plata Varia tio n : 9 lbe1 lb d 7 1 0 iL e 3

A) 1 3 b4 lbg6 14 as l:If7 1s b4 lbf6 16 cs which is

discussed in the notes to Black's 13th
move in Line E.

This obvious preparation for cs has

fallen out of favour since the classic
game Piket-Kasparov. However, it is not Black should probably prefer 1s ...h s
completely harmless and Black should and then:
not take it too lightly. a) 16 as (16 cxd6 cxd6 17 as g4 is
13 lbf6 14 cs lbg6 15 cxd6
... the same) 16 ... g4 17 cxd6 cxd6 18 lbbs
After 1S .l::t c1 .l::tf7 16 cxd6 play sim­ g3 19 hxg3 (19 il..x a7! ?) 19 ...fxg3 20
ply transposes, while 16 a4 gives Black il.. x g3 a6 21 lba3 h4 22 il..f2 lbhs 23 f4
the opportunity to capture on d6 with lbhxf4 24 lbc4 'ifgs gave Black an at­
his bishop after 16 17 cxd6 tack in L.Ftacnik-1. Smirin, Biel Inter­
il..xd6! ?. This has been tried a few zonal 1993.
times, but actually I prefer 17 ...cxd6, b) 16 h3 l:If7 17 as il..f8 18 c6 .l::th 7!
transposing to the main line. (Black has plenty of support for ...g4, so
Instead 1S a4 is important. Black the rook is better off on the-soon-to-be­
has to be careful here or else he could opened h-file) 19 cxb7 (19 bS b6! 20
be tricked into a line that he may not axb6 cxb6 21 .l::ta3 'ii' c7 22 lbd3 g4 23
be prepared to play. fxg4 hxg4 24 hxg4 'ii' g 7 2S il..xb6 'ii'h 6
gave Black a strong attack in l.Hausner­
(seefollowing diagram) S .Dolmatov, German League 1993. de­
spite his ugly queenside) 19 20
Gallagher points out that the natu­ bS (or 20 a6 il.. c8 21 lbbs g4) c8
ral 1 s ... .l::tf7 may well be unnecessary with counterplay.
when White has not played .l::t c 1 be­ c) 16 c6 il..h 6 (Black could also try
cause the c7-square does not need ad­ 16 ...aS!?) 17 bS b6 18 as g4 (unfortu­
ditional defence. Indeed, after 16 as we nately 18 ... .l::tb8 19 axb6 axb6 20 'ii'a4
have transposed to the risky line 13 a4 g4 2 1 'ii'a7 gxf3 22 'ii'xb8 fxe2 23 lbxe2

A ttacking C h e s s : Th e Kin g 's I n d i an, Vol u m e 1

lllxe4 24 l::t a7 does not look sufficient) j_e1 h3 3 0 gxh 3 'ii'h4 31 lllf2 @h8 -
19 axb6 cxb6 20 fxg4 hxg4 21 g3 l::t f7 31 ... 'ii'h s runs into 32 j_c8! - 32 �c8!
was unclear in V.Korchnoi-Xie Jun, .U.xc8 33 j_xc8 j_e7 34 lllh 1! saw White
Marbella 1999. defend in J.Piket-P. Paneque, Adelaide
1s...cxd6 16 Itel l::tf7 17 a4 1988) 24 j_b6?! 'ii'f8 2S j_e2 h4 26
@xh2 lllh s 27 llld3 j_f6 28 .U.g 1 lllg 3 29
'ii'd2 j_d8! gave Black good play in
G.Burgess-W.Watson, British Champi­
onship, Plymouth 1989.
18 as j_d7 19 lllb s
This allows Black to execute his
main idea. Instead 19 j_bS j_c8 20 j_e2
repeats (if this is too off-putting then
there is 17 ... h s), while White has also
tried 19 @h1 l::t g 7 (19 ...'ii'e8 was sug­
gested by Kasparov), and here:
17 ...j_fB! a) 20 lllb s g4 21 lllxa7 g3 22 j_b6
This is Kasparov's clever idea. For 'ii'e 8! 23 .i::te7 gxh2 24 l::t x b7 lllh s gave
the moment Black leaves the h S-square Black some attacking chances in
open for his knight. The alternative is G.Burgess-B.Badea, Prestwich 1990.
the immediate 17 ... h s which also b) 20 j_bS g4 21 j_xd7 'ii'xd7 22 fxg4
seems playable. After 18 as Black has: lllxg4 23 lllf3 j_e7 24 j_g1 lllh 4 was
a) 18... g4 19 lllb s g3 20 j_xa7 lllh 7 good for Black in D.Gurevich­
21 @h1 (to make room for j_g1) H.Gruenberg, New York 1991.
21 ...l::t xa7 22 .U.xc8 'ii'xc8 23 lllxa7 'ii'd 8 19...g4
24 h3 lll g s (White was much better
after 24 ...'ii'M 2s 'ii'c2 lt:lgs 26 j_bs lllf8
27 @g1! in L.Bass-K. Spraggett, New
York 1983) 2s 'ii'a4 'ii' b 8 26 lllbs lllx h3
27 llld3 (27 gxh3? 'ii'c8 wins) 27 ...lll g s
28 l:Ic1 'ii'd8 29 'if c2 lllf8 30 'ii'c8 'ii'f6 31
@g1 l::t d7 and Black developed coun­
terplay in L.Ftacnik-T.Oral, Slovakian
League 199S.
b) 18...j_d7 19 lllb s j_xbs 20 j_xbs
g4 21 @h1 g3 22 j_g1 gxh2 23 j_f2 a6
(23 ...h4 24 @xh2 lllh s 2s �g1 lllg 3 26 20 lllc 7?!
a6 bxa6 27 j_xa6 j_f8 28 llld3 l::t h 7 29 This appears to be too greedy, but

The M a r def Plata Variati o n : 9 t:De1 tiJd7 1 0 iLe3

as Bareev once pointed out, the pieces B) 13 CiJd3

move differently when one is playing
against Kasparov.
White can try to bail out with 20
fxg4, although Kasparov's suggestion
20 ... t:Dxe4 21 t:Dc7 il..a4 22 'ii'xa4 �xc7
looks at least equal for Black.
Instead 20 t:Dxa7 g3 21 il.. b6 'ii'e 7 22
'it>h 1 t:Dh 5 23 t:Db5 'ii'h 4 (23 ...gxh2 24
fi.f2 t:Dg3+ 25 il.. x g3 fxg3 with a clear
advantage for Black is given by Kas­
parov) 24 il.. g 1 I:tf6 25 t:Dc7 t:De7 26
t:Dxa8 ?! gxh 2 27 il..f2 'ii'xf2! 0-1 was This old move seems to mix up sys­
K.Friesen-M.Lomineis hvili, Rotterdam tems, but t:Dd3 is almost always a use­
1998. Here White could look into 26 ful move, so it can hardly be that bad.
CiJd3, which was suggested by Galla­ 13 ... t:Df6 14 cs t:Dg6 15 a4
gher. The idea is to meet the move Instead 15 cxd6 cxd6 could easily
26 ...I:th6 with 27 'ifel or 26 ... gxh2 with transpose just about anywhere, but
27 fi.f2. there is no benefit for White in giving
20 ... g3! 21 t:Dxa8? up all of his flexibility - see too note 'a',
This just carries things too far. In­ below.
stead 21 hxg3 fxg3 22 il.. xg 3 il..h 6 15 Itel is a valid alternative, how­
(22 ... t:Dh 5) 23 t:Dxa8 t:Dh 5 2 4 fi.f2 t:Dgf4 ever. Black should reply 15 ...Itf7.
25 t:Dd3 Itg7 26 t:Dxf4 fi.xf4 27 g4 il.. xcl
28 'ii'xcl t:Df4 29 'ii' e3 h5 with an attack
is given by Kasparov, but this was still a
better try.
21 .t:Dhs 22 @h1

22 fi.xa7 'ii'h 4 23 h3 fi.xh 3 24 gxh3

'ii'xh 3 25 .U.f2 gxf2+ 26 @xf2 t:Dh4 27
ii.fl (27 t:Dd3 'ii'g 3+ 28 @f1 t:Dg2!)
27...'ii'h 2+ 28 t:Dg2 Itg7 looks very
22...gxf2 23 I:txf2 t:Dg3+ 24 @g1 'ii'xa8
Black is much better. The game Now 16 'iih 3 allows 16 ... g4 immedi­
J.Piket-G.Kasparov, Tilburg 1989, fin­ ately, so White might prefer:
ished 25 il..c4 a6! 26 'ii'd 3?! 'ii' a 7 27 b5 a) 16 cxd6 is the most common sta­
axb5 28 il..x b5 t:Dh1 ! 0-1. tistically, but it is never played by

A ttacking C h e s s : Th e King 's I n dian, Vol u m e 1

Grandmasters. When White gives up nov-M.Najdorf, Zurich 19S3.

his queenside flexibility, it makes i t
easier for Black to formulate h i s plans.
After 16 ...cxd6, 17 lllb s, for example,
allows 17 ... g4 with good play.
b) White has also tried 16 a4 .Jtf8 17
as hS 18 h3 which can be met by
18 ...l:Ig7 or even 18 ...g4.
c) The immediate 16 lllb S can, of
course, be met with 16 ... g4. By now we
can see a kind of relationship between
the bS- and g4-squares. If White plays
lllbs too early, Black can often play ... g4 e) 16 @h1 is often a useful prophy­
without further preparation because lactic move, but here 16 ....Jtf8 17 a4 h S
the e4-pawn is loose. After 17 a4 g3 18 1 8 a s g4 1 9 cxd6 cxd6 2 0 lllb s g3 21
hxg3 fxg3 19 .Jtxg3 lllh s 20 .Jth2 .Jth6 .Jtxa7 (21 hxg3 fxg3 22 .Jtxg3 is met by
21 .U.c3 a6 22 llla3 .Jte3+ 23 lllf2 .Jtf4 22 ...h4 and 23 ... lllh s) 21 ... lllh7
Black had a winning attack in Y.Du (21 ...'ii'xas!?) 22 .Jtb6 (or 22 .Jt g 1 h4 23
Bois-V.Bologan, Bern 1997. h3 I:txas with the idea of ... .Jtxh3 and
d) 16 .U.c2 intends to double rooks . ..'ii'd7, which gives Black an attack)
on the c-file. This move was played in 22 ... 'ii'h4 23 .Jtg1 lllg s 24 llle 1
one of the most famous King's Indian
'prototype' games: 16.. ..i.f8 17 cxd6
cxd6 18 'ii'd2 g4 19 .i::tfc1 g 3 ! (the the­
matic sacrifice) 20 hxg3 fxg3 21 .Jtxg3
lllh s 22 .Jth2 .Jte7 23 lll b 1?! .Jt d7 24
'ii'e 1 .Jigs 2s lll d 2 .Jte3+ 26 @h1 'ii'g s 27
.Jtf1 I:taf8 28 .U.d1 bS 29 a4 a6 30 axbs
axbs 31 Ite7 Itg7 32 lllb3 lllh4 33 I:tc2
.Jth3! (every one of Black's pieces par­
ticipates in the attack)

(seefollowing diagram) 24 ... lllh 3! 2S gxh3 was A. Shirov­

T.Radjabov, Baku (rapid) 2009, and now
34 'ii'e2 lll xg2 3 S .Jtxg2 .Jtxg2+ 36 Golubev suggests 2s ...'ii'xh3 26 .Jtd3
'ii'xg2 'ii'h 4 37 'ii'x g7+ @xg7 38 .U.g2+ lllh4 27 .i::t c2 .i::t g7 with good attacking
@h8 39 llle 1 lllf4 40 Itg3 .Jtf2 41 l:tg4 chances.
'ii' h3 42 llld2 h s 43 I:tgs 0-1, M.Taima- We now return to 1S a4:

Th e M a r def Plata Varia tion : 9 t:De1 t:D d 7 1 0 iL e 3

After 16 as Black has:

bl) Following 16 ...a6! ? 17 cxd6 (17
@h1 l:Ig8 18 cxd6 cxd6 19 b4 il.. d 7 20
.U.a3 il..f8 gave Black counterplay in
Z.Kozul-Z.Lanka, Batumi 1999) 17 ...cxd6
18 t:Da4 g4 19 t:Db6 .U.b8 20 t:Dxc8 I:txc8
White is probably a bit better, although
Black has counterplay. This could also
come about after 13 a4 t:Dg6 14 as a6
1S CiJd3 @h8 ! ? in the notes to Black's
13th move in Line E.
1s ... hs b2) 16 ... .U.g8 17 cxd6 cxd6 18 t:Dbs g4
This is the most obvious choice, but 19 fxg4 t:Dxe4 20 il..xa7 il.. d7 21 il..b6
there are other ideas: 'ii e7 22 t:Da I:taf8 is very messy. A cou­
a) 1s ...I:tf7 16 as il..f8 17 cxd6 il.. xd6 ple of practical examples:
(I would prefer 17 ... cxd6) 18 t:Dcs t:Df8 b21) 23 I:ta3 t:Dgs 24 b4 e4 2s t:De1
19 'ii b3 'ii e 7 20 t:De6?! t:Dxe6 21 dxe6 was V.Korchnoi-Z.Lanka, Linz 1997.
il..xe6 22 'ii xb7 'ii f 8! gave Black coun­ Now Black should play 2S ...t:Des with
terplay in M.Golubev-S.Ovsejevitsch, the idea of .. .f3 according to Lanka.
Alushta 1999. However, 20 t:Dxb7! looks b22) 23 t:De6 iLxe6 24 dxe6 t:Dgs 2s
strong. t:Db4 e4 26 t:Dds (better is 26 il..c4 with
b) 1s ...@h8! ? introduces a different an unclear position in T.Chmielewski­
plan and might even be Black's best. D.Lybin, correspondence 199S)
Because the move ...@h8 is useful and 26 ...'ii xe6 27 t:Da 'fies and Black had
in this position White is not targeting the initiative in P.Kiriakov-Y.Zimmer­
C7, Black plays ...@h8, ... l:Ig8 and ... il..f8, man, Sochi 2004.
instead of ...Itf7, ... il..f8 and ... Itg7. 16 a 5
Instead 16 cxd6 cxd6 17 as g4 1 8
t:Dbs g3 transposes t o note 'a' to
White's next move.
Instructive is 16 h3 Itf7 17 c6 as!.

(seefollowing diagram)

This fine move stops the advance of

White's a-pawn and also slows down
White's t:Db4-c6 manoeuvre:

A ttacking C h e s s : Th e King 's I n dia n , Vol u m e 1

(19 hxg3 fxg3 2 0 .i.xg3 h 4 with the idea

... lllh s is the usual theme) 19 ... lllh7
(19 ...llle 8!? is also worth considering)
20 h3 'ii'h4 21 .i.b6 .i.xh3 22 gxh3 'ii'x h3
23 I:tf2 lllh4! 24 'ii'f1 gxf2+ 25 lllxf2
'ii'g 3+ 26 @h1 was B.Larsen-E. Torre,
Bauang 1973. Now both 26 ... lllg s and
26 ... lllxf3 should win for Black.
b) 17 a6! ? is probably the most test­
ing. Black has:
bl) 17 ...bxa6 18 lllb4 g3 19 hxg3
18 cxb7 .i.xb7 19 b4 .i.c8 20 bxas fxg3 20 .i.xg3 h4 21 lllc6 is annoying
�h6 21 lllb 4?! (better is 21 a6 .i.xa6 22 for Black.
lllb4 .i.c8 with an unclear position ac­ b2) 17 ...b6 18 cxd6 cxd6 19 lllb4 g3
cording to Kasparov) 21 ...g4 22 lllc6 20 hxg 3 fxg3 21 .i.xg3 h4 and here, in­
'ii'f8 23 fxg4 hxg4 24 hxg4 2 5 �f3 stead of 22 .i. h2 lllh s 23 lll c6 'ifgs, the
'ii'h6 26 .i::te1 lllh4 27 .i.xh4 .i.xh4 28 g s immediate 22 lll c 6! is again problem­
'ii'xgs 2 9 �e2 lllg4 30 I:tb1 .i.g3 31 �d3 atic.
'ii'h4 0-1, V.Korchnoi-G.Kasparov, Am­ b3) 17 ... g3 18 hxg3 fxg3 19 .i.xg3 h4
sterdam 1991. 20 .i.h2 lllh s 21 axb7 .i.xb7 22 f4
16...g4 lllhxf4 23 l2Jxf4 lllxf4 24 �xf4 exf4 25
�g4 was W.Harper-K.Waidyaratne,
Connecticut 2007. This is a favourable
version of the main line for White.
11 ... g3 18 hxg3 fxg3 19 .i.xg3 h4 20
.i.h2 lllh s 21 cxb7 .i.xb7

Black has easily achieved his king­

s ide break and White must decide how
to continue on the queenside.
17 c6
a) 17 cxd6 cxd6 18 lllb s g3 19 .i.xa7 Black's kingside play looks ominous,

Th e M a r def Plata Varia tio n : 9 lb e 1 lb d 7 1 0 iL e 3

but White has a way to fight for space Korchnoi's idea looks odd at first,
on that flank by sacrificing the pawn but it is actually quite thematic. White
back immediately goes after the valuable c8-
22 f4! bishop. Eventually Korchnoi gave it up,
Both sides must always keep this but the American GM Dmitry Gurevich
idea in mind. White returns the pawn still frequently plays this line.
in order to fight on the kingside. 13 ...b6
22...lbhxf4 23 lbxf4 ibxf4 24 il..xf4 exf4 Instead 13 ...a6 falls in with White's
2s �g4 'if gs plans: 14 lba7 l:Ixa7 (otherwise White
will just take Black's precious bishop on
c8) 1s il.. xa7 b6 16 b4 il.. b7 17 cs dxcs
18 Itel! and White maintained the ini­
tiative in V.Korchnoi-K.Hulak, Zagreb
lnterzonal 1987.
The pawn sacrifice 13 ...lbf6!? 14
lbxa7 i..d7 is interesting:

26 �e6+ @h8 27 I:tf3 il..c8 28 Ith3 �xe6

29 dxe6 �ae8
Here Black retained sufficient play
in V.Korchnoi-Xie Jun, Amsterdam

C) 13 lbbs
a) 1s lbbs lbg6 16 'ii'c2 hs 17 cs g4
18 c6 bxc6 19 dxc6 il..c8 20 il..c4+ @h8
21 lbd3 g3 was D.Gurevich-M.Golubev,
Biel 1992, when White should probably
test Black's idea with 22 hxg3 fxg3 23
il..x g3 h4, although Black certainly has
b) 1s cs l:Ixa7 16 cxd6 lbc8 17 dxe7
'ii'xe7 18 il.. xa7 lbxa7 19 'ii'b3 'ii'cs+ 20
@h1 lbhs 21 lbd3 and here, instead of
21 ...'ii'e 3 22 lbf2 'ii'x e2 23 'ii'xb7 l:If7 24

A ttacking C h e s s : Th e King 's I n d ian, Vol u m e 1

'ii'xa7 g4 2s l:tfe1 'ii'bs 26 fxg4 with a 18 @h1 bS 19 cxd6 cxd6 20 lllb1 lllf6 21
big advantage in J.Benjamin-J.Nunn, llld2 'iVhs 22 .Jt g1 'ii'g 6 (with the idea of
Hastings 1987/88, Black should play ...g4) 23 Itel lllh S! and White was in big
'ii'd6! (with the idea of ... lllg 3) 22 @g1 trouble in M.Gavilan Diaz­
lllb s with counterplay. C.Matamoros Franco, Malaga 2010.
14 b4
After 14 a4?! a6 1S lllc3 as White
has lost two tempi and Black only one.
Thus White is just a tempo down on
Line E.

1s l:If6!?

Black goes for the attack along the

h-file. Usually I am sceptical about this
idea, but here I think there is consider­
able justification in playing this way.
14 a6
... This is about as good a version as Black
The knight has to get booted at will get of this plan because he has
some point. After 14... lllf6 1S cs g4?! played ... a6 for free. This fits in with the
(better is 1S ... a6 16 cxd6 cxd6 17 lllc3 straightforward plan to play ... .l::th 6 and
hS, heading for the main line) 16 cxd6 ...'ii'e8 because White can no longer
cxd6 17 Itel White has the initiative. attack c7 with lllb s.
1s lllc3 The other logical approach is 1S ...hs.
The alternative is 1S llla3 and here Here White's queenside play is not ob­
1S ... as 16 cs axb4 17 cxd6 cxd6! ? with vious, but Dmitry Gurevich has demon­
the idea ... lllcs is interesting, while strated a dangerous positional concept
practice has seen: after 16 cs (White has tried other
a) 1S ...h s 16 cs bS 17 llla c2 lllf6 18 moves such as 16 Itel and 16 a4, while
a4 bxa4 19 .l::txa4 lllg 6 20 bS g4 21 lllb4 one high-level example continued 16
g3 22 hxg3 fxg3 23 .Jtxg3 h4 24 lllc6 @h1 lllf6 17 cs g4 18 cxb6 cxb6 19 Itel
'ii'd7 2S .Jth2 .Jth6 was a complete mess g3 20 .Jtg1 gxh2 21 .Jtf2 h4 22 llla 4 .l:Ib8
in A.Huzman-1. S mirin, Sverdlovsk 1987. 23 bs axbs 24 .Jtxbs lllhs 2s @xh2 lllg 3
b) 1S ...l:If6 ! ? 16 llld 3 l:Ih6 17 cs 'ii'e8 26 .l::t g 1 lllg 6 and Black had counterplay

Th e M a r def Plata Varia t i o n : 9 t:De1 tiJ d 7 1 0 iL e3

in V.Korchnoi-Ye Jiangchuan, Novi Sad Black should play 22 ...t:Dxc6 23 dxc6

Olympiad 1990) 16 ... t:Df6 17 cxb6 cxb6. il.. e 6 24 il..xb6 (24 t:Dxb6 fails to 24... g3)
24...l:Ixb6 2s 'ii'xas CiJd7 when White
has the initiative, but at least the posi­
tion remains complicated.
16 c5
After 16 t:Dd3 l::t h6 17 cs 'ii'e8 18 @h 1
bxcs 19 bxcs t:Df6, the idea of ...'ii' hs
and ... g4 gave Black an attack in J.Kiltti­
V.Maki, Tampere 1998
16... l::t h 6 17 cxd6 cxd6 18 CiJd3 'ii'e 8 19
Instead 19 @h 1 'ii'h s 20 il.. g 1 t:Df6 21
It looks like White will have trouble t:Df2 'ii'e 8 22 t:Dg4 iLxg4 23 fxg4 l::t c8 24
breaching the queen side, but in fact his Itel l::t xc3 2S .U.xc3 t:Dxe4 gave Black
play is very fast The b6-pawn will come good compensation in D.Khismatullin­
under fire and White can also exploit F.Amonatov, Moscow 2009.
the weakened c6-square: 18 t:Da4! l::t b8 19...t:Df6 20 t:Df2 l:Ig6
19 b S ! (the key move) 19 ... as 20 l:Icl g4
21 'ii'd2 (now Black has to watch out for
ideas of capturing on b6 followed by
'ii'x as) 21 d7 (after 21 ... t:Dg6 the blow
22 l::t c6 is even stronger) 22 l::t c6! .

Compared to the positions arising

after lS ...h S, White's bishop has been
diverted from attacking the b6-pawn.
After 21 Itel hS 22 h3 Black played
22 6 in J.Federau-K.Klundt, German
Now instead of 22 xc6 23 dxc6 g3 League 1988, while 22 d7 was seen
24 hxg3 fxg 3 2S il.. x g3 t:Dg6 26 'ii'c2 in X.Sole Fabregat-A.Pablo Marin, Cata­
which was very good for White in lonian League 1998. Both moves look
D.Gurevich-E.Yanayt, Las Vegas 2006, good - Black has a strong attack

A ttacking Chess: The King 's I n d i a n, Vo l u m e 1

D) 13 Itel practical try, but I have my doubts

about its true worth. White can pre­
pare the CS-break OT play it straight­
away as a pawn sacrifice:

This is a very natural and obvious

move, and it may transpose to other
lines if White follows up with a quick
b4 or llld 3. However, there is one im­ a) 14 b4 has scored badly for White
portant independent idea connected overall, but matters are not so clear.
with a pawn sacrifice. 14... .l:th6 1S cS! 1S ... a6 (1S ...'ii'e 8 16 lllbS
13 lllg6
... 'ii'h S 17 h4 lllg6 18 g4 fxg3 19 .Jt xg3
This is the main line. Black has also lllf4 20 lllx c7 may look risky for White,
recently tried 13 ... a6, which seems in­ but it is hard to find anything convinc­
sufficient to me: 14 llld3 b6 (14....l:tf6 1S ing for Black) 16 cxd6 cxd6 17 llld3 lllf6
cs .l:th6 transposes to 13 ....l:tf6, below) 18 llla4! 'ii'e8 19 lllb6 �S 20 h4 was
1S b4 .l:tf6 16 cs .l:th6 17 @h1 ! ? 'ii'e 8 18 S.Atalik-1. Nikolaidis, Halkida 1997.
cxd6 cxd6 19 bS lllf6 20 bxa6 'ii'h s 21
.Jt g 1 'ii'g6 22 lllf2 lllh s 23 lllh 3 lll g 3+ 24
hxg3 .Jtxh3 2S gxh 3 fxg 3 ? (this is bad,
but both 2S ...�xh3+ 26 @g2 and
2s ...'ii'h s 26 I:tf2 fxg3 27 .i:Ig2 'ii'xh 3+ 28
.Jth2 lll g 6 29 'ii' e1 gxh 2 30 .l:txgs also
look insufficient) 26 f4! exf4 27 .Jtg4
with a clear advantage to White in
J.Berkvens-E.lnarkiev, Kerner 2007.
The main alternative is 13 ....l:tf6.
Black intends to play ....l:th6 and
...'ii'e8 when ... 'ii'h s could be a big This is a good example of White's
threat. This is a popular way of meet­ 'desperate' h4 defence, leaving Black's
ing 13 Itel and may well be a good attack looking too speculative. Now

Th e M a r de/ Plata Va ria t io n : 9 lbe1 lb d 7 1 0 iL e 3

2 0 ... l::t b8 21 lbxc8 .U.xc8 22 .U.xc8+ lbxc8 il.. xds 'ii'h s 32 .U.a1 l::tf8 33 l::t a7 g4 34
23 'ii'c2 lbe7 24 'ii'c7 lbg6 2S 'ii'xd6 fxg4 'ii'xg4 3S il..xd4 exd4 36 l::t f3 with a
lbxh4 26 'ii'e6+ @h8 27 il..xh4 'ii'xh4 28 clear advantage for White in
'ii' h3 'ii'xh3 29 gxh3 l::t xh3 30 Itel hS A.Lenderman-R.Barcenilla, Mesa 2009.
was suggested by Atalik, but this looks c) With 14 cs White continues his
in sufficient for Black. plan and sacrifices a pawn to open the
b) 14 lbd3 also has scored badly c-file. Then 14 ... lbxcs 1S b4 lba6
overall, but looks promising for White reach es a key theoretical position for
as well after 14 ...l::t h 6 1S cs when again the 1 3 - Af6 sub-variation .
Black has to decide whether or not to
prevent lbbS:
bl) 1S ...'ii'e 8 still seems too optimis­
tic after 16 lbbs 'ii'h s 17 h4 lbg6 18 g4
fxg 3 19 il.. x g3, which looks ugly, but
Black's queen side is melting away and
19 ...lbf4 20 lbxc7 is almost identical to
variation 'a' above. Even though
White's king position looks bad, it is
not clear how Black should continue as
his pieces are awkwardly placed.
b2) 1S ... a6 and now White has two White has:
promising moves: cl) 16 lbbs l::t h 6 17 lbxa7 il..d7 18
b21) 16 cxd6 cxd6 17 a4 'ii'e8 18 il.. xa6 (White goes after the c-pawn;
ii.el! lbf6 19 ibf2 l::t g6 20 as il..d7 21 instead 18 il.. bs lbxb4 19 'ii'a4 il.. xbS 20
lbb1 (with the idea lba3-c4-b6) 21 ...hS 'ii'xbS b6 21 a3 lba2 ?! 22 lbc6 lbxc6 23
22 h3 g4 23 fxg4 hxg4 24 hxg4 il..h 6 2S .U.xc6 .U.xa3 24 'ii'b2 l::t a7 2S lbd3 g4 26
l::t c7 l::t c8 26 .i::txc8 lbxc8 27 lbd2 f3 28 fxg4 'ii'g S was l.Rajlich-M.Al Sayed, Bu­
gxf3 lbhS 29 lbh3 gave White a win­ dapest 2001, and here 27 h3 would
ning position in M.Dambacher­ give White a clear advantage) 18 ...bxa6
R.Ponomariov, European Club Cup, Kal­ 19 'ii' c2 g4 20 'ii'xc7 'ii'e8 and now:
lithea 2008, which was a game with a c11) 21 g3? (this turns out badly)
large rating disparity. 21... @h8 22 .U.c2 lbg8 23 fxg4 il..xg4 24
b22) 16 c6 bxc6 17 dxc6 lbf8 18 lbb4 il..b6 'ii'g 6 2S lbc6 fxg 3 26 lbd8 l::t xd8 27
'ii'e8 19 @h1 lbe6 20 lbcdS @h8 'ii'xd8 gxh2+ and White had seen
(20 ... lbxdS 21 lbxdS is similar) 21 lbxe7 enough in V.Korchnoi-1.Cheparinov,
'ii'xe7 22 lbds 'ii'f7 23 il..g 1 lbd4 24 il..c4 Amsterdam 2008. After 28 l::t xh2 il.. f3+
il.. e6 2S 'ii'd3 as 26 b3 l::t b8 27 'ii'd2 l::t a8 29 .U.g2 il.. xg2 30 lbxg2 'ii'xe4 Black has
28 a3 l::t h4 29 b4 axb4 30 axb4 il..xdS 31 the initiative and is up material.

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

c12) 21 fxg4 il.xg4 and Black is not the advantage i n R.Vera Gonzalez-
threatened by either 22 tllc6 tllxc6 23 1.Nataf, Havana 2002.
dxc6 'it'h5 24 h4 il.f6 (24..Jk8!? was c22) 18 was Natafs improve­
suggested by Mikhalevski) 25 'it'b7 nf8 ment against the same opponent, but
26 nC3 il.xh4 27 il.xh4 'it'xh4 28 tllf3 as Golubev points out, this may still be
hf3 29 ncxf3 'it'h2+ 30 �2 ng6 3 1 insufficient: 19 'it'b3 �h8 20 tll b 2?!
ng1 ng3 32 �f1 �h8 33 'it'e7 nfg8 34 (better is 20, as there is no ...tllh 5
'it'f6+ n8g7 35 'it'f8+ ng8 36 'iVf6+ n8g7 to worry about and 20 ... 'it'h5 is met by
37 'iVf8+ V2-V2, V.lotov-V.Saravanan, 21 tllf2) 20 ... 'it'h5 21 il.g1 g4 22 fxg4
Kalamaria 2008, or 22 il.b6 'i!Vh5 (if il.xg4 23 il.xg4 'it'xg4 24 tll b 5 il.f6 2 5
Black wants to play for a win then there tll x e7? tllxe7 26 nxa 'it'e2 27 nd1 'it'xe4
is Mikhalevski's suggestion of 22 ...tllc 8! 28 tllc4 tllf5 29 il.f2 il.h4 30 il.xh4 was
23 tllc6 tllxb6 24 'it'xb6 'i!Vh 5 25 tllf3 R.Vera Gonzalez-1.Nataf, Montreal
iLxf3 26 nxf3 'it'xh2+ 27 �fl il.f6 and 2003, and now the easiest win is
Black retains chances to attack) 23 tllf3 30...tllxh4 31 'it'h3 ng8 32 nc8 tllf5, as
il.xf3 24 nxf3 'it'xh2+ 2 5 �fl ng6 26 pointed out by Gallagher. Instead 19
nc2 ng3 ! 27 nxf3+ 28 gxf3 'ilr'h3+ b5?! 'it'h5 20 il.g1 tll c 5 is good for Black,
29 �e2 'it'g2+ (or 29 ...tllf5 immediately) but Golubev points out that 19!
30 il.f2 tllf5 3 1 exf5 e4, which leads to threatens b5 when 19...'ilr'h 5 just trans­
perpetual check, S.Arounopoulos­ poses to variation 'c21'.
M.Ghinda, correspondence 2007. Now we finally return to 13 ...tllg 6.
c2) The assessment of 16 tlld3! ? has
gone back and forth a bit. 16...nh6 17
a4 'it'e8 18 �hl and now:

Here 14 b4 and 14 tlld3 will trans­

pose to Lines A and B respectively, but
White has another idea.
c21) 18... 'it'h5 19 il.g1 il.d7 20 tll a2 14 cs!?
nc8 21 tllf 2! 'it'e8 22 tllg4 ng6 23 b5 Kozul's pawn sacrifice rekindled in­
tll c 5 24 il.xc5 dxc5 25 nxc5 gave White terest in 13 nc1. 1t is clear that 14... dxc5

Th e M a r def Plata Va riatio n : 9 tD e1 tb d 7 1 0 ii. e3

lS b 4 is no fun after either 1S...cxb4 16 'it'b6+ 37 �f2 'it'e3 was level in

tDbs or 1S ...b6 16 t:Dd3, so Black has to V.Korchnoi-0.Cvitan, Pula 1997.
take the pawn with his knight. 16 ii..d 7 17 tDxa7

14 tDxcs 15 b4 tDa6
••• Instead 17 a4 l:tf7 18 t:Dd3 would
Not 1S ...tDd7? 16 tDbs. This is the point lead us back to 16 t:Dd3, above, but
of White's sacrifice - the black knight Black could also try 17 ... h s or 17 ... 'it'b8.
has been sidelined for the foreseeable 11 h s

future. Now White has a choice: he can This looks like the most accurate
immediately regain the pawn, or he move. Instead 17 ...'it'b8! ? 18 tDbs (or 18
can play slowly and count on gradual l:tf7 19 ii..xd7 �xd7 20 'it'a4 �f7)
pressure against Black's queenside. 18 ....l:tf7 19 a4 h S 20 tbd3 ii..h6 21 �c4
16 tDbs �g7 22 �hl 'it'd8 was the actual move
order of Atalik-Gufeld below, which
leads us back to the main line.
Black has also tried 17....l:tf7 and
a) 18 ii..xa6 ! ? looks a bit better for
White: 18 ... bxa6 19 'it'C2 (19 tDc6 ! ?)
19 ... ii..e8 20 'it'e2 h S 21 'it'xa6 g4 22 tDc6
ii..xc6 23 'it'xc6 �xa2 24 �c2 �a3 2 S bS
ii..f8 26 b6 g3 27 ii.cs (27 hxg3 fxg3 28
ii..x g3 cxb6 29 ii..f2 tDf4 30 ii..xb6 'ifgs
31 'it'c8 was suggested in New in Chess
This is the most forcing continua­ and looks strong) 27 ...�b3 28 'it'a4 �bl
tion. 29 'it'a2 �bS 30 'it'a6 �bl and Yi-Y2 was
White can also play 16 tDd3. Then D.Gurevich-J.Becerra Rivero, Las Vegas
16 ...l:tf7 17 t:Dbs 1'..d 7 (17 ... hs 18 a4 ii..h6 2007.
19 t:Dxa7 1'..d 7 would be the same) 18 b) 18 a4 h s 19 tDbs ii..f 8 (19 ...ii..h6 20
a4 hS (18 ...'it'e8 has also been played, �c4 is the main line after 17...h s) 20
but it is probably too artificial) 19 t:Dxa7 t:Dd3 (also good is 20 tDc3 �g7 21 tDd3
ii..h6 20 l:tc3 �g7 21 t:Dbs tDf8 ! ? 22 h3 �h8 22 �hl 'it'f6 23 as 1'..e 7 24 bS tDb8
tDh7 23 ii.el tDf6 2 4 tDf2 (otherwise 2S �al with some advantage in
... g4 comes) 24...tDxb4 2S �xc7 tDa6! 26 A.Yermolinsky-J.Fedorowicz, North Bay
�xb7 tDcs 27 �c7 g4! 28 hxg4 hxg4 29 1998) 20 ... �g7 21 �hl lDM (insuffi­
fxg4 tDfxe4 30 tDxe4 tDxe4 31 'it'b3 tDg3 cient is 21 ... tDh8 22 g3 fxg3 23 ii..xg3
32 ii..x g3 fxg3 33 'it'xg3 ii..xbs 34 �xg7+ tDf7 24 �gl �c8 2S ii..f2 'iVf6 26 ii..e 3
�xg7 3S gs (3S ii..xbs? 'it'b6+ 36 �f2 �g6 27 'it'd2 1'..e 7 28 tDc3 �h7 29 bS
ii..e3 wins for Black) 3 S ... ii..x gs 36 ii..x bs tDcs 30 t:Dxcs dxcs which occurred in

A ttacking C h e s s : The King 's I n dia n, Vol u m e 1

A.Yermolinsky-H.Yoshiharu, Chicago g3 22 'i:Vxb7 gxf2+ 23 .llxf2 f3 24 tllxf3

2000, as 31 as looks much better for il.h6 2 5 l:tc3 'it'b8 26 'it'xb8 .llfxb8 27
White) 22 g3 fxg3 23 il.xg3 tllg 6 24 l:tgl tllc6 l:tbl+ 28 l:tfl l:txfl+ 29 �xfl l:txa4
'it'f6 25 'it'd2 l:tc8 26 il.f2 il.e7 27 il.e3 30 il.c4 il.e3 in V.Erdos-P.Acs, Hungar­
and here, instead of the 27 ...tllh 4? 28 ian League 2005.
il.xg S ! of A.Yermolinsky-J.Becerra 19 .llc4 l::t f7 20 tllbs l:tg7 21 �h 1 tllf8
Rivero, Chicago 2008, Black should play
21 ...tllf4 28 tllxf4 gxf4 29 il.f2 l:txg1+ 30
l:txgl+ �h8 31 tlla7 l:tg8 32 .llx g8+
�xg8 33 il.xa6 bxa6 34 tllc6 when
White is better, but Black is still in the
18 a4
Capturing on a6 is still an option: 18
il.xa6 bxa6 19 tllc6 'it'f6 20 tlla s 'it'd8 21
tll c 6 'iVf6 22 tlla7 'it'd8 23 tll c 6 was
drawn in D.Gurevich-J.Becerra Rivero,
US Championship, Stillwater 2007. Black intends to play ... g4 with an
Here too White might have tried 19 attack However, because one of Black's
'i:Vc2!?. knights is so far away, White can afford
to fight on the kingside directly.
22 g3! fxg3 23 il.xg3 tllg6 24 tlld 3 tllf4
This is better than 24...�h7 2 5 l:tgl
.llf7 26 il.f2 with an edge in
A.Yermolinsky-S.Kindermann, Gronin­
gen 1997.
Instead 18 ....llf7 brings us back to
17 ... .llf7, but an interesting idea of
Golubev's is 18 ...tllx b4!? with the idea
19 'it'b3 g4!. This has been seen in prac­
tice and the complications led to a
draw following 20 fxg4 hxg4 21 'i:Vxb4 2 5 il.xf4

Th e M a r def Plata Va riati o n : 9 llJe1 lLl d 7 1 0 ii.e3

This is the most common, but prac­ 30 ii.. g2 �h6 31 'it'f1 'it'e7 32 ii..h 3 l:tg8
tice has also seen 2S ll:lxf4 gxf4 26 ii..f2 33 ii..x d7 'it'xd7 34 lLlC 3. Both feel White
�h8 (Black could try 26 ... M with the is a little better, but the position is still
idea of ... h3) 27 l:tg1 28 1'.. f l nc8 29 very complicated.
'it'e1 'it'e7 30 ii.. g 2 ii..f6 31 'if cl 1'.. h 4 32 29... �h6! 30 ll:lc3 'it'e7 31 ll:le2 b6 32
ii..xh4 'it'xh4 33 ll:lxd6! cxd6? (a better nc6
chance was 33 ... ncg8 34 ll:lfs ii..xfs 3 S We've followed the game S.Atalik­
exfs 'it'f2 3 6 nc2 'it'b6) 3 4 nxc8+ ii..xc8 G.Timoshen ko, Timisu de Sus 1998.
3S 'it'xc8+ �h7 36 ii.fl and White was Here Atalik suggests 32 ... ng8! with an
wining in l.Rajlich-A.Brustman, Ostrava attack
2s ...gxf4 26 ng1!? E) 13 a4
This looks better than 26 ...nxgl+ 27
"i' xgl+ �h7 28 ii.fl 'it'e7 (Black could
try 28 ... ii..g s or 28 ... nc8) 29 ll:lxc7! l::tc8
30 'it'b6 ii..xa4 31 ll:lb2 (this leads no­
where, so 31 b S ! is better) 31 ...ii..b3 32
ll:lxa6 ii..xc4 33 ll:lxc4 bxa6 34 'it'xa6 ng8
3S 'it'b6? (3S 'it'xd6 'it'xd6 36 ll:lxd6 ii..f8
is about even) 3 S ... 'it'g7? (3S ... 'it'h4! with
the idea of ...'if el wins) 36 'it'f2 (36
ll:lxd6 ! ?) 36 ...'it'b7 was drawn in
S.Atalik-E.Gufeld, Honolulu 1997.
27 ii.fl 1'.. h 4 28 nxg7+ �xg7 This thoroughly modern continua­
tion is the main line and has been
played more than all the other 13th
moves combined. The move a4 is al­
ways useful and White may be able to
play the cs-break without any prepara­
tion, either through tactical means or
after Black moves the d7-knight to
support ... g4. The a-pawn may advance
to as, which often helps White by con­
trolling the b6-square if he manages to
play ll:lbs followed by capturing on a7,
29 'it'c2 especially because the f2-bishop will
Instead Atalik suggests 29 'if cl nc8 have the b6-square to go to in case of
30 nc2, while here Timoshenko gives an eventual ... g3 by Black. The possibil-

A ttacking C h e s s : Th e K i n g 's I n dian, Volu m e 1

ity of a quick tllbs remains, and this Here Black ignores White's play for
sortie could give White an improved the moment and continues his own
version of Line C. campaign. It is very risky, but it is excit­
13 as
... ing. After 14 as (White can also lunge
Eventually this positional move has forth with 14 tllb s ! ?) Black has:
come around to establishing itself as a) 14... tllf6 is too compliant, as it al­
the main line. Black prevents White's lows 15 cs immediately. Following
further advance of the a-pawn and 1s ... hs 16 cxd6 cxd6 17 tllbs g4 both 18
fights for control of the dark squares on tllxa7 and 18 ii..x a7 have scored heavily
the queenside. White will still break for White.
through of course, but Black hopes that b) 14...a6!? is not so bad and could
the time White spends breaking down use further tests: 15 tlld3 tllf6 16 cs
his queenside pawn structure can be �h8 17 cxd6 cxd6 18 tlla4 g4 19 tllb6
used to create counterplay. �b8 20 tllxc8 �xc8 is probably better
It is too early to play 13 ... h s because, for White, although the position is not
as we shall see, sometimes Black can do so clear. We also considered this posi­
without this move. There is another tion via a different move order in the
plan with 13 ...�f6 14 as a6, with the notes to Black's 15th move in Line B.
familiar idea of ... �h6 and ...'it'e8-hS. c) 14...h S is usually a signal for
The play is similar to that discussed in White to play tllb s. This is because if
the notes to Black's 13th move in Line White plays tllb s too early (as in Piket­
D, but here Black has even scored Kasparov), Black may be able to get ... g4
worse, so we will not spend time look­ in without preparing it with ...h s be­
ing at it. cause the e4-pawn is not defended.
Black's most principled alternative This is usually an achievement for
is 13 ... tll g 6, which is the move perhaps Black, not only because he saves a
most in the King's Indian spirit. tempo on ... h s, but also because the
move ...tllh s remains a possibility.
However, with Black spending time on
...h s, White should always consider this
lunge. Indeed, here 15 tllb S! tllf6 (or
1s ...a6 16 tlla7) 16 tllxa7 has scored
heavily in White's favour.
d) 14...�h8 intends ...�g8 and ... ii..f8.
Markos considers this to be Black's best
try, before recommending for White 15
tlld3 �g8 16 cs tllf6 17 cxd6 cxd6 18
tll b s g4 19 fxg4! tllxe4 20 'it'c2. After

Th e M a r def Plata Va riatio n : 9 tDe1 tbd 7 1 0 ii. e 3

Golubev's 20...t:Dxf2, 21 t:Dxf2 controls for White) 18 a6 a critical position is

e4 and gives White an edge according. reached.
Instead 20.. Ji.d7? runs into 21 t:Dxes
and 20 ...'it'e7 21 tDc7 tDxf2 22 t:Dxf2 nb8
23 tDe4 1'..d 7 24 'it'b3 ! (with the idea
tDe6) 24 ...ii..xg4 2S ii..xg4 'it'xc7 26 ii.. e 6
is also good for White according to
Markos. The best try may be giving up
the exchange with 20... tDf6 !? 21 tDc7
ii.xg4 22 t:Dxa8 ii..x e2 23 'it'xe2 e4 24
lDb4 'it'xa8 2S 1'..d4, but again Markos
thinks White is somewhat better and I
tend to agree.
e) The main line has always been Black has:
14...nt1. ell) 18...b6?! 19 d6 nd7 (or 19 ...nf8
20 ii..d s nb8 21 t:Dbs cxd6 22 t:Dxa7) 20
'it'ds c6 (forced, because 20 ... nb8 21
dxe7 'it'xc7 22 'it'g8 is mate!) 21 'it'xc6
nb8 22 lDbS ii..f8 was D.Vigorito­
E.Yanayt, Philadelphia 2003. Now 23
'it'c7! ii..xd6 24 'it'xd8+ nxd8 2S t:Dxa7
gives White a big advantage.
e12) 18...nf6?! 19 axb7 ii..xb7 20
tDd3 ii..f8 21 nt2 (also possible is 21
nas ii..c 8 22 t:Dxcs, as in D.Gurevich­
A.Sherzer, Durango 1992) 21 ... ii..d6 22
Now White has: lDa4 'it'e7 23 nc2 ng8 24 tDf2 hS 2S
el) Just like in Line D, Kozul has ii..e 2 ii..c 8 occurred in D.Vigorito­
gone lS cs here. I have actually tried J.Fedorowicz, US Championship, San
this several times with White, but now Diego 2006, and here 26 nacl! looks
I suspect lS b4 is just stronger. Follow­ good for White.
ing lS ...tDxcs 16 ii..x cs dxcs 17 1'..c4 (af­ e13) 18 ... bxa6 19 tDd3 ii..f8 20 tDa4
ter 17 a6 b6 18 1'..c4 ii..f 8! 19 d6 c6! 20 (20 Ji.xa6 Ji.xa6 21 nxa6 C4 22 lDf2 Ji.CS
'it'b3 tDh8!, as in Z.Kozul-A.Kuzmin, 23 'it'e2 g4! gives Black counterplay ac­
Oberwart 199S, Black will pick up the cording to Bologan) 20 ... g4 (20 .. Ab8
d6-pawn and have two good pawns for has played by Popovic, and after 21 nt2
the exchange) 17 ... �h8 (17...ii..f8 18 d6 or 21 b3 I would suggest 21 ... c6! ?) 21
c6 19 tDd3 ii..xd6 20 tDa4 is much better fxg4 'if gs 22 tDf2 ii..d6 23 na3 nb8 24

A ttacking C h e s s : The King 's I n dia n, Vo l u m e 1

nc3 nb4 was M.Llopis de Aysa-K.Volke, �hl! il.d7 2 1 g4 2 2 il.xd7 'it'xd7 23
Biel 1993, which is unclear according to fxg4 tll x g4 24 tllf3 tllxf2+ 2S nxf2 il.e7,
Bologan. as in P.H.Nielsen-Y.Kruppa, Minsk 1996,
e2) The calm move lS b4 may be although White looks somewhat better
best. here as well, and Markos's suggestion
of 22 il.c6 may be even better) 18 �hl
and White looks faster, but Bologan
has some ideas.

Here Black has:

e21) 1s 16 cs dxcs! was sug­
gested by Markos, leading to 17 d6
il.xd6 18 il.c4 cxb4 and now we have Black has:
the following: e221) 18...nh1 19 tll b s g4 20 tllxa7
e211) 19 'llt'b 3 tllh 8! 20 tllb s il.cs! 21 g3 21 tllxc8 bxc6 22 il.g1 gxh2 23 il.f2
il.xcs tll x cs 22 nd1 'iVf8 23 il.xf7+ tllxf7 'it'xc8 24 dxc6 h4 2 S il.c4+ �h8 26 'llt'b 3
24 'it'xb4 tlle6 2S 'it'xf8+ �xf8 26 l:tf2 was good for White in P.Kiriakov-M.Al
gives Black compensation according to Sayed, Dubai 2002.
Markos. e222) Bologan's idea is to play on
e212) 19 tllb s tllcs (the 19 ... tlldf8 20 the g-file after 18...ng1!?. His main line
'it'b3 'it'e7 of P.H.Nielsen-M.Solleveld, goes 19 cxb7 il.xb7 20 a6 il.c8 21 tll bs
German League 2002, can be met with g4 22 tll x a7 g3 23 il.g1 l:txa7 24 il.xa7
21 tllxd6) 20 tllx d6! (20 il.xcs il.xcs+ is cs 2s il.xcs dxcs 26 d6 tllh 7 27 il.c4+
check!) 20...'it'xd6 21 'it'xd6 cxd6 22 �h8 28 na2 'i(h4 29 h3 il.xh3 30 gxh3
il.xcs dxcs 23 tll d 3 looks a little better 'it'xh3+ 31 �gl tllh4 32 'it'e2 cxb4 33 a7
for White. nxa7 34 nxa7 il.xd6 3S na8+ �g7 36
e22) 1s ...tllf6 is queried by Markos, nas 'it'c8 when Black is two rooks for a
but this is the only move Bologan con­ piece down, but has tremendous coun­
siders: 16 cs il.f8 17 c6! (17 cxd6 is also terplay.
dangerous) 17 ... h s (Markos suggests Now we return to the relatively­
11 ... ng1 18 cxb7 il.xb7 19 a6 il.c8 20 safer 13 ... as:

Th e M a r def Pla ta Varia tio n : 9 tDe1 tb d 7 1 0 ii.e3

and cs-breaks.
15 ...tDf6 16 tDf2
The immediate 16 b4 axb4 17 t:Db s
transposes to Line E21, below.
16... h 5 17 h3

14 lDd3 b6
Black must not allow cs. Now White
can regroup his pieces for both attack
and defence, or he can strike at once on
the queenside.
11 ...�hS!
El: 15 .i.e1 This is a very important move. The
E2: 15 b4 natural 17... tDg6?! 18 t:Dbs leaves Black
struggling to achieve ... g4, whereas
El) 15 ii.el White is ready to break open the
queenside with b4. Thus the text in­
tends to regroup with ...tDeg8, ...tDh6,
....llf7, ... ii..f8 and ....ll g7 when ... g4 will
come with great force.
18 t:Db5
Both sides can employ a few differ­
ent move orders here. One example is
18 b4 tDeg8 19 bxas bxas 20 cs tDh6 21
t:Dbs llf7 22 cxd6 cxd6 23 .ll c l ii..f8,
transposing to the main line.
18...tDeg8 19 b4 .llf7!
This is a flexible, yet somewhat slow It is important for Black to avoid
continuation. White wants to bring the capturing on b4, because then White's
d3-knight to f2 in order to hold up dark-squared bishop would spring to
Black's ...g4-break. He also prepares to life. After 19 ... axb4?! 20 ii..xb4 the aS­
recapture with his bishop after playing break is coming and Black's play is too
b4 which would then support the as- slow. Instead 19 ...tDh6 is possible,

A ttacking Chess: Th e Kin g 's I n dian, Vo l u m e 1

though, and will lead to the main line 2 9 fxg4 tllf6) 2 s ...ng1 2 6 nc6 il.d7 27
after 20 bxas (or 20 cs bxcs 21 bxcs) il.d2 il.xc6?! 28 dxc6 'it'b6 29 nc1 g4 30
20...bxas 21 cs nf7 22 cxd6 cxd6. fxg4 il.e7 31 'it'e6 was good for White
20 bxas in P.H.Nielsen-M.Golubev, Internet
Again 20 cs bxcs 21 bxcs makes no (blitz) 2004. Instead Golubev suggests
difference. 27 ... g4, while I think Black could also
20...bxas 21 cs il.f8 22 cxd6 investigate 27 ...tllf6!?.
White should not become too fancy: The text reaches a critical position.
a) 22 tlla 3?! tllh 6 23 tllc4 ng1 24
na3 �g8 2S tllx as g4 26 fxg4 hxg4 27
hxg4 tllh xg4 28 tllx g4 tllxg4 29 tllc4
'it'g s 30 il.xg4 il.xg4 31 'it'c2 il.e7 32 nf2
'ikh S 33 g3 il.dl! 34 'it'b2 nf8 and Black
had a strong attack in S.Ghane Gardeh­
V.Spasov, Izmir 2003.
b) 22 c6?! tllh 6 23 na3 ng1 24 il.d2
g4 2S fxg4 hxg4 26 hxg4 tllh xg4 2 7
tllxg4 tllxg4 28 'it' el (this prevents
... 'it'h4 and also eyes the a7-pawn)
28 ... �g8! (after breaking with ... g4, this 24...g4!
is a typical move to avoid any na3-h3+ This is better than 24...ng7?! 2s 'it'c2
business) 29 nff3 (29 il.xg4 il.xg4 30 g4 which runs into 26 il.xas!. After
il.xas 'it'g s 31 nf2 nh1 32 tllx c7 nxas 26 ... 'it'e8 (even worse is 26 ...'it'xas 27
33 'it'xas 'it'h4 34 nf1 'it'h2+ 3S �f2 f3 nxc8 gxh 3 28 tllxh3, J.Fang-B.Dean
gives Black a winning attack according Kawamura, Parsippany 2008), White
to Mikhalevski) 29 ... il.e7 30 nh3 tllf6 (or should play 27 fxg4! hxg4 28 tllc 7 'it'g6
30 ... tlle 3 31 nhxe3 il.h4 32 g3 fxe3 33 (or 28 ... nxa 29 il.xc7 gxh 3 30 tllx h3
nxe3 'it'g s 34 nf3 'it'g6 with a clear ad­ tllh g4 31 il.b6) 29 tllx a8 gxh 3 30 il.f3!
vantage - Mikhalevski) 3 1 iLfl ! ? il.xh3 when he survives the attack and keeps
32 nxh3 'it'c8 33 'it'h4?! �f7 was much a decisive material advantage.
better for Black in S.Krivoshey­ Instead 27 h4? gxf3 28 il.xf3 il.g4 29
G.Schebler, German League 2004. il.xg4 tllfxg4! 30 il.b6 'it'e7 31 tllxg4
22 ... cxd6 23 nc1 tllh 6 24 nc4 tllx g4 32 il.f2 f3 33 g3? (this loses
After 24 'it'c2 tlle 8 2 S 'it'b3 (instead quickly, but Black is still much better
2S 'it'd3 ng1 26 nc2 �g8 27 'iVdl looked after 3 3 'JJ..C7 fxg2 34 ne1 'iVf6 3 S nxg7
very strange in A.Bachmann-A.Frolov, il.xg7 36 �Xg2 il.h6 OT 33 nc3 fxg2 34
Berlin 1994, and here Black should na1 ng8 3S il.g3 il.h6 36 naa3 il.f4)
probably continue 27 ... g4 28 hxg4 hxg4 33 ... tlle 3 ! 34 il.xe3 nxg3+ 3 S �2 ng2+

The M a r def Plata Va ria tio n : 9 tDe1 tb d 7 1 0 ii.e3

36 �Xf3 nxc2 37 nxc2 'ifxh4 38 ngl Aguirre, Sort 2006. Black should now
nxa4 39 tDc3 nc4 40 �e2 ii..h 6! 41 ii..b6 play 30 ... t:Dxfl 31 �xfl ng1 with good
nxe4+ and 0-1 was the well-known chances, such as after 32 t:Dxa8 ii.. a 6.
game P.H.Nielsen-V.Kotronias, Hastings
2003/04. 82) 15 b4
25 fxg4
After 2S 'it'c2 1'..d 7 26 tDC7 Black has
the shot 26 ...ii..x a4!.
25 ...hxg4 26 hxg4 ng1

This is a more direct approach from

1s ...axb4
Black should certainly take now, be­
21 gs!? cause otherwise White will either cap­
Worse is 2 7 'it'c2 t:Dfxg4 28 t:Dxg4 ture on as himself or else even break
tDxg4 29 ii..x g4 ii.xg4 when Black has with cs immediately, when the f2-
good attacking chances. bishop remains very active. Now White
21.. Jbgs 28 'it'c2 has a further choice:
After 28 tDc7 na1 29 'it'c2 (or 29 tDe6
ii.xe6 30 dxe6 nag7 31 ii..f3 tDhg4 with E21: 16 "llbS
counterplay) 29 ...1'..d 7 30 tDe6 ii.xe6 31 E22: 16 "bxb4
dxe6 ng8 both sides have chances ac­
cording to Avrukh. E21) 16 t:Dbs
28...tDhs White is hoping to obtain a favour­
All of 28...t:Dfg4, 28...tDhg4 and able version of Line El by playing ii.el
28...ng8 deserve attention as well. and ii..xb4, but the early knight hop
29 ii..f3? gives Black an opportunity to acceler­
29 ii..xhs nxhs 30 na 1'..d 7 is un­ ate his attack and this method of play
clear. for White has a bit of a suicidal reputa-
29... tDg3 30 tDc7 tion. In this variation both sides must
This was S.Krivoshey-N .Ortiz try to break through quickly.

A ttacking C h e s s : Th e King 's In dian, Vo l u m e 1

18 fxg4 smacks of desperation, but

really it is just unclear: 18 ...tllxe4 19
il.xb4 il.d7 (or 19 ...tll g6 20 as tllcs 21
tllx cs bxcs 22 il.c3 e4 23 il.xg 7 �xg7 24
'it'c2 'it'e7 2S 'it'c3+ tll e s with the initia­
tive, A.Nikitin-S.Soloviov, Alushta 2002)
20 'it'c2 tll gs 21 h4 tlle4 22 il.e1 tllg 6 23
h s tll g 3 24 hxg6 'it'h4 2s tllf2 il.xbs 26
gxh7+ (26 cxbs e4 with an attack)
26 ... �h8 (26 ...'it'xh7 wins immediately
after 27 �xh7+ �xh7 28 il.d3+ e4 29
16...tllf6! 17 il.e1 tllxe4 tllxfl, but the text is certainly
Instead 17 tllxb4 looks misguided, good enough) 27 cxbs e4 28 nd1 f3 29
but matters are not so clear after gxf3 tllx e2+ 30 'it'xe2 exf3 0-1,
17 ... g4 18 il.h4 g3 19 hxg3 (after the 19 J.Chabanon-J.Degraeve, French League
h3 of P.H.Nielsen-H.Harestad, Gausdal 1999.
1996, Black could consider 19 ...'it'd7! ?, 18...g3
with the idea of ...tll g 6, or Gallagher's Another idea is 18 ...tll g 6 as sug­
19 ... il.d7) 19 ... tllg6 20 tllc6 �d7, which gested by Golubev. Indeed, the further
is murky, E.Kobylkin-S.Soloviov, Alushta 19 as bxas 20 nxas l:tb8 21 l:ta7 g3 22
2001. 'it'c2 nf7 (better than 22 ...tllh s 23 nxa
gxh2+ 24 �xh2 'it'h4+ 2S �gl tll g 3 26
il.xd6) 23 nb1 gxh2+ (23 ... tllxds 24
exds e4 2S fxe4 nxbs 26 na8!) 24 �xh2
tllxdS !? 2S exds e4 certainly deserves
19 h3
Instead 19 as should be met with
19 ...bxas! (19 ...gxh2+ 20 �xh2 bxas 21
.l:txas nxas 22 il.xas tlle8 23 cs is better
for White according to Markos) 20
nxas (20 il.xas nxas 21 nxas c6)
17 ...g4! 20 ...l:txas 21 il.xas tlle 8 (with the idea
Because the e4-pawn is loose, Black of ...tll g 6) 22 cs tll g6 23 'it'c2 'i!fh4 24 h 3
does not need any further preparation il.xh3 2S gxh3 'it'xh3 2 6 il.d1 tllh4 2 7
for his key kingside break and can ad­ ne1 dxcs, which i s a little better for
vance at once. Black according to lkonnikov in New in
18 il.xb4 Chess.

Th e M a r def Plata Varia t i o n : 9 tDe 1 tb d 7 1 0 ii.e3

up on the queenside) 20 ... tDg6 (20 ... cs

was suggested by Markos) 21 ntc1 lDh4
22 tDel nf6 23 ii.fl should be some­
what better for White, even though
Black eventually made a sacrificial
breakthrough in N.Sulava-V.Neved­
nichy, Bad Worishofen 2000.
20 gxh3 'it'd7 21 'it'c2
This is the best defence. White will
try to defend along the second rank.
Instead 21 �g2? gets crushed: 21 ... tDg6
The text reaches a controversial po­ 22 nh1 tDh4+ 23 �g1
sition. Black's play is very far advanced,
but with the kingside closed he will
have to sacrifice to break through.
19 ii..x h3!

The assessment of this move has

gone back and forth from '?!' to '!'.
Black will clearly obtain an attack, but
will White be able to defend? For a
while I believed that this sacrifice was
not quite good enough, but upon fur­
ther investigation I believe it is indeed
the best, both from a practical and an 23 ...t:Dxe4! 24 fxe4 f3 25 ii.. d 2 (or 25
analytical point of view. We will see, ii..xf3 nxf3 26 tDel 'it'f7 27 t:Dxf3 t:Dxf3+
however, that Black's order of moves is 0-1, S.Ghane Gardeh-S.Krivoshey, Dubai
extremely important. 2006) 2s .. .f2+ 26 t:Dxf2 nxf2 27 l:th2
If Black does not sacrifice immedi­ gxh2+! 28 �xf2 'it'xh3 29 'ikh1 nf8+ 30
ately, White may be able to hold the �el 'it'g2 31 ii..f3 t:Dxf3+ and 0-1 in
kingside together: for example, R.Pogorelov-C.Matamoros Franco, Dos
19 ... tDg6 20 ne1 tDe8 21 ii.fl tDh4 22 Hermanas 2003.
'it'e2 nf6 23 nebl ng6 24 tDel ii..f8 25 21 'it'd2 is better, but unimpressive:
as gave White all the play in 21 ... t:Dg6 (or perhaps even 21 ... 'it'xh3) 22
D.Vigorito-S.Muhammad, Chicago ii..dl ii..h6 23 tDel �h8 24 na2 ng8 25
2007. Instead 19 ... tDe8 was praised by 'it'd3? (25 'it'g2 tDh4 26 'it'h1 is unclear ­
Markos, but 20 'it'd2 (20 ne1 with the Gallagher) 2s ... 'it'xh3 26 .i:tg2 tDh4 27
idea of j,.f1, ne2, tDel and nea2 can be 'it'e2 ngs was winning for Black in
met with 20 ...cs!?, trying to hold White S.Krivoshey-Xie Jun, Linares 1997.

A ttacking C h e s s : Th e Kin g 's I n dian, Vo l u m e 1

21...tllg 6! 'it'xf3+ 38 �c2 was winning for White

This move is very important. The in C.Persson-G.Calzolari, correspon­
obvious 21 ...'it'xh3 is not as good, but it dence 1998.
is worth looking at it to see some typi­ Now back to 21...tllg 6!:
cal ideas for both sides. After 22 il.d1
tllg 6 23 'it'g2 'it'h6 practice has seen:
a) 24 'iVhl tllh4 25 tlle l?! (25 ne1!)
25...tllxe4! 26 tllg 2 was S.Koutsin­
A.Frolov, Kiev 1995, and now
26 ...tllxf3+! 27 il.xf3 'it'xhl+ 28 �xhl
tllf2+ 29 nxf2 (29 �gl e4) 29 ... gxf2 30
nt1 (30 il.e4 nxa4!) 30 ...nxa4 31 il.a3
nta8! (this seems even better than Gal­
lagher's 31 ..Axc4 32 nxf2) 32 il.b2
nxc4 33 nxf2 na2 should win for Black.
b) 24 tll x c7! tllh4 25 'it'd2! (but los­ 22 ntb1
ing is 25 'it'e2? tllxe4! 26 'it'xe4 tllg 2! as This move is often recommended.
given by Gallagher) is a key idea - Instead:
White lines up against Black's queen to a) 22 il.d1? loses to 22 ...tllM 23 tll e1
avoid the ... tllxe4 and .. .f3 bomb. 'it'xh3 24 tllxe7 (24 na2 tllh 5) 24...tllh 5!
25 tllxa8 g2 2 6 tllx g2 tllg 3 and 0-1 in
M.Dziuba-J.Czakon, Koszalin 2005.
b) 22 nfdl is similar to OUT main
line: 22 ...'it'xh3 ! (22 ...tllh4 23 tll e1 'it'xh3
24 iLfl at least staves off the first wave
of the attack) 23 il.f1 'it'h5 24 tllxa (or
24 tlle1 tllxe4 25 fxe4 f3) 24...tllxe4! 25
tllxa8 tllf2 wins, since after, for exam­
ple, 26 il.g2 'it'h2+ 27 �fl tllh4 28 tll e1
e4 White will be crushed.
22 .. 'it'xh3!

Indeed, 25 ... g2 (after 25 ... nac8 26 This move was mentioned by

tlle6 g2 27 ne1 nxc4 both 28 tllxf8 and Matamoros back in 2003 in New in
28 tllf2 give White a big advantage) 26 Chess, but it seems to have gone unno­
ne1 tllxf3+ 2 7 il.xf3 'it'hl+ 28 �2 'it'h4+ ticed. Weaker is 22 ...tllM 23 tll e 1 with a
29 �e2 tllxe4 30 il.xe4 f3+ 31 �dl f2 clear advantage for White as given by
32 il.xg2 fxel'iV+ 33 tllx el 'it'g4+ 34 il.f3 Markos.
�f3 3 5 tllxf3 nt8 36 na3 �f3 37 nxf3 23 il.f1

Th e M a r def Plata Vari a t i o n : 9 llJ e 1 lLl d 7 1 0 ii. e 3

After 23 ii..d 1 Black has 23...ll:lh4! E22) 16 ll:lxb4

(23 ... ll:lxe4 24 'it'g2) 24 ll:le1 ll:lxe4! with
familiar devastation.

While both lS ii.el and lS b4 axb4

16 ll:lbs are very ambitious, the simple
23... 'it'hs 24 ll:le1 recapture with 16 ll:lxb4, while very
No better is 24 ll:lxc7 ll:lxe4! 2S ii..g 2 obvious, almost looks naive. The knight
(otherwise, 2S ll:lxa8 ll:lgs 26 ll:lel e4 cannot defend the kingside and it is
wins, but 2 s ... ll:lf2 may be even better: also seems harder for White to support
for example, 26 ii..g 2 'it'h2+ 27 �fl ll:lh4 the advances a4-aS and c4-cS. How­
28 ll:le1 e4) 2s ... ll:lgs 26 ll:le1 e4 27 l:ta3 ever, as is often still playable and the
ll:lh4 and White cannot defend. b4-knight may hop to c6. As always, a
lot will come down to specifics.
Instead 16 ...ll:lcs 17 as! bxas
(17 ...�xas 18 �xas bxas 19 ll:ld3 is
similar) 18 ll:ld3 ll:lxd3 19 'it'xd3 gives
White a slight but clear edge according
to Markos. White will double rooks on
the a-file to win back the pawn and
Black's counterplay is not so easy.
17 ll:ld3
White has several alternatives:
24...ll:lxe41 25 fxe4 a) 17 ll:lc6 is the most direct:
2 S ii..g 2 ll:lf2 26 ll:lxe7 ll:lh4 winning. 17 ...ll:lxc6 18 dxc6 'it'e8 (18 ...ii..e 6 is also
2s ...f3 26 ll:lxf3 �xf3 27 'it'g2 ll:lf4 28 possible) 19 ll:lds �f7 20 as bxas 21 'it'a4
'it'h1 'it'g4 29 �b2 �f8 30 ll:lxc7 �f2 g4 22 'iYbs (or 22 1'..h4 ll:lxds 23 cxds g3
Black has a decisive attack - Mata­ 24 hxg3 fxg3 2S ii..x g3. as in P.Kiriakov­
moros. D.Lobzhanidze, Groningen 1996, and

A ttacking Chess: Th e K i n g 's I n dian, Vo l u m e 1

now Black should go 2S ... il.h6) 22 ...tllxds gave Black counterplay in A.Yusupov­
23 cxds g3 24 hxg3 fxg3 2s il.xg3 il.h6 G.Kasparov, Yerevan Olympiad 1996)
gives Black counterplay. 19 ...nas (or 19 ... g4 20 as gxf3 21 il.xf3
b) 17 as bxas 18 tllc6 tllxc6 19 dxc6 bxas 22 nxas nb8 23 l:ta7) 20 tlld3 g4
'it'e8 20 cs dxcs 21 il.xcs nf7 22 'it'b3?! 21 il.el na6 22 as is a line given by
'it'xc6 23 'it'bs 'it'xbs 24 il.xbs il.f8 was Markos. White's play looks faster.
better for Black in O.Wilgenhof­ d2) 18 ... g4!? and then:
F.Nijboer, Dieren 2009. d21) 19 fxg4 tllxe4 20 il.d3 tllxf2 21
c) 17 il.el and then: l:txf2 il.f6 is level according to Yusupov.
cl) 17... hs 18 h3 nf7 (worse is d22) 19 il.el gxf3 20 il.xf3 �h8 21
18 ...�h8?! 19 tlld 3! tlle g8 20 tllf2 tll h6 tlld3 tllg 6 22 as bxas 23 il.xas il.xbs 24
21 tllbs .l:tf7 22 as bxas 23 nxas nxas cxbs 'it'd7 2S 'it'b3 nfb8 26 nb1 tllh4
24 il.xas il.f8 2S cs!) 19 tll d3 il.f8 20 gave Black counterplay in M.Bley-
tllf2 is a little better for White. 1.Novak, correspondence 200S.
c2) 17 ...'it'd7 18 tlld3 tllg 6 19 tllf2 h S d23) 19 il.h4 g 3 ! ? (Black could also try
2 0 h3 tllh4 21 a s ! nf7 (or 21 ...nxas 2 2 19 ... gxf3 20 il.xf3 tll g 6 21 il.f2 'it'c8 or
nxas bxas 23 tllb s) 22 axb6 nxal 23 19 ...nf7!? with the idea 20 il.xf6 gxf3
'it'xal cxb6 24 'it'b2 'it'd8 2S tlla4 looks 21 il.xe7 fxe2 22 il.xd8 exdl'it' 23 nxdl
shaky for Black, although he drummed nxd8 24 tllxe7 nc8) 20 hxg3 tllg 6 was
up some counterplay after 2S ...g4 26 suggested by Yusupov. Then 21 tllc6
fxg4 f3 27 gxf3 tll h7 28 'it'xb6 'iVf6 in il.xc6 22 dxc6 tllxh4 23 gxh4 tllh s 24 as
O.Gladyszev-S.Soloviov, Soc hi 200S. is given by Markos, but after 24 ... bxas
d) 17 na3 il.d7 18 tll b s was Yusu­ 2S cs tllg3 I think I would prefer Black.
pov's choice and has been endorsed by 11 hs

Markos. Black should also consider

17 ... tllg 6!? with the idea 18 tllbs g4!.
1s tt::i bs tllg6
After 18 19 il.el 'it'b8 20 �hl
g4 21 as bxas 22 il.xas 'it'b7, as in
S.Ghane Gardeh-V.Kotronias, Cappelle
la Grande 2008, Markos suggests 23
'it'c2 ntc8 24 nfbl when White is way
ahead. Black could, though, consider
18 ...g4 19 il.h4 gxf3 (after 19 ... g3 20
hxg3 tllg6 21 as tllxh4 22 gxh4 Black
does not have ...tllh s available, so it is
Now Black has: not so easy to attack) 20 il.xf3 tllg4
dl) 18 ...�h8 19 'it'al! (19 il.el ng8 with counterplay.

Th e M a r def Plata Va ria t i o n : 9 tDe1 tb d 7 1 0 ii.e3

Dembo also mentions 24...h 3 ! ?.

2 s nxas?!
After 2S cxd6! cxd6 26 tDc7 nb8 27
tDe6 ii.xe6 28 dxe6 tDg3 the position is
very messy.
2s nxas 26 ii.. x as ii..h 6!

19 a 5
Instead 19 'if el is interesting, taking
a look at both the as- and h4-squares:
19...g4 20 as h4! ? 21 ii..xh4 (or 21 axb6
nxal 22 'it'xal g3 23 ii.el h3 21 fxg4
tDxe4 22 axb6 .l:txal 23 'it'xal cxb6 with
counterplay - Golubev) 21... tDxM 22 Black has excellent attacking chances.
'it'xh4 nxas 23 lDb4 was E.Guseinov­ 27 ii..d 2
M.Golubev, Moscow 2006. Now 23 ...1'..d7 Dembo also gives the lines 27 t:Df2
24 t:Dc6 ii..xc6 2S dxc6 'it'a8 26 tDxe7 (or h3, 27 t:Dxd6 h3!, 27 tDxe7 ii..e 3+ 28 �h2
26 nxas bxas with the idea of 27 cs?! 'it'gs, 27 cxd6 ii.. e 3+ 28 �h2 'it'gs and 27
'it'xc6 28 1'..c4+? dS 29 exds 'it'xcs+, win­ ii.xc7 ii..e 3+ 28 nf2 (or 28 tDf2 'it'gs 29
ning) 26 ...'it'xc6 27 tDe6 ne8! ii.. x d6 tDgf4 30 g4 hxg3) 28...'it'gs. In all
(E.Guseinov) 28 t:Dxg7 �xg7 29 fxg4?! cases Black retains a strong attack
t:Dxe4 gives Black the initiative. The 27 ii..x d2 28 'it'xd2 t:Dgf4 29 tDxf4

threat is 30 ... 'it'cs 31 �hl tDg3!. If 29 cxd6? 'it'gs 30 t:Dxf4 t:Dxf4 threat­
19. bxas 20 cs
•• ens both mate on g2 and 31...tDh3+. After
20 ii.el g4 21 nxas .l:tb8 22 lDb4 nb7 31 �2 'it'xg2+ 32 �e3 (or 32 �el ii..h 3,
23 tDc6 'it'e8 24 na8 gave White a big opening the way for Black's rook to a8)
advantage in Z.Kozul-D.Rogic, Bled 32 ...cxd6 Black's attack is decisive.
1997, but Black should try 20...a4 or 29 tDxf4

21 ...1'..d 7. Black is much better. Indeed, after

20 g4 21 'it'c2 g3! 22 hxg3
••• 30 .l:tdl 'it'gs 31 ii.fl 'it'g3 32 nc1 tDh3+
After 22 ii.el t:De8 23 hxg3 fxg3 24 33 �hl t:Df2+ 34 �gl h3 White re­
ii..x g3 h4 2S ii..f2 'it'gs Black has attack­ signed in K.Nikolaidis-Y.Dembo, Corfu
ing chances. 2007, in view of 3S 'ikh6 hxg2 36 ii..x g2
22 fxg3 23 ii..x g3 h4 24 ii.el tDhs
••• Ji.h3 OT 3S 'ifxf2 h2+.

Chapter 4
The Mar del Plata Variation
9 '2Jd2

1 d4 ll:lf6 2 c4 g6 3 ll:lc3 il.g7 4 e4 d6 5 White intends to force through c4-c5,

ll:lf3 o-o 6 il.e2 es 7 o-o ll:lc6 8 ds ll:le7 9 even as a pawn sacrifice, after which
ll:ld2 the d2-knight will take up a tremen­
dous post at c4. From there it will pres­
sure d6, while its control of b6 often
ends up being important as well.
At first 9 ll:ld2 looks more active
than 9 ll:lel, but if White cannot effec­
tively reposition the knight, White's
development can become a bit clogged
up. Logically Black should look to pre­
vent White from playing cs and ll:lc4
effectively. In that sense 9 ... cs might be
considered Black's most natural re­
By playing 9 ll:ld2 White takes a very sponse, blocking the position. At one
optimistic view towards his queenside point this was, indeed, the main line,
play. This line was the main battle­ but after 10 �bl ll:le8 11 b4 b6 12 bxcs
ground in the early 1990s until 9 b4 bxcs 13 ll:lb3 White's knight finds use­
experienced a tremendous boom in ful employment via another route. For
popularity. Like 9 ll:lel, 9 ll:ld2 prevents example, 13 .. .fs 14 il.g s! h6 15 il.xe7
Black's ... ll:lhs and at the same time 'Wlxe7 16 ll:las and the knight becomes
prepares to bolster his centre with f2- a menace on c6.
f3. In this case, however, instead of re­ Another idea is to play a 'straight
routing the knight to help support the race' with 9 ...ll:ld7 (or 9 ...ll:le8 - the two
c4-c5 break (for example with ll:le1-d3), moves will generally transpose after

The M a r def Plata Vari a t i o n : 9 tll d 2

the knight returns to f6) 10 b4 fS 11 cs 11...c6 when in each case White does
�f6 12 f3 f4 13 tllc4 gs 14 a4 ( 14 il.a3 is not really have anything better than 12
also possible) 14...tll g 6 lS il.a3 l:tf7. a3, which will just transpose back into
Here both Beliavsky's 16 as and the our main lines, below.
main line, 16 bS, are theoretically very After 10 a3 Black faces a very impor­
dangerous, although Black has had tant strategic decision. First of all, he
some recent successes, especially in the must avoid the race. Even though 9 ... as
later variation after 16 ... dxcs! 17 il.xcs has gained a tempo of sorts, it has
hS 18 as g4 when he has dangerous weakened Black's queenside and after
counterplay. the analogous 10 a3 tlld7 11 l::tb l fS 12
9 as
... b4 tllf6 13 f3 f4 14 cs axb4 lS axb4 gs
16 tllc4 tllg 6 17 bS! (worse is 17 il.a3
�f7) 17...dxcs 18 b6! Black's queenside
is destroyed before he could even
dream of a kingside attack. The block­
ing plan with 10 ... cs is more respect­
able, although after 11 l:tbl tlle 8 12 b4
axb4 13 axb4 b6 14 bxcs bxcs lS tllb3
White again does pretty well. Black
should still consider this structure in
some cases, though, and we will allow
it sometimes if we can obtain a rela­
This is established as the main line. tively favourable version.
Black prevents b4 and in a sense he Traditionally the big choice has
gains a tempo, because White will have been between 10...tlld7 and 10 ... il.d7.
to play both a3 and l:tbl in order to Originally I had hoped to include both
achieve the b4-advance. In the 9 tllel lines, but then decided that, with 9
line, this move would be ineffective be­ tlld2 being considered less critical than
cause White could play something like 9 tllel and 9 b4 these days, one main
a3 and il.d2 when the al-rook is pro­ line should suffice. I've always kind of
tected by very natural means. After 9 liked the move, because Black
tlld2, however, the d2-knight prevents may actually play on either side of the
the cl-bishop from moving, so White board. Instead 10 ...tlld7 actually mir­
does not have this plan available. rors White's developmental logjam, but
10 a3 because 10 ...tlld7 is twice as popular
White can also play 10 l::tb l, when we and a more traditional 'attacking'
can still play 10 ... il.d7. After 11 b3 (11 a3 move, I began to delve into its intrica­
a4) Black can play 11 ...tll c 8, 11 ...tll e 8 or cies first.

A ttacking C h e s s : Th e King 's I n dia n , Vo l u m e 1

I decided to avoid 10 ... ll:ld7 because This line was Gallagher's recom­
of 11 nb1 fS 12 b4 �h8 mendation back in 2004 and it was
played successfully by Kasparov. How­
ever, the more I delved into White's
possibilities, the more disheartened I
became. White can prepare the cs­
break in various ways, such as with 14
tllb3, 14 tlla4 and 14 tll bs, and in fact
even the immediate 14 cs!? is danger­
ous. As it turns out, even if Black tries
to prevent this break with ...b6, White
can almost always sacrifice a pawn
with c4-cS anyway. Once that knight
13 f3 (after 13 'it'c2, I do not like gets to c4, White obtains excellent play.
13 ... tllg 8 14 exfs gxfs 1s f4, but Although Black has not done so
13 ... tllf6! 14 f3 axb4 l S axb4 c6! is an badly in practice with 10 ... tlld7, while
effective way for Black to deal with studying the relevant games I found it
White's threat to play cs and tllc4, and much easier to find improvements for
16 dxc6 ll:lxc6 has proved to be satisfac­ White than for Black Black's pieces all
tory for him) 13 ...f4 (after 13 ... axb4 14 seem to be a bit jumbled up. I once
axb4 c6 lS dxc6! ? ll:lxc6 White has came across a comment by Golubev
gained as his queen has not gone to c2 that said something to the effect that
and Black's knight is still on d7, so he is he was not convinced that combining
quicker to attack the d6-pawn) 13 .. .f4 ... as and ... tlld7 had to be logical, and I
(after 13 ...tll g 8 lvanchuk's 14 'it'c2 ll:lgf6 have come to understand this reason­
l S il.d3 ! forces 1S ...f4 when Black's ing. After all, the 'logical' 9 ... cs and
knights are suddenly misplaced). 9 ... as 10 a3 cs do not stop White's ini­
tiative even though they prevent the
c4-cS advance. While 9 ... as and
10 ...tlld7 seems to discourage White's
breakthrough, they do not prevent it.
So after a lot of time spent on
10 ... tlld7, I felt I had to abandon it, al­
though I think an understanding of
these positions is still beneficial, espe­
cially the idea of playing a pre-emptive
strike with ... c6. However, when I came
around to investigating 10 ... il.d7, I was

Th e M a r def Pla ta Varia tio n : 9 tll d 2

quite pleased with the play that arose. initiative i n L.Brunner-J.Nunn, Nurem­
10 il.d7
... berg (rapid) 1990) 17 ...tllf6 is fine for
Black. White's queenside play has not
really gone anywhere.
a2) 16 f3 tllf6 (alternatives are
16 ...�h8 and 16 ... cs) 17 il.e3 cs!? (or
17 ... tllh s ! ?) 18 dxc6 il.xc6 was fairly
level in E.Gleizerov-1.Nataf, Montecatini
Terme 1997.
b) 11 .ll a2 hopes to obtain an im­
proved version of variation 'a' by main­
taining the rook on the a-file: 11 ... cs (I
still think 11 ...a4 is playable and after
This move not only develops a piece, 12 b4 axb3 13 tllxb3 Black could inves­
but in effect it gains another tempo tigate 13 ... c6, 13 ... cs. and 13 ... b6) 12
because White usually chooses to pre­ dxc6 (after 12 b3 tlle 8 13 .llb2 fS 14 b4
vent Black from damaging his pawn axb4 lS axb4 b6 White has wasted a lot
structure with ... a4. of time, so Black should be fine)
11 b3 12 ...bxc6! (this is an ideal circumstance
This by far the most common move, for this recapture, as White won't re­
preventing Black from locking down gain control of the dS-square) 13 b4
the queenside with ... a4. It is also pos­ axb4 14 axb4 .llxa2 lS tllxa2 'it'a8 (also
sible to ignore Black's idea: possible are 1s ...'it'b6!? and 1s .....lfi.e6!?)
a) 11 .l:tbl a4 12 b4 axb3 13 tllxb3 16 tllc3 dS and Black had the initiative
(after 13 .llxb3 b6 14 a4 Black should in G.Flear-M.Gurevich, Tel Aviv 1989.
not bother with 14 ... il.xa4 lS tllxa4 Instead after 11 b3 White is all set
.llxa4 16 .l:txb6!, but instead just play to play 12 llbl and 13 b4.
14 ... tlle8 lS .lla3 fS with counterplay)
13 ...b6 (alternatives are 13, 13 ... c6
and 13 ... cs 14 dxc6 il.xc6 lS 'it'd3 tll e 8!
16 .lld l fS) 14 .ll al (White is ready now
to push the a-pawn, but he has wasted
a lot of time) 14 ... tlle 8 (stopping the a­
pawn with 14...'it'e8!? is also an attrac­
tive idea) lS a4 fS and now:
al) 16 as bxas 17 tllxas (after 17
.llxas .llx as 18 tll x as 'it'a8 19 tll b 3 tllf6
20 il.d3 .llb8 Black already had a nice

A ttacking C h e s s : Th e Kin g 's I n dian, Vo l u m e 1

One thing I like about the 10 ... 1'.d7 ...ll:ib6 which will allow Black to meet
line is the fact that Black's play is flexi­ b4 with ...ll:ia4. Exchanging this knight
ble and there are a few decent options not only provides Black with some re­
here: lief from his spatial disadvantage, but
he may sometimes even take the initia­
A: 11 2t)c8
... tive on the queenside. We will see a
8: 11-.lhel similar plan frequently in the Petrosian
C: 11 c6... Variation, although there Black plays
...ll:ia6-cS-a4. Black may also initiate
It is also possible to play 11 ... cs, be­ play with ... c6. If White is not careful he
cause Black will end up a tempo ahead can very easily lose the initiative on the
of the line 11 a3 cs as White will queen side.
have to spend two moves with his b­ 12
pawn. Still, after 12 ll:ie8 13 b4 This is the natural move, but there
axb4 14 axb4 b6 lS bxcs bxcs 16 ll:ib3 are others:
(16 'i¥b3 ll:ic8! exploits Black's extra a) 12 l:ta2 always looks strange to
tempo by covering b6) 16 ...fs 17 f3 lLif6 me. If White plays b4 soon he will end
18 ii.d2 f4 19 hi.xal 20 'ifxal g s 2 1 up with a knight on a2, which hardly
Va s White has the initiative. I n general seems ideal for him. After 12 ...ii.h6
Black should be willing to accept this (this is a typical idea for Black in the 9
structure, but only on his own terms. ll:id2 variation) 13 'ifc2 ll:ib6 14 htdl
Here even the extra tempo does not 'i¥e7 lS 'ifd3 ll:ie8 16 h!.c2 fs Black is al­
guarantee Black equality. ready comfortable. White felt the need
to complicate with 17 cs dxcs 18 ll:ic4
A) 11 ll:icS
... ii.xcl 19 d6 ll:ixd6 20 ll:ixb6 fxe4 21
ll:ixe4 ii.ts 22 'ifds+ �g7 23 l:tcxcl
ii.xe4 24 'ifxcs cxb6 2S 'i¥xd6 'i¥xd6 26
h!.xd6 ii.c6, but then Black was just up a
pawn in S.Skembris-V.Kotronias, Xanthi
b) 12 ii.b2 ll:ib6 (12 ... cs and 12 ...ii.h6
have also been tried) 13 'ifc2 'ife7
(again 13 ... ii.h6 is possible) 14 b4 ii.h6
(or 14... axb4 lS axb4 ii.h6 16 hi.fdl c6!?)
1S ll:ib3 axb4 16 axb4 ll:ia4 17 ll:ias (17
ll:ixa4 ii.xa4 is fine for Black) 17...ll:ixb2
I have always been attracted to this 18 'i¥xb2 h!.ab8 19 bs ii.e8! (Black is
funny-looking move. The idea is to play ready to block the queenside with

Th e M a r d e f Plata Va ria t io n : 9 lLi d2

...ll:id7 and this causes White to panic) ll:ixb6 'ifxb6 23 ll:ic3 Vas is fine for
20 b6?! cxb6 21 ll:ib3 (21 'i¥xb6 ii.d2 Black) 22 ll:ibc3 ii.e8 23 ll:ibs ll:ic6 24
wins the e-pawn) 21 ... 'ifc7 left Black ll:idc7 Itb8 2S ll:ixe8 Itxe8 26 'i¥d3 ll:id4
with an extra pawn and the bishop­ 27 ll:ixd4 exd4 and Black was at least
pair in S.Kishnev-B.Damljanovic Sibenik equal in l.Cosma-B.Damljanovic, Bel­
1990. grade 199S.
c) 12 'i¥c2 ll:ib6 13 a4! ? sees White c3) 13 ...ii.h6 ! ? is untried, but 14 1'.a3
switch plans, hoping to make the b6- 'i¥e7 lS b4? would run into 1S ... ll:ixa4!,
knight look silly. He may continue with exploiting the loose knight on d2.
ii.a3 and b4. Still, this is time­ 12 ll:ib6

consuming and freezing the queenside

looks strange.

13 'i¥c2
After 13 b4 axb4 14 axb4 ll:ia4 lS
Black has: 'i¥c2 Black has:
cl) 13 .. lte8 14 ii.a3 ii.f8 1s ll:i bs cs?! a) 1s ...ii.h6 16 ll:ibs ii.xbs (16...'ife7?!
(1s...ii.xbs 16 cxbs ll:ibd7 is better, but 17 cs! is variation 'b21' to Black's 13th,
after 17 b4 White still has an edge) 16 below) 17 cxbs ll:ib6 18 ll:ib3 ii.xcl 19
dxc6 ii.xc6 17 Itfdl 'i¥b8 18 ll:ic3 ll:ibd7 Itbxcl ll:ie8 20 ll:ias Itb8 was perhaps
19 ll:if1 ll:ics 20 ii.xcs! dxcs 21 ll:ie3 gave slightly better for White in
White a positional advantage in D.A nagnostopoulos-1.Nikolaidis, A no
D.Komarov-Kr.Georgiev, Lyon 199S. Liosia 199S, but Black is very solid.
c2) With 13 ...ll:ic8 Black is satisfied b) 1s ... ll:ixc3 16 'ifxc3 ii.h6 (this is
that the knight has done its job and logical, but Black could consider 16 ...c6,
now prepares to block the queenside: 16 ...h!.a2 or 16 ...'ife7) 17 ii.b2 ! ? (White
14 ii.a3 b6 lS b4 axb4 16 ii.xb4 cs 17 intends f4, so Black is obliged to grab a
dxc6 ii.xc6 18 ll:ids ll:id 7 19 1'.g4 ll:ics 20 pawn) 17 ... ii.xd2 18 'i¥xd2 ll:ixe4 19 'i¥e3
ii.xcs bxcs 21 ll:ib1 ll:ia7 (or 21 ...ll:ib6! ? fs 20 f4 exf4 21 h!.xf4 gives White com­
when 22 ll:ibc3 fails to 22 ...ll:ixc4 and 2 2 pensation for t he pawn.

A ttacking C h e s s : Th e King 's In dian, Vo l u m e 1

13 'i¥e7
•.. a4!? ll:id7 18 ll:if3 ll:ics 19 ii.a3 b6 20
Black connects his rooks and covers ii.xcs bxcs 21 g3 ll:if6 the position was
the cs-square. He also has: fairly level in M.Gurevich-1.Smirin,
a) 13 ... ll:ie8!? 14 b4 axb4 lS axb4 Haifa 199S.
ll:ia4 16 ll:ixa4 ii.xa4 17 'if c3 ii.d7 18 cs 14...axb4 15 axb4
fs 19 f3 ii.h6 (better is 19 ...ll:if6) 20 c6
and White was ahead in K.Lerner­
G.Pieterse, Amsterdam 1988.
b) 13 ... ..ih6 and here:
bl) 14 b4 axb4 lS axb4 ll:ia4
(1s ... 'i¥e7 16 c s ! ) 16 ll:ibs leads back to
13 b4, above, while avoiding 1S ... ll:ixc3,
although this is no special accom­
plishment for White.
b2) 14 ll:ibs 'i¥e7 lS b4 (White can­
not take on c7 because of ... l:tfc8, so he
just continues his queenside play) 1s ...ll:ia4
1S ... axb4 16 axb4 and here: Or 1S ... l:tfc8 16 ll:ib3 ll:ia4 17 ii.d2
b21) 16 ... ll:ia4?! 17 cs! is a trick Black ll:ixc3 18 ii.xc3 c6 19 dxc6 ii.xc6
must watch out for. After 17 ... dxcs 18 (19 ...bxc6 !?) 20 f3 dS! 21 cxds ii.xds 22
ll:ixc7 l:tac8 19 ll:ic4! he was clearly 'i¥b2 ii.e6 23 ll:id7 with equality in
worse in G.Kamsky-V.Spasov, Tilburg A.Beliavsky-Ki.Georgiev, Groningen
1992. 1994.
b22) 16 ...h!.fc8 17 cs ll:ia4 was 16 ll:ixa4
L.Psakhis-V.Spasov, European Team White acquiesces to the exchange
Championship, Debrecen 1992, and of knights. Instead 16 ll:id1?! is too pas­
now the simple 18 lLif3 ii.xcl 19 l:tfxcl sive. After 16 ...ll:ig4!? (16 ... c6 also looks
gives White the initiative. good) 17 f3 (or 17 h3 ll:ih6 with the idea
b23) Best is 16 ... c6!, a typical pre­ of ... fs) 17 ...ll:if6! 18 ll:ib3 ll:ihs (the point
emptive strike: 17 dxc6 ii.xc6 18 ll:ic3 of Black's manoeuvres) 19 ll:ias ll:if4!
l:tfc8 19 ll:ib3 ii.xcl 20 hi.fxcl ll:ia4 was Black had the initiative in
fine for Black in Wu Shaobin-M.Al Mo­ R.Kasimdzh anov-1.Smirin, Elista Olym­
diahki, Beijing 199S. piad 1998.
14 b4 After 16 cs Black should avoid
This is consistent. Instead 14 ii.b2 16 ... ii.h6 17 ll:ibs! and play either
ii.h6 lS l:tbel 1'.g4!? is possible, which 16...h!.fc8 or 16 ... ll:ixc3 17 'it'xc3 c6!?,
is an idea we will see too in the Petro­ with the idea of 18 cxd6 'it'xd6 19 lLic4?!
sian Variation. After 16 ii.xg4 ll:ixg4 17

Th e M a r def Plata Va riati o n : 9 lLid2

16 ... ii.xa4 17 'ifc3 26 �g2 'ifd7 21 ll:ig1 ii.xe2 28 ll:ixe2 fs

After 17 'it'd3 the moves 17 ... c6, Trading light-squared bishops is not
17 ...l:tfc8, and 17 ...1'.d7 all look reason­ always fatal in the King's Indian if Black
able. can generate counterplay. Here Black
11...ii.h6 could also consider 28 ... 'ifb5 29 ll:ic3
Pressuring the e4-pawn. 17... c6!? is 'i¥c4 with equal chances.
another idea. 29 ll:ic3 ll:if6 30 f3 l:tfc8 31 'i¥d3 l:ic7 32
18 'i¥d3 ll:ie2 h!.xcl 33 h!.xcl h!.a2 34 h!.c2
Instead 18 f3 is met by 18 ...ll:ih5!. After 34 ll:ic3 fxe4! 35 fxe4 l:tb2 36
White can again consider sacrificing l:tbl l:txbl 37 'ii' x bl 'i¥c7 38 'it'd3 'ifb6
a pawn with 18 ii.b2 ii.xd2 19 'i¥xd2 Black has counterplay. His queen is ac­
ll:ixe4 20 'it'e3 f5 21 f4, although this is tive and the e4-pawn will always need
somewhat speculative. attention. Given time, Black can also
18 ... c6 19 l2Jf3 consider something like ...h 5, ...�h7 and
34...fxe4 35 fxe4 'if a4
Black had sufficient play in
M.Gurevich-1.Smirin, Elenite 1994.

B) 11...ll:ieS

19 ...1'.g7!?
Black keeps pieces on the board. Af­
ter 19 ...ii.xcl 20 l:tfxcl Mikhail Gure­
vich believes White is better, but Black
looks okay to me.
20 cxds 21 cxds h6 22 ii.d2 1'.d7
Black could also consider playing This typical King's Indian move has
22 .. JUc8 immediately. recently garnered some attention.
23 h!.bcl 'ifdB 24 'it'b1 Black simply prepares .. .f5.
24 b5 can be met by 24...h!.a4! when 12 f5 13 b4
the e4-pawn is in trouble. Instead 13 f3 is well met by
24... ll:ihs 2s g3 1'.g4 13 ... ii.h6, while Black could also con­
Not 25 ... f5? 26 ll:ih4. sider 13 .. .f4.

A ttacking C h e s s : The King 's I n dian, Vo l u m e 1

(the complications begin) 18...fxe4 19

fxe4 dxcs 20 d6 cxb4 21 dxe7 'ifxe7 22
ll:ia4 ll:ig4 23 ll:ib6 ll:if2+ 24 h!.xf2 h!.xf2
2 s 'it'e1 (2s ll:ixa8 'ii'g s!) 2 s ...l:taf8 26
ll:ixd7 'ii'g s ! 27 g3 hi.8f7 28 ll:ics 'it'e3 29
1'.c4 was R.Nolte-W.So, Tagaytay City
2010. Now 29 ... 'ifxcs 30 ii.xf7 hi.xf7
would be unclear.

13 ...axb4 14 axb4 ll:if6

This is by far the commonest move,
but Black could also consider 14...�h8
1s cs ll:ig8 16 f3 ll:igf6 17 ll:ic4 ll:ihs or
14... ii.h6 lS cs fxe4 16 ll:idxe4 ii.xcl 17
'ifxcl ll:ifs.
15 cs
The most direct. Instead lS f3 c6!?
(or just 1 s...ii.h6), and here: 15 ...fxe4
a) 16 cs!? cxds 17 exds ll:iexds 18 Black gives up the e4-square in or­
ll:ixds ll:ixds 19 1'.c4 ii.e6 20 ll:ie4 (this is der to generate active piece play him­
the point of White's play, but it does not self. There are a couple of alternatives:
lead to an advantage) 20...fxe4 21 ii.xds a) 1s ...�h8 16 f3 ll:ihs 17 ll:ic4 ll:if4
'ifd7 22 'tlfb3 ii.xds 23 'ifxds+ �h8 24 18 ii.e3 ll:ic8 (Black covers the d6-pawn

'ifxe4 ds 2 S htdl l:tad8 was level in and intends ...'if gs with counterplay)
C.Bemard-A.David, Andorra 1996. 19 �hl! g s 20 ll:ias (20 b s also looks
b) 16 dxc6 ll:ixc6 17 ll:ibs (17 ll:ib3 good) 20...'ife8 21 ll:ixb7 h!.a3 was
ii.e6 18 ll:ibs fxe4 19 ll:ixd6 exf3 was M.Gurevich-B.Gelfand, Munich 1993.
unclear in E.Gleizerov-1.Zaitsev, Podolsk Now 22 l:tcl seems to refute Black's
1992) 17 ... ii.e6 18 'it'c2 'ife7 (or 18 ... l:tc8, play because after 22 ... h!.xc3 23 h!.xc3
as in R.Scherbakov-J. Nunn, H astings ll:ixe2 24 'ifxe2 2S h!.c4 Black can­
1993/94) 19 ii.d3 ll:ihs (both 19 .. .f4!? not exploit the pin and White will re­
and 19 ...ll:ie8 ! ? are possible too) 20 cs main a good pawn up.
dxcs 21 'ifxcs was V.Chuchelov-J.Nunn, b) 1s ... ii.h6!? is Black's latest at­
Leeuwarden 199S, and now 21 ...ll:if4 tempt to fight for the initiative: 16 ll:ic4
would give Black a slight initiative. (16 ii.b2 and 16 f3 are possible as well)
c) 16 �h1 �h8 17 ii.b2 h s 18 c s ! ? 16 ... ii.xcl (perhaps Black should prefer

Th e M a r def Plata Va ria t io n : 9 lLid2

16 ... ll:ixe4!? 17 ll:ixe4 1'.a4 18 'ifel ii.xcl 11...h6

19 l:txcl fxe4) 17 'ifxcl (after 17 h!.xcl This natural move has always been
ll:ixe4 18 ll:ixe4 fxe4 19 cxd6 cxd6 20 played, even though 17 ... 'i¥e7 and
ll:ixd6? is not possible because of 17 ... 'i¥e8 look reasonable. In fact I be­
20 ... ii.a4) 17 _.fxe4 18 cxd6 and here: lieve the strongest move is 17 ... ll:ixe4!
bl) 18 ...ll:iexds 19 ll:ixds ll:ixds 20 18 ll:ixe4 (18 ii.xd8 ll:ixc3 19 'if el ll:ixe2+
ll:ixes ii.e6 21 dxe7 'ifd6 (no better is 20 'i¥xe2 l:taxd8 gives Black three won­
21...'ifxe7 22 'i¥xc7 ll:ixc7 23 l:tbcl) 22 derful minor pieces for the queen)
'i¥b2 and White is more comfortable. 18 ... 'i¥e8, which looks fine for Black,
b2) 18 ...ll:ifs! ? and then: while here 18 ... 1'.a4!? may be possible
b21) 19 ll:ixes?! cxd6 20 ll:ic4 l:tc8 as well.
(20... bs!?) 21 'if d2 hi.e8 22 l:tfel and 1s ii.xf6 ii.xf6 19 bs
now 22 ... e3! 23 ll:ixe3 ll:ixe3 24 fxe3 White's most recent try was 19
'ifb6 2 S ii.f3 hi.c4 gave Black excellent and 19 ... l:txal 20 'ifxal! 21 'ifa7
play for the pawn in A.Khalifman- was much better for him in A.Beliavsky­
1.Cheparinov, Amsterdam 2007. Z.Lanka, Austrian League 2007. Black
b22) 19 dxe7 'i¥xc7 20 d6! ? (20 l:tdl can improve with 20 ... ll:id4 21 ii.d3 ii.el
e3! 21 ll:ixe3 hi.fc8 22 ll:ic4 bS is not when White has a small edge at best
clear, but both 20 'i¥d2 and 20 'i¥b2 look after 22 'i¥a2.
somewhat more pleasant for White) 19...b6 20 c6 ii.cs 21 ii.d3 22 l:ta1
20 ... ll:ixd6 21 ll:ixes ii.ts? (21 ... hi.ac8) 22 l:txal 23 'ifxal
ll:ixe4! won a pawn in V.Chuchelov­
J.Degraeve, German League 2008.
16 ll:idxe4 ll:ifs 17
White is willing give up his dark­
squared bishop to consolidate his grip
on e4.

23 ... hs
This is Nunn's recommendation. In­
stead 23 ...1'.f4?! 24 ll:ie2 'ifh4 was
G.Kamsky-J.Nunn, Monaco (blindfold)
199S, when Nunn gives 2S g 3 ! 'ifh3 26
ll:ixf4 exf4 27 hi.el! ll:ih4 28 gxh4 'ifxd3

A ttacking Chess: Th e Kin g 's I n dian, Vo l u m e 1

29 ll:if6+ �f7 30 'i¥a7! and White is allows Black t o show another idea be­
winning. hind 11 ... c6.
After 23 ... hs Lukacs and Hazai give
24 ll:ixgs 'ifxgs 2s f4 (?!) exf4 26 ll:ie4
'i¥e7 27 h!.xf4 as slightly better for
White, but Black can play the more ac­
tive 26 ...'t!Vh4! 27 'ii'a1 ll:ie3 28 'if xe7 ii.ts
with the idea ... 'ifg4. After 29 g3 fxg3
30 hxg3 'i¥g4 Black has the initiative.

C) 11...c6

12 ... bs1
This opportunistic move gives Black
good counterplay. He can also play
more quietly with 12 ...'ifb8 13 b4 axb4
(after 13 ... cxds 14 cxds hi.c8 White can
play lS ll:ic4! targeting b6 and as, as in
D.Giacomazzi-A.Shchekachev, Metz
199S) 14 axb4 (White may be better off
with 14 l:txb4 as in M.Notkin­
This move was once rather uncom­ V.Bologan, Saint Petersburg 199S,
mon, but has become established as when Black should probably play
the main line. Black creates more ten- 14 ... l:ta7) 14... cxds 1s cxds l:tc8 (or the
sion in the centre and keeps his options immediate 1s ... bs) 16 ii.b2 bs
open on both sides of the board. White (16 ... ii.h6!?) 17 ii.d3 'ifb6 18 ll:ib3 ii.h6
now has a choice: 19 ll:ias hi.Cl gave Black good play in
l.Nemet-J.Gallagher, Swiss League
C1: U .:.%b1 1994.
C2: 12 lla2 13 dxc6
C3: 12 1'.b2 Taking the pawn with 13 cxbs gives
Black excellent chances after 13 ...cxds
Cl) 12 l:tb1 14 exds ii.ts! (but not 14... ll:iexds? 1s
ll:ixds ll:ixds 16 ll:ie4) 1s l:tb2 ll:ifxds 16
(seefollowing diagram) ll:ixds ll:ixds.
Ignoring Black's play with 13 b4 also
This is the most natural move, but it comes to nothing after 13 ... axb4 14

Th e M a r def Plata Va ria tio n : 9 l2Jd2

axb4 bxc4 15 l2Jxc4 cxds 16 exds ii.ts Benasque 1999.

17 hi.b3, which was V.Chuchelov­ c) 14 axb4 axb4 and now:
J.Filipek, Leuven 1999, and here Black cl) 15 cxd7 fails again after
has 17 ... l2Je4!, with the idea of 18 ltJas 1 5 ... bxc3 16 l2Jf3 l2Jxe4.
'if c8! when he enjoys excellent play. c2) 15 ltJds l2Jxc6 16 l2Jxf6+ ii.xf6 is
13 . b4!
.. certainly fine for Black
c3) 15 C7 'ifxe7 16 ltJbs (after 16
ltJds ltJexds 17 cxds l:ta2 18 ii.d3 hi.b8
19 Itb2 Itxb2 20 ii.xb2 ii.e8! 21 l2Jc4
l2Jd7 22 ltJcs Black was totally
equal in J.Gokhale-P.Mahmoud, Cal­
cutta 1996), and here 16 ... ii.xbs 17
cxbs ds gives Black counterplay accord­
ing to Bologan, but 16...'ifcs ! looks even
14...l2Jxc6 1s l2Jxf6+
Instead 1 5 ii.b2 ii.e6 16 l:tal l2Jd7 17
White's jumbled queenside pieces f4?! exf4 18 ii.xg7 �xg7 19 hi.xf4 ltJcs
allow Black to fight for the initiative. 20 l2Jf1 'ifgs was much better for Black
14 ctJdS in L.Ftacnik-R.Gadjily, Moscow Olym­
Instead 14 cxd7 bxc3 15 l2Jf3 l2Jxe4 piad 1994, and Black indeed managed
is the main tactical justification of to score a notable upset.
Black's play. White does have a few 15 ...ii.xf6 16 l2Jf3 'i¥e 7 17 a4
other tries:
a) 14 C7 looks too fancy after
14 ... 'ifxe7 1 5 ltJds ltJexds 16 cxds a4!
(Black's better development allows him
to take over the initiative on the
queenside) 17 axb4 h!.fb8 18 (or 18
bxa4 ii.xa4 19 'ifel ii.c2 20 h!.b2 l2Jxe4
when Black is much better) 18 ... 'i¥c3 19
ii.a3 axb3 20 l:tcl 'i¥d4 and White posi­
tion fell apart in G.Grigore­
V.Nevednichy, Calarasi 1995.
b) 14 ltJbs is fairly safe. After This structure is fine for Black He
14 ... ii.xc6 15 ii.f3 'tlfb8 16 a4 l2Jd7 17 has good squares for his pieces and can
Itel ltJcs 18 l2Jf1 l:td8 the position was easily cover his only weakness on d6.
level in G.Grigore-J.Baron Rodriguez, 17 ...l2Jd8!

A ttacking C h e s s : Th e King 's I n dian, Vol u m e 1

The knight belongs on cs. b ) 12 .. JWb6 13 hi.c2 cs 14 l:tb2 l:tfb8 lS

18 ii.h6 l:te8 19 'ifd2 ii.c6 20 ii.d3 ll:ie6 ll:idbl 'i¥d8 16 b4 axb4 17 axb4 b6 was
Black had an excellent position in very solid for Black in M.Gurevich­
the game P.Lukacs-W.Uhlmann, Vienna A.Shirov, Prague (rapid) 2002. Black has
199S. He intends moves like ...'ifb7, at least managed to connect his rooks
...ii.d8 and ...ll:ics. here and can hold firm on the queen­
C2) 12 l:ta2 c) 12 ...ll:ie8 is very sensible and looks
fine. After 13 b4 axb4 14 axb4 l:txa2 l S
ll:ixa2 cxds 16 cxds f s Black had good
play in M.Gurevich-V.Babula, German
League 2000.

This move always looks strange, but

it does prepare the b4-advance without
allowing Black the counterplay we saw
in Line Cl.
12...'ifbS 13 hi.c2 hi.cs 14 ll:idb1
Here 12 ... bs? fails to 13 dxc6 b4 14 A good example of Black's play was
cxd7 bxc3 because White has lS ll:ibl! 14 ii.d3 cxds 1 S cxds bs 16 b4 axb4 17
when 1S ... ll:ixe4 loses to 16 ii.f3 fs 17 axb4 ll:ihs 18 hi.el ii.h6 19 ll:idbl ii.xcl
ii.xe4 fxe4 18 'i¥xd6. 20 'ifxcl hi.c7 21 'i¥e3 fs 22 f3 f 4 23 'i¥e2
Instead 12 ... cs is as usual solid but gs 24 l:tec1 l:tb7 2s ll:id1 ll:if6 26 g4 h s
stodgy after 13 h!.b2 ll:ie8 14 b4 axb4 lS 27 h 3 h!.a4 2 8 l:tb2 'i¥a7+ 2 9 'iff2 'i¥d4
axb4 b6 16 bxcs bxcs. Others: with good play in V.Chuchelov­
a) 12 ... ii.h6 13 h!.c2 cs 14 ll:idbl! 1'.g7 V.Bologan, Istanbul 2003.
(14...ii.xcl lS l:txcl with the idea of f4 14 ... cxds
gives White the initiative) 1 S l:tb2 ll:ie8 Black can also play 14 ... bs when lS
16 b4 was M.Gurevich-R.Kasim­ cxbs cxbs leads to the same position.
dzhanov, Hoogeveen 1999. and here 15 cxds bs
16 ...axb4 17 axb4 b6 18 bxcs bxcs 19 Black has a piece formation that is
ll:ibs fs 20 ll:i1c3 looks better for White. very typical of the Ruy Lopez or the Old

Th e M a r def Plata Varia tio n : 9 lLid2

Indian Defence. With such good devel­ Black had good counterplay in
opment, Black is not afraid to initiate A.Beliavsky-D.Stellwagen, Amsterdam
play on the queenside. 2006.
16 b4
C3) 12 ii.b2

This is a very flexible move, not This move makes a lot of sense.
rushing to determine the queenside White completes his development and
pawn structure. Black has also resolved does not declare his intentions just yet.
the tension immediately: 12 ...ii.h6
a) 16 ... a4 is probably a little prema­ This is a typical deployment, al­
ture, as White is not badly placed to though Black must mind the long di­
fight on the kingside: for example, 17 agonal that White has just occupied.
ii.e3 l:tc7 18 l:tc1 'i¥e8 19 'if d3 l:tb7 20 f4 Others:
and White had the initiative in a) 12 ... 'ifb8 is the thematic move,
M.Gurevich-Y.Vovk, Cappelle la Grande but White is better developed here and
2010. after 13 b4 axb4 14 axb4 h!.xa1 1s 'if xal
b) 16 ... axb4 is not so bad. After 17 cxds (Bologan recommends Black try to
axb4 l:ta1 18 ii.d3 ll:ihs 19 g3 l:ta7 20 draw by sacrificing a pawn starting f6 21 ii.e3 hi.ac7 22 f3 fs Black had with 1s ... bs!?) 16 cxds hi.c8 17 Vas
counterplay in D.Sharavdorj-M.Al Mo­ White has maintained control of the
diahki, Yangon 1999. queenside and enjoys a slight edge.
17 ii.e3 f5 18 f3 ll:if6 19 'ifd2 a4 b) 12 ... 'ifb6 13 dxc6 (13 l:tcl cs is
Now Black decides to close off the solid) 13 ... bxc6 (Black could consider
queenside because he will have good recapturing with a piece) 14 ll:ia4 'ifc7
play on the other side of the board. 1s cs ds 16 ll:ib6 l:tad8 17 ii.c3 with
20 h!.fcl h!.c7 21 'if d3 htb7 22 ll:id1 f4 23 complications in V.Malakhov­
ii.f2 gs V.Zvjaginsev, Poikovsky 2004. Here

A tt a c k i n g C h e s s : The King 's In dian, Vo l u m e 1

Black could consider 11 ... ii.e6 or 17 ... a4. White had the initiative in
c) 12 ... cs is common, but after 13 R.Sherbakov-V.Syrtlanov, Koszalin
ll:ibs ll:ie8 14 b4 axb4 lS axb4 l:txal 16 1996) 14 ll:ic4 cxds 1s exds e4! (after
'ii'xal ii.h6 (16 ... b6 17 bxcs bxcS 18 ii.c3 1s ... ll:ifs 16 ll:ixes ll:id4 17 1'.c4 ii.ts 18
gives White a pleasant edge), 17 ll:ixd6 ! ll:ie2 ii.c2 19 'ifel 1'.g 7 20 ll:ixd4 cxd4 21
is a strong sacrifice, M.Ulibin-A.Galkin, ii.xd4 ll:ixds 22 ll:ixf7! White was doing
Russian Championship, Elista 1996. well in R.Sherbakov-A. Korotylev, Par­
13 'it>hl dubice 1996) 16 ll:ia4 (after 16 ll:id6
White takes a moment to make a Black played 16 ... ii.g7 in F.Atakis­
prophylactic move and again does not T.Remmel, correspondence 2003, but
disclose his intentions. This is a patient 16 ... ii.fs, with the idea of 17 ll:ixb7?
approach, but many other moves have 'ii' b6, looks even better) 16 ... ll:iexds 17
been tried too. ll:ixcs ii.c6 18 'it'd4 ll:if4 19 'it'xd8 ll:ixe2+
a) 13 dxc6 ll:ixc6 (this is fine, but 20 �hl l:tfxd8 21 ii.xf6 .rf.e8! ? 22 ll:id6
both 13 ... ii.xc6 and 13 ... bxc6 are possi­ ii.f8 23 ll:ixe8 ii.xcs gave Black compen­
ble too) 14 ll:ibs ii.e6 1s ii.d3 ll:id7 16 sation for the exchange in R.Jansen­
ll:ib1 ll:ics 17 ll:i1c3 ll:ia7 18 ll:ixa7 l:txa7 F.Nijboer, Dutch Championship, Rot­
19 ii.c2 h!.a6 20 'ii'e2 h!.c6 21 l:tadl 'ii'b6 terdam 1999, and 21 ... l:tds 22 ll:ixe4
22 ll:ibs a4! 23 b4 ll:ib3 gave Black good l:tbs with a strong initiative is a good
counterplay against the c4-pawn in alternative.
Shen Yang-L.Van Wely, Moscow 2009. 13 cs

b) 13 ll:ic8!? (again we see this

manoeuvre) 14 dxc6 ii.xc6 lS ii.d3 ll:ib6
16 b4 axb4 17 axb4 ll:ihs (after 17 ... 1'.d7
18 ll:ib3 ll:ia4 19 ll:ixa4 .Jha4 20 ii.c3
bS!? 21 'ii' e2 - or 21 cxbs 'ii' b6 -
21 ...bxc4 22 ii.xc4 ii.d7 23 hi.fdl 'ii'b6
Black was also doing well in Li Wenli­
ang-F.Nijboer, Groningen 2003) 18 g3
ii.d7 19 'ii'e2 l:tc8 20 ll:ib3 ll:ia4 21 ll:ixa4
ii.xa4 22 ii.xcl 23 l:tfxcl ii.xb3! 24
Itxb3 ll:ig 7 and with the knight bound
for e6, Black had good play in L.Ftacnik­ White's slow play finally convinces
V.Topalov, Polanica Zdroj 199S. Black to acquiesce to the blocking of
c) 13 cs!? is a sharp try: 13 ...dxcs the queenside.
(better than 13 ... cxds?! 14 cxd6 ll:ic8 l S 14 ii.d3
ll:ixds ll:ixds 16 exds f6 17 a4 ll:ixd6 1 8 After 14 'it'c2 ll:ie8 1s ll:ibs fs 16 ii.d3
ii.a3 ii.ts 19 ll:ic4 ll:ixc4 20 ii.xc4 when fxe4 the game was agreed drawn in

Th e M a r def Plata Va ria tio n : 9 lLid2

K.Sakaev-F.Amonatov, Dagomys 2009. nate the d4-knight, but this allows

If we continue with 17 ll:ixe4 ii.xb5 18 Black to contest White's domination of
cxb5 both 18 ...ll:if5 and 18...ll:if6 look the e4-square.
fine for Black.
White can also swing back his
knight with 14 ll:if3, but 14...ll:ih5 (more
active than 14 ... ll:ie8 15 ii.xcl 16
'it'xcl ll:ic8 17 ll:ig1 ll:ig7 18 f4! with the
initiative in A.Beliavsky-M.Erdogdu,
European Championship, Budva 2009)
15 g3 f5 gave Black good play in Zhao
Xue-Wenjun Ju, Jiangsu Wuxi 2008.
14...ll:ihs 15 g3 1'.h3 16 hi.el fs 11 exfs
ll:ixf5 18 ll:ide4
Of course not 18 g4? ii.xd2. 19 ll:ibs ll:ixbs 20 cxbs ll:if6
18 ...ll:id4 By now Black already had a slight
Black has active pieces and no struc­ initiative in V.Kramnik-L.Van Wely, Nice
tural problems. White decides to elimi- (blindfold) 2008.

Ch,a,pte'(c 5
Tfte Mar· def Plata Variation

1 d4 ll:if6 2 c4 g6 3 ll:ic3 1'.g7 4 e4 d6 5 will devote our attention. Black takes

l2Jf3 0-0 6 ii.e2 es 7 0-0 ll:ic6 8 ds ll:ie7 9 advantage of the possibility to jump to
b4 a more active square than d7 or e8,
eyes the f4-square and prepares .. .fs.

The Bayonet. White had tremen­

dous success in the second half of the It is easy to forget that the critical
1990s with this move which forced 10 hi.el of our next chapter is a rela­
Black to really shore up his defences. tively new move. Even though it only
White's king is usually (but not al­ came to prominence as late as 1995, it
ways!) pretty safe and despite the rela­ has been played more than all the
tively closed nature of the position, other moves combined and by quite a
there is often a lot of play in the centre. wide margin. Nevertheless, White has
9 ll:ihs
... other continuations on move 10 that
This is the challenging main line, Black cannot ignore and it is these lines
which will be the variation to which we thatare examined in this chapter.

Th e M a r def Plata Vari a t i o n : 9 b4 tl'i h s with o u t 1 0 hi. e l

B.Gelfand, Dortmund 1997. As the

A: 10 �2 main line is fine for Black, we will focus
8.: 10 tlt'c2 our attention there.
C: 10 C5 12 a4 gs
0; 10 g3 This is commonest and quite sound,
but Black can already consider experi­
A) 10 tl'id2 menting:
a) 12 ...�h8!? 13 tl'ib3 g s 14 exfs
tl'ixfs 1s g3 ll'ih3+ 16 �g2 tl'ih6!? 17
1'.e4 ll'ig4 18 h!.a2 'ife8 19 f3 ll'if6 20 'ifhs was unclear in G.Sosonko­
F.Nijboer, Dutch Championship 1996.
b) 12 ...ii.d7 13 cs (13 ii.a3 gs 14 exfs
tl'ixfs 1s g3?! ll'ih3+ 16 �g2 'ifc8! is
good for Black) 13 14 exfs tl'ixfs 1s
g3 ll'ih3+ 16 �g2 'ifc8 17 1'.e4 g4 was
also messy in D.Zagorskis-G.Beckhuis,
Muenster 1994.
White forces Black's knight to f4. 13 exf5 ll'ixfs 14 g3
This old move has been considered in­
ferior since the well-known game Pet­
rosian-Gligoric, Zagreb 1970, but it is
likely that this attractive game does
not represent best play for either col­
our. Nowadays this move is very rare
and this is understandable because
Black should not really experience any
problems.'if4 11 ..if3
White can also play 11 a4 fS 12 ii.f3,
transposing to the main line. This 14... ll'ih3+
avoids the possibility given in the note This looks sounder than 14 ... ll'id4!?
to Black's 11th move, but gives Black lS gxf4 tl'ixf3+ (1s ...exf4!? has been
the option of playing 11 ... tl'ixe2+ 12 suggested as a better move order, be­
'ifxe2 fs. cause 16 tl'ide4 tl'ixf3+ 17 'ifxf3 g4 leads
11...fs to variation 'a', below, but White could
Black can also play 11'id3 12 ii.a3 try 16 l:ta3 ! ?) and here:
as 13 bxas hi.xas, as in V.Anand- a) 16 'ifxf3 g4 17 't!fh1? exf4 18 ii.b2

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

ii.ts 19 l:tfel f3 20 ll:ide4 'ifh4 21 h3 a) 16 1'.g4? is bad because 16 ...ll:ixf2! gave Black a strong attack in does work now. After 17 �xf2 ll:id4+ 18
T.Petrosian-5.Gligoric, Zagreb 1970. ii.f3 g4 Black is winning.
b) However, 16 ll:ixf3 ! gives Black b) 16 ll:ide4 is natural, but White's
problems: 16 ... e4? (this loses, but light-squared bishop becomes a target
16 ... exf4 17 ii.b2 g4 18 �hl! gxf3 19 after 16 ... l2Jd4 when 17 ii.hs ? ll:if4+ 18
l:tg l also gives Black big problems) 17 gxf4 'ifh3+ 19 �hl g4 gives Black a
ll:ixg s ! ii.xc3 18 l:ta3 1'.g7 19 htg3 and, winning position.
facing an overwhelming attack, Black c) 16 ll:ib3 ll:id4 (16 ... as!?) 17 ll:ixd4
resigned in A.Martin-R.Britton, British exd4 18 ll:ibs c6 19 ll:ia3 (instead 19
Championship, Eastbourne 1991. ll:ixd4 l:txf3 20 �xf3 'if g4+ 21 �e3 looks
Black could, however, throw in the crazy, but 21...cS 22 bxcs dxcs 23 ii.b2
move 14 ... as!? to disrupt White's is actually not so clear) 19 ... h!.xf3 (sim­
queenside. In fact, I think this move pler is 19 with an excellent posi­
would improve Black's cause in almost tion) 20 'ifxf3 g4 with compensation in
any position that follows. R.Keene-L.Kavalek, Teeside 197S.
15 '>t>g2 'it'd7!
Instead 1 S...g4?! 16 ii.xg4 ll:ixf2
backfired after 17 'it>xf2 ll:ih6+ 18 ii.f3
ll:ig4+ 19 '>t>gl ll:ie3 20 'it'e2 ll:ixfl 21
ll:ide4 1'.h3 22 'i¥d7 23 ii.g2 with a
clear advantage for White in J.Parker­
J.Gallagher, Hastings 1991/92, but here
too 1s ... as!? is possible.

Yet again 16 ... as looks good, while
16 ... 'iff7 is also interesting. Instead
16 ...ll:ixf2 17 �xf2 ll:ixg3+ 18 '>t>gl ll:ixfl
19 ll:ixfl was unclear in D.Ruzele­
D.Lapienis, Lithuanian Championship
1993, but I would rather not trade
away Black's active minor pieces, and
16 1'.e4 16 ...ll:id4 17 f3 Itf6 (of course 17 ... as! is
Other moves are likely to bring again a likely improvement) 18 ll:ib3
White trouble: l:th6 19 l:ta2 ll:ixb3 20 'i¥xb3 'iff7 21 cs

Th e M a r def Plata Va riatio n : 9 b4 tl'i h s with o u t 1 0 l:t e1

was a little better for White in tects the c3-knight in anticipation of

V.Grabliauskas-R.Zukauskas, Vilnius ...ll'if4; ilxf4 exf4, after which he will try
199S. to overrun Black in the centre. One
17 ll'ib3'i¥e7 18 'ifd3 hS 19 cs downside to this move is that with his
queen off the d-file, White will lack
control of dS and perhaps d6, and this
could give Black some additional possi­
bilities to undermine the centre.
A similar idea for White is 10 'ifb3
ll'if4 (Black can also consider 10.. .fs! ? or
10 ... h6) 11 ii.xf4 exf4 12 l:tadl h6 13 cs
gs 14 es! (this is basically forced, as
Black would otherwise play 14'ig6
with a great position) 14... dxes lS d6
cxd6 (after lS'ifs both 16 dxc7 'ifxe7
19...ll'if4+ 17 tl'ids and 16 h 3 ! ? cxd6 17 ll'ie4 ds 18
This is a bit speculative. Black could l:txds 'ife7 19 .�c4, as in S.Skembris­
consider 19 ... as or 19 ... tl'igs with coun­ A.David, Cannes 1991, give White com­
terplay. pensation for the pawn) 16 l:txd6 and
20 gxf4 exf4 21 f3 here both 16 ...'ifC7 and 16 ...'i¥e8 have
Panczyk and llczuk suggest 21 'it>hl been tried with unclear play.
ii.xc3 22 i.xfs i.xal 23 ii.xc8 l:taxc8 24
tl'ixal h!.ce8 2S 1'.xf4 'i¥e4+ 26 'ifxe4
h!.xe4 27 ii.g3 when White may be a
little bit better in the endgame.
21... 22 hi.a2 'i¥g7 23 tl'id1
Instead Smirin gives 23 ii.xfs ii.xfs
24 ll'ie4 l:tae8 2S �hl 'i¥g6 with com­
pensation in Info rmant 69.
23 ... g3 24 htgl gxh2+ 25 �xh2 ll'ig3
Black had some compensation for
the piece and managed to overwhelm
his opponent in J.Manion-1.Smirin, Las 10.. .fs
Vegas 1997. I like this straightforward move. In­
stead'if4 11 1'.xf4 exf4 12 h!.adl
B) 10 'i¥c2 seems to play into White's hands: for
Epishin's move looks funny but it example, 12 ... as 13 bs h6 14 ll'id4 gs 1s
should be taken seriously. Black pro- ii.h s 16 h3 tl'ig6 17 tl'if3 'i¥e7 18

A t ta c k i ng C h e s s : Th e King 's Indian, Vo l u m e 1

b6!? cxb6 19 ll:ia4 hta6 20 'it'b3 and tending ...ll:ifs looks like a good answer,
White had the initiative in A.Zhigalko­ while Black could consider throwing in
V.Kovalev, Minsk 2010. 12 ... as!? too.
A decent alternative, however, is 12 ...exf4 13 h!.ael
10 ... as 11 bxas hi.xas 12 hi.el cs 13 a4
h!.a6!? with a position similar to the line
10 hi.el as 11 bxas hi.xas 12 ll:id2 ll:if4
13 ii.fl cs 14 a4 h!.a6, but in this case
Black will not have to waste as much
time playing the typical retreat ... lLif4-
hS-f6. As that line is not part of our
repertoire after 10 l:tel, I will stick with
the text move, which I think gives Black
good chances.
11 ll:igs
After 11 cs I would propose White is loading up in the centre,
11 ...ll:if6!? 12 ll:igs fxe4 (this is logical but Bl ack has a good tactical answer.
because playing 'ii'c2 and cs has left 13 ...ll:ic6! 14 dxc6
the dS-pawn needing White's atten­ After 14 ll:ie6 Black has a couple of
tion), with the idea of ... ll:ifs when the good responses:
queen on c2 is a target.. a) 14...ii.xe6 lS dxe6 ii.xc3! 16 'ii' x c3
11...ll:if4 fxe4 17 1'.g4 hs 18 1'.h3 (or 18 ii.d1
Also good is 11 ... ll:if6 when 12 ii.f3 'ii'f6) 18 with good play.
(possible too is 12 f3, but I cannot see b) 14 ... ll:id4 lS ll:ixd4 ii.xd4 16 ii.f3
how White's queen on c2 can be an 1 7 cs as 18 'it'b3?! (18 a3 is more
improvement on the normal hi.el in solid) 18 ... axb4 19 'ii'xb4 was V.Epishin­
this position) 12 ... h6 (again 12 .. .fxe4 V.Menoni, Bratto 1999, and here
with the idea ... lLifS-d4 is also a reason­ 19 ... dxcs 20 'ii'xcs 'ii'd6 looks good for
able idea) 13 ll:ie6 ii.xe6 14 dxe6 was Black.
played in A.Yermolinsky-D.Sharavdorj, 14...'ii' xgs 15 ii.f3
Berkley 200S, and here Panczyk and After lS cxb7?! ii.xb7 White has
llczuk suggest 14...'it'c8 ! lS lLids ll:ifxds. problems on the long diagonal.
This is rarely this a good idea, but it 1s ... bxc6 16 bs 1'.d7
works here: 16 cxds c6 and Black has A decent alternative is 16 .. .fxe4 17
good counterplay. ll:ixe4 (17 ii.xe4 f3!) 17...'it'fs 18 bxc6
12 ii.xf4 h!.b8 with good play for Black.
This is always played, but 12 ii.f3 17 exfs 'ii'xfs 18 1'.e4 'ifgs 19 bxc6 1'.h3
must also be considered: 12 .. .fxe4 in- 20 'ii'd 3 ii.xg2! 21 ii.xg2 ii.xc3 22 h4

Th e M a r def Plata Varia tio n: 9 b4 tl'i h s with o u t 1 0 hi. e l

'it'f6 23 hi.e6 'ifg7 24 ii.ds This is by far the most common

Here the game V.Epishin-F.Nijboer, move. White logically breaks the pin
Apeldoorn (rapid) 2001, was agreed along the long diagonal, but he can
drawn, although after the forced also protect the loose c3-knight with
24...�h8 I prefer Black. his queen:
a) 12 'it'd2 h6 (oth erwise, 12 ... 1'.g4
C) 10 cs 13 ll'id4 is annoying, but 12 ... as!? is
worth considerin g) 13 .rf.adl gs 14 es
g4!? lS exd6 cxd6 16 ll'iel and here
Black can play 16 ... tl'ig6 or 16 ...dxcs!? 17
d6 tl'ig6 18 bxcs (not 18 d7 ii.xd7 when
the c3-knight is loose) 18 ... ii.d7 with an
unclear position.
b) 12 'tlfb3 1'.g4 (12 ... h6 13 h!.adl was
considered via the move order 10 'ii'b3
in Line B) 13 hi.adl ii.xf3 14 ii.xf3 gs 1s
ii.h s (or 1s 1'.g4 tl'ig6 16 tl'ie2 lLle s)
1 S'ig6 16 ii.xg6 hxg6 was pretty level
This move is the most direct. Al­ in A.Schneider-E.Gufeld, Helsinki 1992.
though it is not bad, this line is not 12... as
seen very much nowadays. This move scores well and was'if4 played by Kasparov, which is quite a
Black can also play 10 .. .fs, but I pre­ good endorsement. Again 12 ... 1'.g4 is
fer the text move, which is considered met by 13 ll'id4, but Black has a couple
the main line. of other lines worth considering:
11 ii.xf4 exf4 a) 12 .. .fs is worth looking at, espe­
cially to compare it to 12 ... as 13 a3
axb4 14 axb4 fs below. After 13 es
dxes 14 d6 cxd6 1s cxd6 tl'ic6 16 'ii' ds+
�h8 17 bs ll'id4 18 tl'ixes 'it'e8! 19 'it'xd4
'ii' x es 20 'ii'xes ii.xes 21 hi.fd1 hi.d8 22
hi.ds 1'.g7 23 ii.f3 ii.e6 24 hi.d3 the posi­
tion was unclear in J.Vilela-L.Perez, Villa
Clara 1998; Black has the bishop-pair,
but White has a passed d-pawn and
some queenside pressure.
b) 12 ... h6 is the main line. White has
12 some choice here:

A ttacking C h e s s : Th e Kin g 's I n dian, Vo l u m e 1

bl) 13 ll:id2 g s 14 ll:ic4 ll:ig6 (14 ... a6 sibility: 2 o 'ifd2 (20 l:tfdl 1'.d7) 2 o...ll:id4
l S ll:ia4! ?) lS 1'.g4 ll:ies 16 ll:ixes ii.xes 21 ll:ixd4 exd4 22 ll:ids f3 ! (better than
17 ii.xc8 h!.xc8 18 ll:ibs and here, rather 22 ... ii.e6 23 ll:ixf4) 23 ii.xf3 (after 23
than 18 ...'it'f6 19 c6!, Black should play gxf3 ii.e6! 24 ll:if4 'it'gs+ wins for Black
18 ... dxcs 19 bxcs c6! with counterplay. and here 24 ll:ic7 Ita2 2s ll:ixe6 Itxd2 26
b2) 13 lLid4 g s 14 ii.hs as lS a3 axb4 16 ll:ixd8 h!.xd8, with the idea of, is
axb4 c6! 17 ll:ide2 would transpose to also much better for him) 23 ...l:txbs 24
A.Beliavsky-V.Spasov, Manila Olympiad 'i¥e2 1'.d7 and Black has an extra pawn.
1992, and here the simplest move is b) 13 ll:ibs could also be played im­
17 ... with equality according to mediately. Then 13 ... axb4 14 cxd6 cxd6
Beliavsky. l S 'it'b3 1'.g4 16 htc7 ll:ic8! is a very typi­
b2) 13 h 3 ! ? g s 14 a4 ll:ig6 l S as l:te8 cal move. The knight heads to a7 to
(both 1S ...'ife7 and 1S ... a6 are possible challenge White's own knight and un­
too) 16 ll:id2 ll:ies 17 cxd6 cxd6 18 ll:ibs dermine the Cl-rook. Now White has:
l:te7 19 'i¥c2 a6 20 ll:id4 was M.Diesen­ bl) 17 ll:ifd4 ii.xe2 (Black could also
L.Day, Lone Pine 1977, and here Black consider 17 ... ii.xd4 18 ii.xg4 ii.cs, with
should play the simple 20 ... ii.d7! when the idea 19 l:txb7 'it'as!) 18 lLixe2 f3 19
the position is fairly level. gxf3 ll:ia7 20 hi.xb7 ll:ixbs 21 Itxbs 'ifgs+
22 �hl hi.a3 23 'it'xb4 l:txa2 gave Black
compensation for the pawn in
F.Schirm-G.Schmid, German League
b2) 17 h!.xb7 Vas 18 1'.c4 and here:

13 cxd6
Other moves are also possible:
a) 13 a3 is the obvious move, but
opening the a-file should favour Black
compared to the lines above. After
13 ... axb4 14 axb4 fS! ? (of course, 14 ... h6 b21) 18...ii.xf3 19 gxf3 ! ll:ib6 20 l:tbl
is also possible) lS es dxes 16 d6 cxd6 ll:ixc4 21 'ifxc4 (Grivas suggests
17 cxd6 ll:ic6 18 'it'ds+ �h8 19 bs the 21 ... h!.fc8 22 'it'xb4 l:tcl+ 23 �g2 'i¥xb4
idea of 19 ... Itas!? becomes a new pos- 24 Itxb4 Itxa2 2S htb8+ ii.f8 26 ll:id4

Th e M a r def Plata Variati o n : 9 b4 tl'i h s with o u t 1 0 hi. e 1

Itdl, but after 2 7 tl'ie6! fxe6 2 8 dxe6 unclear in N.Petre-L.Vajda, Sovata 1999.
�g7 - 28 ... l:ta7 29 l:ta4! is an important 14...1'.g4
point - 29 l:t4b7+ �f6 30 l:txf8+ �xe6 Also possible is 14... axb4 which
31 l:tff7 �es 32 l:tbe7+ �d4 33 l:tf6 leads us to note 'b' to White's 13th,
White can still try to press) 22 'it'xb4 above.
'it'xb4 23 l:txb4 l:txa2 24 J:ic4 was a little 15 l'.!c7 axb4 16 'ii' d2
better for White in E.Grivas-J.Cooper,
Novi Sad Olympiad 1990.
b22)'ib6! is very strong: 19 l:ic7
tl'ixc4 20 'ii'xc4 l:tfc8! 21 l:tc6 l:txc6 22
'ii'xc6 (22 dxc6 ii.e6!) 22 ...'ii' xa2 (or
22 ...ii.xf3 23 gxf3 'ii'xa2 24 ll'ixd6 b3) 23
tl'ixd6 was E.Grivas-J.Murey, Tel Aviv
1991. Here Black shouldn't play
23 ...ii.xf3 24 gxf3 b3, transposing to the
previous note, but rath er the immedi­
ate 23 ... b3! with a clear advantage.
13 ... cxd6 16...ii.xf3
Th is was Kasparov's choice, but I
think that Black may be better off with
the older 16 ... tl'ic8!? and then:
a) 17 l:tfcl ii.xf3 18 ii.xf3 ii.c3 !? (I
think Black can improve with 18'ia7
19 'it'xb4 tl'ixbs 20 'it'xbs l:txa2 21 'ii'xb7
1'.d4 when he is obviously better - f2 is
weaker than f7 and Black has much the
better bishop) 19 'it'e2 (better is 19
'ifxf4 tl'ia7 20 h!.7xc3 tl'ixbs - after
20 ...bxc3 21 tl'ixd6 White has compen­
14 tl'ibs sation - 21 l:tb3 when White is at least
After 14 'ii'd 2 axb4 1s tl'ibs Black equal) 19 ...ll'ia7 20 l:txb7 tl'ixbs 21 l:txbs
could play 1S ... 1'.g4 16 h!.c7, which is was B.Malich-L.Vogt, East German
our main line below, or lS ...fS !? 16 ii.d3 Championship, Erfurt 1973, and here
fxe4 17 ii.xe4 h!.a4 with counterplay. 21 ... l:ta3 would be better for Black.
In stead 14 bS limits White's possi­ b) 17 ll'ifd4 ii.xd4!? (17 ...ii.xe2 18
bilities on the queenside and 14... a4!? tl'ixe2 was agreed drawn in
1s 'it'd3 (1s tl'ixa4 'it'as) 1s...ii.d7 16 K.Grigorian-E.Geller, USSR Champion­
tl'id2 fs 17 hi.fe 1 tl'ic8 18 ii.f3 tl'i b6 was ship, Leningrad 1971) 18 ii.xg4 ii.cs

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

(with the idea of ... ll'la7) 19 es (19 .i::!.xb7 'ii'xb7 21 'ii'x as .i::!.a8 22 'ii'd2 when he
�as is better for Black) 19 ... ll)a7 20 has an extra pawn, although it will be
exd6 tllxbs 21 .i::!.x cs tllxd6 22 'ii'xf4 difficult to convert.
.i::txa2 23 'ii'xb4 'ii'f6 with equality. 18 'ii'as 19 ll'ld4 'ii'x a2 20 'ii'xa2 .i::tx a2

17 �xf3 �es 21 .i::txb4

Again 17 ...tllc8! ? looks good: 18
'ii'xb4 (18 .i::!.x b7 'ii'a s) 18 ...ll'la7 19 .i::!. x b7
tllxbs 20 'ii'xbs .i::tx a2 and if anyone is
better it's Black.

21 .i::!.fa8

Black has also tried 21 ... 3'.xd4 22

.i::!.xd4 g s 23 .l::!.b4 (White should really
prevent Black's next by playing 23 �hS
18 .i::!.xb7?! with equality) 23 ...tll g 6 24 .i.g4 tll e s 2s
This may be good enough to main­ h3 .i::!.fa8 with an edge in H.Herndl­
tain equality, but White will have to S.Kindermann, Austrian League 1996.
work for it. Instead after 18 .i::!.fcl Galla­ 22 �g4
gher gives 18 ....i::t a s! and then: Gallagher suggests that White
a) 19 'ii'xb4 .i::txa2 20 .i::!.xb7 .i::!.b2 21 should at least activate his bishop with
'ii'a3 'ii'a 8! 22 .i::!.x e7 (or 22 'ii'x a8 .i::txa8 22 tllc6 tllxc6 23 dxc6 .i::tc2 24 .i.d1 .i::!.xc6
23 tllxd6 .i::tc2! ) 22 ....i::t x bs when Black is 2S �b3, although he will clearly be
better. grovelling here as well.
b) 19 .i::!.xb7 �a8 20 �xb4 (not 20 22 .i.xd4 23 I!xd4 gs 24 h4?!

.l::!.ce7 .i::txa2 winning, while 20 .i::!. x e7 After 24 �hS .l::!.b8 Black can try to
.i::!.xbs also favours Black) 20...'ii'x b7 21 improve his position with ... '>t>g7 (Gal­
'ii'xas .i::!.a 8 22 'ii'b4 ( 2 2 .l::!.e7 'ii'b 8!) lagher), but White should hold here.
22 .. 1Ixa2, again with a slight advan­ 24 gxh4 25 'it>h2 tll g6 26 '>t>h3 .i::t b2 27

tage for Black. �fs tll es 28 'it>xh4 h6 29 �h3 '>t>g7

White has an improvement on all of Now White's king is in some trouble
this, however, as he can play the clever and Black went on to win in G.Kamsky­
18 .i::tb 1 ! ? .i::t a s 19 .i::!.xb7 'ii'a 8 20 'ii'xb4 G.Kasparov, New York (rapid) 1994.

1 00
Th e M a r de/ Plata Var i a t i o n : 9 b4 li'i h 5 with o u t 1 0 I! e 1

D ) 10 g3 ...bxc3 and ...'ii' c s+) 16 ... h6 17 li'if7+

This favourite of Van Wely is (Nunn gives 17 dxe7 'ii'xe7 18 li'if7+
White's most dan gerous alternative to I!xf7 19 �xf7 'ii'xf7 20 'ii'd8+ 'ii'e8, win­
10 I!el and sometimes the play is simi­ ning, and 17 li'ixc7 hxgs 18 li'ixa8 li'ic6
lar. Rather than discouraging'if4, 19 li'ic7 !i'id4 as good for Black, because
White prevents it altogether. 20 li'ibs? li'ixbs 21 �xbs 'ii'b6+ wins)
10 fs 11 li'igs li'if6 12 f3 f4
... 11 ...I!xf7 18 �xf7 cxd6 19 li'ixd6 �h3
gave Black good compensation for the
exchange in Cu.Hansen-J.Nunn, Wijk
aan Zee 1991.
b) 13 'it>g2 is actuallythe most popu­
lar move.

This is Black's most popular choice

and is favoured by the top players. In­
stead 12 ...c6 is also popular, but White
has done well after 13 �e3. The imme­
diate 12 ... h6 13 li'ie6 �xe6 14 dxe6 c6
lS �e3 also looks good for White, and A good part of the theory of this line
after 12 ...'it>h 8 both 13 �e3 and Van is based on Van Wely's games, but
Wely's 13 'it>g2 score well for White, so more recently he has preferred 13 bS.
we will stick with 12 .. .f4, which looks to Here Black has a wide choice:
take advantage of White's 10th move. bl) 13 ... as 14 bxas (instead 14 �a3
13 bs axb4 lS �xb4 �h6 16 li'ie6? loses to
This has been Van Wely's latest 16 ... .i.xe6 17 dxe6 cs, but 14 bS!?
choice, but we may have seen the end should be considered) 14 ..l!xas lS 'ii'b3
of it. White has a couple of other op­ li'ie8 16 �d2 I!a8 17 cs was a little bet­
tions: ter for White in E.Lobron-B.Gelfand,
a) 13 cs has fallen out of favour be­ Munich 1992.
cause of 13 ... dxcs!? 14 �c4 (14 bxcs h6 b2) 13 ...h6 14 li'ie6 �xe6 lS dxe6 c6
lS li'ie6 �xe6 16 dxe6 'ii'd4+! is Black's (not 1s'ic6?! 16 li'ids) 16 bS 'ii' c 7 17
idea) 14 ... cxb4 lS d6+ 'it>h8 16 li'ibs (af­ bxc6 bxc6 18 �a3 I!fd8 19 'ii'a4 'ii'c8 20
ter 16 dxe7 'ii'x e7 Black threaten s both I!abl 'it>h 8 21 .i::!.fdl 'ii'xe6 (21 ... as? has

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

been suggested, but 22 .i::!.xd6 just wins) ll:lxe6 2 S as ll:lc8 White has compensa­
22 'ii'a 6 hS 23 .l::!.b 7 gave White good tion) 22 'ii'b 3 'ii'e 8 23 a4 a6 24 ll:lc3
compensation for the pawn in L.Van ll:lxe6 2s �xe6 .i::!.xe6 26 ll:lds �d8 27
Wely-Ye Jiangchuan, Biel lnterzonal �b2 with compensation for the pawn
1993, although Black has some coun­ in L.Van Wely-A.Zapata, Matanzas
terplay on the kingside. 199S.
b3) 13 ... 'it>h8 14 cs h6 lS cxd6 (after bS) 13 ...c6 increases the tension and
lS ll:le6 �xe6 16 dxe6 dS 17 exds has scored very well.
ll:lexds 18 ll:lxds ll:lxds 19 �c4 Black
has 19 ... c6 with the idea of ....i::!.e 8)
1s ...�xd6 (now after 1s ...cxd6 16 ll:le6
�xe6 17 dxe6 Black cannot support the
dS-square with ... c6) 16 ll:lbs �b6 17 a4
ll:lfxds 18 exds hxgs 19 as 'ii'f6 20
ll:lxe7 .l::!.b8 21 g4 and now 21 ... �d7? 22
bS .i::!.fc8 23 d6 left White much better
in L.Van Wely-J.Nunn, Wijk aan Zee
1992, but both 21 ...�d6 21 ...e4 improve
and look satisfactory for Black.
b4) 13 ... ll:lhs!? and here: White has:
b41) 14 cs fxg3 lS hxg3 ll:lf4+ 16 bSl) 14 bS cs! lS 'ii'd3 ll:le8 16 ll:le6
gxf4 exf4 17 �el (after 17 lLlxh7 Black �xe6 17 dxe6 was B.Finegold­
can play the simple 17 ... 'it>xh7 18 .l:!.hl+ G.Michelakis, Groningen 1993. Here
'it>g8 19 �b2 gs OT 17 ...�XC3 18 ll:lxf8 17 ... 'ii'c8 18 ll:lds 'ii'xe6 should be
�xf8 19 .l::!.b l �h3+! 20 '>t>gl �xfl 21 somewhat better for Black
�xfl 'ii'f6 22 �h3 '>t>g7 with a good po­ bS2) 14 cs looks logical enough, but
sition) 17 ...ll:lfs 18 �xf4 (18 lLlh3 ! ?) remains untried. Gallagher gives
18 ... �XC3 19 �XC3 ll:lh4+ 20 '>t>f2 (not 14 ... h6 (14 ... cxds! ? 1s exds ll:lfs is an­
20 '>t>g3 .i::!.xf4 21 'it>xf4 ll:lg2+ 22 '>t>g3 other idea) lS �c4 (or lS ll:le6 �xe6 16
�xgs+) 20....i::txf4 21 ll:le6 �xe6 22 dxe6 dxe6 dS!) 1s ... hxgs 16 cxd6 'it>h7 17
'ii'e7 23 cxd6 cxd6 24 .i::ta cl .i::taf8 with a dxe7 �xe7 18 bS (both 18 g4 'ii'xb4 and
messy position, L.Van Wely-0.Cvitan, 18 d6 'ii'd7 should be at least okay for
Moscow Olympiad 1994. Black) 18 ... g4 with counterplay.
b42) 14 g4 �f6! lS ll:le6 �xe6 16 bS3) 14 'ii'b3 h6 lS ll:le6 �xe6 16
dxe6 ll:lg7 17 cs 'it>h8 (17 ...ll:lc6!?) 18 dxe6 �c8 17 .i::!.d1 .l:!.d8 (not 17...'ii'xe6?
cxd6 cxd6 19 ll:lbs ll:lc8 (19 ...ll:lxe6!?) 20 18 .i::!.xd6!), with a final divide:
�c4 .l:!.e8 21 �dS .l::!.e 7 (also after bs31) 18 ll:lds ! ? cxds 19 cxds gs 20
21 ... �e7 22 �b2 ll:lb6 23 a4 a6 24 ll:lc3 �d2 .i::!.f 8 21 g4 hS 22 h3 ll:lg6 23 .i::!.dcl

1 02
Th e M a r de/ Plata Varia t i o n : 9 b4 lLJ h 5 with o u t 1 0 .i::!. e 1

� 8 2 4 ..i.el hxg4 2S hxg4 lLJe8 2 6 a4 19 �xf8 �xf8 with some compensa­

�f6 27 as �d8 was unclear in L.Van tion) 16 lLJe6 �xe6 17 dxe6 'ii'c 8 18
Wely-A.Fedorov, Batumi 1999. lL:ids 'ii'xe6 19 g4 lLJf4 20 lLJxc7 'ii'c 8 21
bs32) 18 cs ds 19 exds lL:ifxds 20 lLJxa8 'ii' cs+ 22 'it>h2 I!xa8 23 'ii'f2 was
lLJe4 'ii'xe6 (instead 20 ...lL:ifs was sug­ better for White in G.Prakash­
gested by Mikhalevski, but after 21 e7!? K.Sasikiran, New Delhi 2001.
lLJfxe7 22 lLJd6 'ii'd7 23 bS White has b) lS 'it>f2! lLJf4? (this fails, but after
the initiative) 21 g4 'it>h8 (probably too 1S ... ..i.f6 16 lLJe6 �xe6 17 dxe6 both
slow; Mikhalevski suggests 21 ... b6, 17 ... lLJg7 18 cs and 17 ... 'ii'c8 18 cs 'ii'xe6
while 21...bS!? is also possible) 22 '>t>hl 19 cxd6 cxd6 20 �3 leave White with
lLJg8 (again 22 ... b6 looks better) 23 �b2 the initiative) 16 gxf4 exf4 17 'ii'd3 h6
lLJgf6 24 gs lLJxe4 2s fxe4 lLJC7 26 gxh6 18 lLJe6 ..i.xe6 19 dxe6 lLJc6 20 .l:thl lL:ies
�f6 27 �c4 'ii'g4 28 .i::!. d 6! With an ini­ 21 'ii'd2 'ii' gs 22 .i::!.h 3 'ii'f6 23 lL:ids �xe6
tiative in L.Van Wely-R.Pruijssers, Dutch 24 'it>g2 �f7 2S 'ii'xf4 'ii'd7 26 'ii'g 3 and
League 2006. 1-0 was L.Van Wely-S.Dyachkov, Rus­
We now return to 13 bS. sian Team Championship 2008.
14 lLJe6 �xe6 15 dxe6 fxg3
Black only plays this because he has
a concrete idea in mind.
16 hxg3 �cS 17 lLJds

At first this move seems illogical

from a positional standpoint, but
White's play is based on his light­
square control.
13 ... h6 17 ...'ii'xe6 18 lLJxc1 'ii'h 3
After 13 .. .fxg3 14 hxg3 lL:ih s (it is This is the point of Black's play.
probably better to head back to the 19 l!f2
main line with 14 ...h6 lS lLJe6 �xe6 16 After 19 lLJxa8 'ii'xg3+ 20 '>t>h l Black
dxe6 'ii'c 8) White has: can, of course, take the perpetual
a) lS 'ii'e 1 h6 (better is 1s ...�f6 16 check, but he could also try 20 ... lLJhS ! ?
lLJe6 �xe6 17 dxe6 lLJg7 18 �h6 lLJxe6 21 'ii' el .i::tx a8 22 .i::tf2 'ii'h4+ 23 'it>g2

1 03
A ttacking C h e s s : Th e King 's I n dian, V o l u m e 1

ll:lf4+ 24 �xf4 exf4 2S .i::!.d1. Then 24 ... h s (24 ... ds has also been consid­
2s ...�d4 26 .i::txd4 'ii'g 3+ 27 'it>h1 'ii'h4+ ered), although this is playable too: 2S
is still a draw, but Black could also go in ll:lxg7 'it>xg7 26 'ii'b3 'ii'g 2 27 'ii'e 3 ll:lg8!
for either 2s ... �es or 2s ....i::!.c 8! ? with 28 cs dxcs 29 � b2 and now, instead of
the idea of ... .i::tc s. 29 ... 'ii'h 2?! 30 .i::!.e l ll:lf6 31 '>t>dl b6 32
�c3 'it>h7 33 'ii'g s when White was
much better in L.Van Wely-T.Radjabov,
Dresden Olympiad 2008, Black should
have played 29 ... ll:lf6! 30 .i::!.el (not 30
�xes? .i::!.x e2+ 31 'ii'xe2 'ii' g s+) 30 ... ll:lg4
when he has counterplay with his ac­
tive pieces and dangerous h-pawn.

19 ll:lxe4!

Correct, whereas after 19 ... .i::tac8 20

.i::th 2 'ii'xg3+ 21 .i::!.g 2 'ii' h3 22 'ii'xd6 .l::!.f7
23 cs �f8 24 ll:le6 White was winning
in L.Van Wely-M.Golubev, Romanian
Team Championship 2000.
20 fxe4
This leads to a very delicate posi­ 2 s ll:lxd6 .i.f6! 26 cs
tion. Instead 20 .i::th 2 'ii'd7 21 ll:lxa8 After 26 '>t>el 'ii'g 3! 26 'ii'e l .i::!.h 2 27
ll:lxg3 22 �xh6 �xh6 23 .i::!.xh6 '>t>g7 24 cs Black has 27...ll:lc8! 28 '>t>dl (or 28
.i::th 2 ll:lefS! gave Black compensation ll:lxc8 'ii'xe4) 28...'ii'g 2 29 'ii d2 ll:lxd6 30
for the exchange in L.Van Wely­ cxd6 'ii'h l+ 31 'ii'e l 'ii'xe4 when White
J.Degraeve, Mondariz Balneario Zonal is in trouble.
2000. 26...ll:lds! 27 exds e4! 28 'it>e1
20 ....i::txf2 21 'it>xf2 .i::!.f8+ 22 '>t>e3 'ii'xg3+ Now 28 ...'ii'g 2 29 'it>d2 �xal was
23 'it>d2 .i::tf2 24 ll:leS good enough to win in L.Van Wely­
After 24 'ii'b3 'ii'g 2 2S 'ii'e 3 h S ! (Gal­ D.Stellwagen, Dutch Championship,
lagher) Black gets his h-pawn moving Amsterdam 2009, but even stronger
and threatens ... �h6. was 28 ... �c3+! with the idea of 29 �d2
24...'ii'f3! 'ii'g 3 30 �xc3 e3 31 'ii'd3 .i::tfl+ 32 'it>xfl
A tremendous improvement on 'ii'f2 mate!

1 04
Chapter 6
The Mar del Plata Variation
9 b4 tbhs 10 l::.e 1

1 d4 ll:lf6 2 c4 g6 3 ll:lc3 �g7 4 e4 d6 5 Truth be told, Kasparov never

ll:lf3 o-o 6 �e2 es 7 o-o ll:lc6 8 ds ll:le7 9 'blamed' this loss to the Bayonet for the
b4 ll:lhs 10 .i::te1 nixing of the opening from his reper­
toire. Since then, however, not only has
Teimour Radjabov resurrected the
King's Indian, but he has had great suc­
cess in this particular 'death knell'
variation from the black side. Kramnik
himself has faced Radjabov twice in the
Classical King's Indian, and both times
he has avoided the Bayonet...
10 fs

The main alternative is 10 ... as 11

bxas .i::txas 12 ll:ld2 ll:lf4 13 �fl cs 14 a4
This move was introduced into .i::ta6 15 .i::ta 3 with a stodgy game ahead.
modem practice by Ivan Sokolov. Soon This has never appealed to me and the
hordes of top GMs, headed by Vladimir fact that nowadays the retreat
Kramnik, started playing this way. The 15 ... ll:lhs, with the idea of a further re­
game Kramnik-Kasparov, Novgorod treat with ...ll:lf6-e8, has become con­
1997, sounded the alarm and Kasparov sidered Black's most reliable method of
quickly stopped playing the King's In­ play furthers my suspicions.
dian. The world took notice and the 11 ll:lgs ll:lf6
King's Indian became a 'dubious' open­ The alternative 11 ...ll:lf4 is not so
ing and virtually vanished from the bad. After 12 �xf4 exf4 13 .i::!. cl Black
upper echelon of the chess world. can choose between 13 ... �f6, 13 .. .fxe4

A ttacking Chess: Th e King 's I n dian, Vo l u m e 1

and 13 ... j.h6!?. The position tends to and this variation tends to lead to forc­
simplify greatly, but Black is rarely in ing play.
any 'danger' of winning the game. 12 c6

Black shuts down the f3-bishop's

diagonal and increases the tension in
the centre. Black may also initiate the
sequence 13 ... h6 14 lL'ie6 j.xe6 15 dxe6
followed by either .. .fxe4 and ... d5 or
... lL'ie8-c7 depending on the circum­
The only other sensible continua­
tion is really 12 .. .fxe4 with the idea of
... lL'if5. After 13 lLicxe4 lL'if5 14 �b2 or
13 lLigxe4 lL'if5 14 .i.g5 lLid4 15 lL'ib5!
After 11...lL'if6, White's e4-pawn is White is somewhat better, although
attacked, so his choice is limited. White Black has a solid position. In general I
has two main lines here and they are do not like this structure when the c­
both quite complicated. pawns are still on the board, because
the advance c4-c5 gains space for
A: 12 .i.f3 White and Black always has to consider
B: 12 f3 the possibility of both cxd6 and c5-c6.
After 12 ...c6 White has several pos-
A) 12 j.f3 sibilities:

Al: 13 'W'b3
A2: 13 b5
A3: 13 .i.b2
A4: 13 .i.e3

Other moves are less dangerous:

a) 13 lL'ie6 looks premature. After
13 ... j.xe6 14 dxe6 fxe4 15 lLixe4 lLixe4
16 j.xe4 d5 17 cxd5 cxd5 18 .i.c2 e4 19
.l::!.bl 'ii'b6 (but not 19 ...'ii'd 6? 20 j.xe4
White increases his piece play on j.e5 21 .i.f3 j.xh2+ 22 'it>h1) Black is
the light squares. The diagonal for the already better.
light-squared bishop can easily be b) 13 .l::!.bl is Bareev's favourite move
opened if White plays lL'ig5-e6 and exf5, in various positions, but here it looks

1 06
Th e M a r de/ Pla ta Va ria tio n : 9 b4 li'i h 5 1 0 I! e 1

irrelevant after 13. . .h6 1 4 li'ie6 �xe6 15 c21) 14 cxd5 h6 15 li'ie6 �xe6 16
dxe6 fxe4 16 li'ixe4 li'ixe4 17 �xe4 d5 dxe6 fxe4 (also possible is 16 .. 1Ic8 17
18 .i.c2 'ii'd6 19 'ii'g4 (E.Bareev­ 'ii'b3 d5, as in A.Shariyazdanov­
T. Radjabov, Sarajevo 2003), when both M.Cebalo, Rabac 2003) 17 .i.xe4 d5 18
19 ...e4 and 19 ... .i::!.f6 look good for Black. b5 li'ixe4 19 li'ixe4 dxe4 and once more
c) 13 .i.a3 is not bad and will gener­ we have transposed to Line A2.
ally transpose to Line A2, although c22) 14 exd5!? e4 15 .i.e2 and then:
there are a couple of independent pos­ c221) 15'ifxd5 16 li'ixd5 .i.xal 17
sibilities: li'ixe7+ 'ii'xe7 18 'ii d5+ �g7 19 .i::!.x al
'ii'xg5 20 .i.b2+ 'it>h6 21 'ii'x d6 gave
White very good compen sation for the
exchange in L.Portisch-P.Acs, Rethym­
non 2003.
C222) 15 ... !i'Jh5 ! ? 16 'ii'd 2 li'if4! 17
'ii'xf4 .i.xc3 18 'ii'h4 h5 19 .i.xh5 was
J.Lautier-T.Moriuchi, Tokyo (simul)
2002, and now 19 ... gxh 5 20 �xh5 .i::!.f6
21 'ii'h 7+ �8 22 'ii'h 8+ li'ig8 23 li'ih7+
�7 24 li'ig5+ is a draw.
c223) 15 ... h6 looks the simplest: 16
cl) 13 ...h6 is the most forcing: 14 li'ie6 .i.xe6 17 dxe6 d5 18 cxd5 li'ifxd5
li'ie6 .i.xe6 15 dxe6 fxe4 16 �xe4 (16 19 li'ixd5 �xd5 is equal according to
li'ixe4 li'ixe4 17 �xe4 d5 18 cxd5 cxd5 19 Portisch.
b5 reaches the main line of 'A2') 16...d5
17 �c2 ! ? (again, both 17 cxd5 cxd5 18 Al) 13 'ii'b3
b5 li'ixe4 19 li'ixe4 dxe4 and 17 b5 li'ixe4
18 li'ixe4 dxe4 19 'ii'xd8 .i::!.fxd8 20 .i.xe7
I!e8 21 .i.c5 cxb5 22 cxb5 bring us to
Line A2) 11 ... 'ii'd6 18 cxd5 cxd5 19 �b3
'it>h7 20 'ii'e2 li'ih5 21 I!ad1 li'if4 22 'ii'f1
and now, rather than 22 ...'ii'xe6 23 b5!
.i::!.ad8 24 �xe7 'ii'xe7 25 li'ixd5 li'ixd5 26
.i.xd5 with an edge for White in J.Chab­
anon-0.Touzane, Marseille 2010, Black
should try 22 ... .i::ta c8! with good play.
c2) 13 ... cxd5 gives White the possi­
bility to deviate, but Black should still This is a tricky move. White creates
be fine: pressure on the a2-g8 diagonal, but

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

now there is no pressure on the d6- well here too) 19 ... .i::tx e7 20 .i::!.x es 'ii'C7
pa wn, a consideration that Black 21 .i::!. e el fxg2 (White's kingside is a
should logically try to take advantage wreck and Black threatens ...lL'ig4) 22 h3
of. .i::!.f8 23 .i::t e 3 lL'ihs gave Black a strong
13 h6
... attack in Y.Pelletier-E.lnarkiev, Istanbul
This is the main move, but 13 ... 'it>h8 2003.
is also possible. b) 16 �a3 is an interesting move
14 lL'ie6 �xe6 15 dxe6 'ii'c 81 order.
Because nothing is attacking the
d6-pawn, Black can attack the e6-pawn
right away. He has a good alternative
in 1s ...fxe4 16 lLixe4 lLixe4 17 .i.xe4
�c8 and now after 18 'ii' h3 dS 19 .i.c2
lL'ifs Black has counterplay, while 18
.i::!.d 1 .i::!.d8 19 bS (after 19 'ii' h 3 dS 20
.i.c2 e4 Black looks to be doing well)
19 ...'ii'xe6 leaves White with some
compensation for the pawn, but
probably no more than that.
After 16 ... 'ii'xe6 17 .i::!.adl .i::!.fd8 18 bS
'it> h7 (note that compared to the main
line, 18...cs?! is not so good here be­
cause after 19 exfs gxfS 20 �xb7 .i::!.ab8
White has better piece control to sup­
port 21 ..i.dS!) 19 .i::t d2 .l::!.d7 20 bxc6 bxc6
21 .i::!.edl .i::!.ad8 22 'ii'a4 f4 White had
some compensation for the pawn in
P.Eljanov-T.Radjabov, Moscow 200S,
although Black has chances here as
16 bs c) 16 .i::!. d1 .i::!.d8 17 bS (instead 17 cs
White plays on the light squares. fxe4 18 cxd6 exf3 19 dxe7 .i::!.x dl+ 20
There are several alternatives: 'ii'xdl 'ii'xe6 21 'ii'd8+ 'it>h7 was H.Kallio­
a) 16 cs?! looks panicky and back­ V.Kotronias, Batumi 2002, when White
fires quickly: 16 ...fxe4 17 cxd6 exf3 18 could not take the rook because 22
dxe7 .i::!.e 8 19 �b2?! (better is 19 gxf3 'ii'x a8? 'ii'g4 wins for Black) 17...'ii'xe6
I!xe7 20 .i::!.x es lL'ihs 21 .i::t e 3 lLJf4 22 .i.b2 18 �a3 (after 18 bxc6 both 18 ... lLixc6
'ii'f8 23 .i::t a el 'ii'f6, but Black is doing 19 exfs gxfs 20 'ii'xb7 lL'id4, E.Bacrot-

1 08
The M a r de/ Plata Va ria tio n : 9 b4 li'i h 5 1 o .i::!. e 1

T.Radjabov, F I D E World Championship, 16...'ii'x e6 17 �a3

Tripoli 2004, and 18 ...bxc6 19 �a3 'it>h7 White keeps the tension, but now
20 'ii'a4 .i::!.ab8 21 �a6 fxe4 22 li'ixe4 li'ifs Black can close the position. The im­
23 li'ixf6+ 'ii'xf6 24 �xc6 .l::!.b6 25 'ii'a4 mediate 17 bxc6 has not been played
li'id4, Y.Pelletier-A.David, French League or mentioned, probably because after
2007, gave Black enough piece play) 17 ... bxc6 18 �a3 .i::!.ab8 19 'ii'a4 li'id7!
and here: White cannot take the a7-pawn and'ib6 is threatened.

cl) 18 ...'it>h8 19 bxc6 li'ixc6?! 20 exfs

gxfs (after 20 ... 'ii'xfs 21 'ii'xb7 li'id4 22 11 . .cs!?

�e4 li'ixe4 23 li'ixe4 .i::!.ab8 24 'ii'ds li'ic2 Black can also play 17...'>t>h7 when
25 li'ixd6 'ii'f6 White has 26 li'if7+!) 21 18 .i::!.adl .i::!.fd8 19 bxc6 bxc6 (after
�ds 'ii'e7 22 li'ibs li'ie8 23 .i::tab1 and 19 ... li'ixc6?! 20 exfs both 20...gxfs 21
White had the initiative in E.Bacrot­ 'ii'xb7 li'id4? 22 .i::!.xd4 and 20...'ii'xfs 21
T.Radjabov, Bled Olympiad 2002. 'ii'xb7 li'id4, as in V.Mikhalevski­
c2) 18 ...'>t>h7 is given by Bologan. C.Matamoros Franco, Drammen 2005,
This is a subtle improvement that I had and then 22 �xd6! are bad for Black)
to figure out and 19 bxc6 (or 19 .i::!.a bl 20 'ii'a4 .i::!.a c8 21 'ii'a6! gave White com­
.l::!.d7 20 bxc6 li'ixc6 21 exfs 'ii'xfs 22 pensation for the pawn in P.Eljanov-
�xc6 bxc6 23 .i::!.xd6 .i::!.ad8 24 cs e4 - 1.Nataf, Mallorca Olympiad 2004. Here
Bologan) 19'ixc6!? 20 exfs 'ii'xfs 21 Black could try 20...f4, with play similar
'ii'xb7 li'id4 22 �xd6 (it transpires that to Eljanov-Radjabov in note 'b' to
this time after 22 �e4 li'ixe4 23 li'ixe4 White's 16th move, above.
.i::tab8 24 'ii'ds li'ic2 25 li'ixd6 'ii'f6 there 18 exfs
is no check on f7 and so Black wins ma­ This wins back the pawn, but Black
terial ! ) 22 ....i::tac8 23 li'ibs li'ixf3+ 24 will get good piece play. Instead 18
'ii'xf3 'ii'xf3 25 gxf3 a6 is equal accord­ .i::!.adl f4 19 .i::t d3 .i::!.ad8 20 h3 was
ing to Bologan. S.Porat-R.Djurhuus, Drammen 2005,

1 09
A ttacking C h e s s : Th e King 's I n dia n , Vol u m e 1

and here Mikhalevski suggests 20 ... g5 perienced certain bouts of popularity.

21 .i::!.edl b6 with the idea of ...h 5 and At first this move looks very strange
... g4. positionally - usually White is aiming
18 ...gxf5 19 �xb7 I!ab8 for c4-c5, but in the Bayonet the ad­
vance b4-b5 is a common theme be­
cause White's initiative is largely based
on his light-square play, especially with
lL'ig5-e6 always in the air. These days it
is known that the main line basically
leads to a drawish ending. Black does
have one main deviation, but it entails
some real risks.
13 ...cxds
Instead 13 ... h6 14 lLie6 �xe6 15
dxe6 fxe4 16 lLixe4 lLixe4 17 �xe4 d5
20 .i.ds 18 cxd5 cxd5 19 �a3 leads to the same
After 20 �f3 e4 Black intends ... lLic6 thing. This move order would deny
or ... lLig6: for example, 21 �e2 lLic6! 22 Black of the option given at move 16,
�b2 lL'id4 23 'ii'dl lL'id7 with excellent however.
piece play. 14 cxds
20 ...lLifxds 21 cxds �f7 22 �b2 e4 23 Sometimes the recapture 14 exd5
.l:!.adl ll'ig6 has to be taken into account when
Black stood well in A.Korobov­ Black plays ...cxd5 before ...h6, but in
D.Yevseev, Sochi 2008. this case White has problems on the
long diagonal after 14 ... e4.
A2) 13 bS 14...h6
If Black really wants to avoid the
forcing lines that follow, the rare
14 .. .fxe4 could be played. After 15
lLigxe4 lLixe4 16 lLixe4 lL'if5 White is a
little better, but these positions are not
so bad for Black when the c-pawns
have been exchanged.
1s lL'ie6 �xe6 16 dxe6 fxe4
Thi s is the theoretically approved
line, but if Black wants to take a chance
there is 16 ....i::!.c 8!?.
This forcing continuation has ex- This looks attractive (at least in the

Th e M a r de/ Plata Va ri a t i o n : 9 b4 li'i h 5 1 0 .i::!. e1

sense that i t does not just lead t o a

level endgame), but I am not so sure it
is worth the trouble ... Here White has:

bl) 20 ... .i::t c3 ?! 21 dxe7 (much

stronger than the 21 �a3? .i::!.xb3 22
dxe7 .i::txa3 23 exd8'ii' .i::!.x d8 24 .i::!.adl of
a) 17 �b2 fxe4 18 li'ixe4 li'ixe4 19 Ki.Georgiev-R.Ponomariov, Istanbul
.l:txe4 (19 .i.xe4 dS 20 �a3 dxe4 21 Olympiad 2000, when both 24...l:te8
'ii'xd8 .i::!.fxd8 22 �xe7 .i::!.e 8 is a familiar and 24....i::!.c 8 are good for Black)
ending, but Black has the extra move 21 ...'ii'xe7 22 'ii'a4 exf3 23 �b2 f4 (or
_ ..i::tac8) 19 ... ds 20 .i::!. x es �xes 21 .i.xes 23 ... �gs 24 g3 'ii'g4 2s 'ii'xg4 fxg4 26
is very similar to Line A3. After 21 ...�b6 .i::tadl .i::ta 27 e7 .i::!.e 8 28 .i::!.d8 1-0,
(21 ... .i::tc4 is also possible) 22 �b2 .l::!.c4 S.Savchenko-F.Guilleux, Le Touquet
(not 22 ...'>t>h7 23 �e2 d4 24 h4 .i::!.f6 2S 2007) 24 'ii'e4 'ii'g s 2s g3 .i::t e 3 26 �c4
.l:!.e1 d3 26 �es when White was win­ was drawn here in R.Szuhanek­
ning in T.Nyback-D.Stellwagen, Ger­ N.Grigore, Eforie Nord 2008, but White
man League 2oos) 23 'ii'e2 'ii'a 24 �e3 is much better.
�f4 2S 'ii'xa7 (after 2S �a3 'ii'd2 26 b2) 20 ... �xal 21 dxe7 'ii'xe7 22 �a3
.l:!.dl l:te4! 27 '>t>fl 'ii' c 2 28 �xe7 'ii'xb2 'ii'f6 23 e7+ .l::!.f7 24 �e2! (Zakhartsov's
29 'ii'd7 .i::!.xf3 30 gxf3 'ii'e2+ 31 'it>g2 idea; instead 24 �dS .i::!.e 8 2S �dl 'ii'c 3
.l::!.g 4+! it's a draw) 2s ....i::t c2 26 �a3 'ii' e s was good for Black in S.Savchenko­
27 .i::tf1 .i::tx a2 28 'ii' c s .i::!.xf3 29 gxf3 �gs+ M.Mozharov, Sochi 2008) 24 ... �es (also
30 'it>h1 'ii'fs 31 'it>g2 �gs+ 32 'it>h1 �fs bad are 24... '>t>g7 2S .i::t xal 'ii'xal+ 26
White could not avoid a draw in .i.fl and 24 ... '>t>h7 2S b6! a6 26 'ii'a4) 2 S
Z.Peng-Li Shilong, Wijk aan Zee 2008. b 6 ! a 6 ( 2 s...�xb6 26 'ii'xb6 axb6 2 7
b) 17 �b3 ds 18 li'ixds li'ifxds 19 26 �c4 '>t>g7 21 �xf7 'ii'xf7 28
exds e4 20 d6 and here: .i::!. d l and the e-pawn will give Black big
(seefollowing diagram) b3) 20...'ii'xd6 21 �a3 �f4 (worse
are 21 ...�dS? ! 22 .l:!.adl 'ii'xb3 23 axb3

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

exf3 24 �xe7 .i::!.fe8 2 S .l::!.d7 �f6, as in .i::!. cl 'l;e8 28 e7 V2-V2 L.Fressinet­

Xu Jun-Ye Jiangchuan, Taiyuan 2004, M.Hebden, Lausanne 2001.
and then 26 �b4!, and 21 ... 'ii'b6 22 22 �cs .i::!.x e6 23 �e3
�xe7 .i::!.fe8 23 �b4 �xal 24 .i::txal exf3 This is White's latest attempt to
2 S e7+ 'l;g7 26 �a3 ! with a strong ini­ keep a little something in the position.
tiative) 22 .i::!.a dl (not 22 �xe7 .i::!.fe8 23 He hopes to develop some pressure
�a3 .i::tc 3 24 'ii'ds exf3 2s �d6 'ii'g s 26 against Black's queenside. In general,
�g3 f4 when Black is better) 22 ....l::!.C 3 taking on e4 does not lead to anything
23 �4 .i::txa3 (better than 23 ... .i::!.e8 24 because Black's king can become active
�e2) 24 'ii'xa3 exf3 2S .l::!.d7 �f6 26 very easily, which makes up for the
'ii'xf3 'ii'b4 27 'ii'e2 with an unclear po­ weak es-pawn. My first experience
sition according to Zakhartsov. with this line was back in 1996 when
11 ll:lxe4 GM Igor Nataf asked me to play the
17 �xe4 ll:lxe4 18 ll:lxe4 dS 19 �a3 white side of this endgame in some
makes no difference blitz games on line to see if it was ten­
17 ...ll:lxe4 18 �xe4 ds 19 �a3 able for Black. It was. A more serious
This leads by force to an ending example: 23 .i::!.xe4 a6 24 b6 .i::!.c 8 2S �e3
where White hopes to be marginally .l:!.d6! and Black was at least equal in
better. Instead he must avoid 19 �c2 M.Brodsky-A. Shimanov, St Petersburg
e4 20 .l::!.bl 'ii'b 6 when Black is hitting 2008; Black's king will come to e6.
both f2 and e6. Instead 23 .i::!.a dl a6 24 bxa6 .i::!.exa6
19...dxe4 20 'ii'x d8 .i::!.fxd8 21 �xe7 2S a3 �f8 was also level in E.Bacrot­
T.Radjabov, Dresden Olympiad 2008.
23 ...a6
Black removes the a7-pawn from
the scope of White's bishop. The care­
less 23 ...b6?! 24 a4 a6? runs into the
breakthrough 2s as! axbs 26 axb6
when Black is in big trouble.
24 b6
The only try to eke anything out of
the position is to advance the pawn,
but it could become weak here.
21 ....i::te s 24....i::!. d S
The alternative 21 ....i::td s!? has only This looks like the simplest move to
seen one outing, although it was suc­ me. Black should certainly not go to
cessful: 22 .i::!.adl .i::!.x bs 23 �d8 .i::!.c 8 24 sleep just yet, though, as these exam­
.i::!.d6 .i::!.c6 2S .i::!.xc6 bxc6 26 h4 �8 27 ples show:

Th e M a r de/ Plata Varia t i o n : 9 b4 li'i h 5 1 o .i::!. e 1

a) 24...�f6 intending ... �gs has al­ �d6 32 'it>g2 'it>e6 33 .l::!.g4 gs 34 h4 the
ways been considered a simple solu­ game was agreed drawn immediately
tion, but White may still push a little in N .Rashkovsky-F.Florian, Biel 2001,
bit here: 2S g3 (2S a4 has also been although Black could have tried
tried) 2s ... �gs 26 �xgs hxgs 27 .i::!.xe4 34 ... �xe7 with the idea of 3S �xc7
.i::!.xb6 28 .i::!.x es .i::t bs 29 .i::t ael (after 29 (better is 3S .l::!. c 4) 3S .. 1Ixc7 36 hxgs
.i::!.xbs axbs 30 .i::t d 1 '>t>g7 31 .l::!. d7+ 'it>h6 'it>f S! when he is suddenly much better.
32 .i::!.xb7 .i::txa2 33 .i::txbs g4 Black drew Another game went 28 g4 'it>e6 29
easily in J .Lautier-V.lvanchuk, Mallorca 'it>g2 .l::!.xe7 30 bxe7 .i::!.c8 31 �b6 �f8 32
Olympiad 2004) 29 ... g4 30 .i::t1e4 .i::!.x es .i::!.xe4 �d6 and Black even went on to
31 .i::!.x es bS 32 .i::tg s �7 33 .i::!.xg4 and win in A.Van Beek-G.Kodentsov, Gron­
White went on to win in S.Savchenko­ ingen 2007.
V.Talla, Legnica 2007.
b) 24....i::!.c 8?! 2s .i::tacl .i::!.ec6 (2s ....i::tcc6 A3) 13 �b2
26 .i::!.edl �f8 27 'it>fl also gives White
some hope) 26 .i::!.c dl! .i::!.f6 27 .l::!. d7 .l::!.f7
28 .i::!.edl .i::!.e8 29 .l::!.c 7 and White main­
tains some pressure.
c) 24 ... '>t>f7!? leads to the main line
after 2s .i::ta cl .i::!.d 8 26 .l::!. e7+ .l::!. e 7 27 �cs
.i::!.e d7.
25 .i::t a cl '>t>f7
Or 2s ... .i::!.ed6 with the idea of .. 1I8d7
and ... '>t>f7-e6.
26 .l::!. c 7+ .l::!. e7 27 �cs .i::!.e d7
This move became popular after
Shirov used it to defeat Radjabov. The
bishop does not look very effective on
b2, but White plans a dangerous ex­
change sacrifice to justify it.
13 h6

Black heads down the rabbit hole.

The alternative is the immediate
13 ... cxds 14 cxds (Black must also
worry about 14 exds) 14...fxe4 (instead
14 ... h6 lS li'ie6 �xe6 16 dxe6 fxe4 17
Black is totally fine and after 28 g3 li'ixe4 li'ixe4 18 I!xe4 dS brings us back
�f8 29 .i::!.xe4 .l::!. xe7 30 bxe7 .l::!. c 8 31 �b6 to the main line) lS li'igxe4 li'ifs 16

A ttacking C h e s s : Th e King 's I n dia n, Vo l u m e 1

ll:lxf6+ and now both 16 ... �xf6 17 lLle4 after 1 7 �d3 e4 1 8 �fl 'ii'b6 19 .l::!.bl
�g7 18 .i::!.cl �d7 (V.Laznicka-G.Jones, ll:lhs in E.Bareev-T.Radjabov, Enghien­
European Team Championship, Herak­ les-Bains 2003) 17 ...cxdS 18 �c2 and:
lion 2007), and 16 ... .i::txf6 17 lLle4 .l::!.f7 18 cl) 18...'ii'd6 19 �b3 .i::!.ad8!? (this
g3 (D.Vigorito-J.Rihel, Somerville 2009) looks better than 19 ... 'ii'xe6 or
18...�d7 are only a little bit better for 19 ...'ii'xb4) 20 ll:la4 ll:le4 21 f3 ll:lgs 22
White. ll:lcs ll:lxe6 looks okay for Black.
14 ll:le6 �xe6 15 dxe6 fxe4 c2) 18 ... e4 19 ll:la4 (it may be better
Black can try to avoid the coming to just play 19 �b3 'ii'b6 20 .l:!.cl 'it>h7 21
complications with 1S ...'ii'C7 16 'ii'b 3 (16 bS 'ii'xe6 22 �a3 .i::!.ad8 with an unclear
cs ds 17 exds ll:lfxds 18 �xds .i::ta d8 position) 19 ... b6 20 f3 exf3 21 'ii'xf3
looks okay for Black), and here �d6! 22 .i::!.adl .i::tac8 23 �bl .l::!.c4 and
16 .. .fxe4!? looks worth a try after either the strong threat of ... ll:lg4 gave Black
17 i.. xe4 .i::!.a e8 or 17 ll:lxe4 ll:lxe4 18 the initiative in A.Mista-R.Antoniewski,
�xe4 .i::!.f6. In practice Black has pre­ Trzebinia 1998.
ferred 16 ... .i::tad8, but 17 .i::!.adl fxe4 18 16...ll:lxe4 17 .i::!.x e4
ll:lxe4 ll:lxe4 19 �xe4 �c8 20 'ii' h3 I!f6 Worse is 17 �xe4 dS 18 cxds cxds
21 ..tc2! .i::!.xe6 22 f4 I!f8 23 fxes (worse 19 �c2 �b6 20 'ii'd2 e4! 21 �xg7 'it>xg7
is 23 cs dS 24 fxes, as in W.Paschall­ 22 �b3 .i::tfs and Black was better in
D.Karatorossian, Budapest 2004, be­ M.Peek-A.David, Amsterdam 2000.
cause of 24...h s with the idea of ...ll:lfs)
23 ... dxes 24 cs ll:lds 2s �b3 .l::!.e 7 26
�xc8 .i::!.xc8 27 �xds+ cxds 28 I!xds e4
29 ..txg7 'it>xg7 30 'it>f2 was clearly bet­
ter for White in L.Vrbica-T.Pupak, corre­
spondence 2004.
16 ll:lxe4
Instead 16 �xe4 gives Black a rather
pleasant choice:
a) 16 ...'ii'b6 17 �b3 dS! 18 cxds cxds
19 �xds ll:lg4 gave Black counterplay
in S.lvanov-B.Avrukh, Beersheba 1998. 11 ...ds
b) 16 ...ll:lxe4 17 ll:lxe4 ll:lfs 18 'ii'g4 The only real alternative is Galla­
'it>h7 (18 ... 'ii'e8 19 e7!) 19 .i::!.adl 'ii'e 7 20 gher's old suggestion of 17 ...ll:lfs. After
Ild3 .i::tae8! (not 20...'ii'xe6 21 .i::!.xd6) was 18 bS! .i::!.c8 19 .i::te 2 .i::!.e8! ? (worse is
satisfactory for Black in O.Averkin­ 19 ...�e8 20 bxc6 bxc6 21 'ii'a4, V.Golod­
V.Zakharstov, Krasnodar 1998. T.Nedev, European Championship,
c) 16 ... ds 17 cxds (Black was better Plovdiv 2008) White has:

Th e M a r de/ Plata Varia tio n : 9 b4 li'i h 5 1 o .i::!. e 1

a ) 20 bxc6 bxc6 21 cs (21 �a4 'ilia and introduces the threat of a battery
22 .l::!.bl .i::!.xe6 23 �c3 .l::!.e 7 24 �e4 'it>h7 on the a1-h8 diagonal. other moves are
2S g4?! li'ih4 26 �el .l::!.f7 27 .i::!.b3 .l::!.f4! less dangerous:
gave Black the upper hand in A.Wirig­ a) 21 �d4 �xe6 22 �cs was played
N.Pokazanjev, Winterthur 2008) 21 ... ds in J.Vilela-R.Vazquez, Cuba 1996. Here
22 �xes �xes 23 .i::!. x es �f6 24 .i::t e 1 Markos suggests 22 ... .i::tfd8 23 �d2 li'ifs
was agreed drawn here in R.Dautov­ when White probably does not have
S.Kindermann, Nussloch 1996. After quite enough compensation for the
24 ... .i::txe6 the position is pretty level. exchange.
b) 20 �a4 'fib6 21 .i::!.d1 (or 21 bxc6 b) 21 'fid2 'illx e6 (21 ... '>t>h7 22 .i::!.e l as
bxc6 22 �e4 I!xe6 23 .l:tb1 �c7 24 �c3 was played by a young Radjabov, but is
li'id4 2S .l:teb2, which gives White com­ not as trustworthy) 22 .i::!.e l (or 22 �d4
pensation for the pawn) 21 ... .i::txe6 22 I!xf3) 22 ....i::txf3 ! is a thematic idea.
�b3 �d8 23 cs ds 24 .l:txds! cxds 2 s Black gives back the exchange to ruin
�xds .i::!. x cs 2 6 �xe6+ '>t>h7 27 g3 li'id4 White's pawn structure and give his
28 �xd4 �xd4 29 'it>g2 and White kept knight a strong outpost. After 23 gxf3
an edge in R.Buhmann-M.Klenburg, li'ifs the knight is well placed to both
Neustadt an der Weinstrasse 2009. attack and defend. One example, which
18 cxds was a serious upset, T.Nyback­
Retreating with 18 .i::!. el gives Black M.Matthiesen, Copenhagen 2003: 24
no problems after 18 ... 'fid6. �g3 �f7 2s .i::t c1 d4 26 .i::t a 'fids 27
18...cxds 'fid3 .i::!.e8 28 a4 hS 29 'it>g2 �e6 30 .i::!.xb7
h4 31 .l:txa7 h3+ 32 '>t>gl 'flel+ 33 'fifl
li'ixg 3 34 fxg3 'fle3+ 0-1.

19 Ilxes!
The point of White's play.
19...�xes 20 �xes 'fib6 21 �b2! 21....i::ta d8
This was Shirov's innovation. White This is considered best by Markos.
opens the e-file to protect the e6-pawn Black creates the possibility of shutting

A ttacking C h e s s : The King 's I n dian, Vo l u m e 1

the long diagonal with ...d4. others: 'ii'b 2+ (Bologan only gives 2 S h4 b6 as
a) 21 ...'>t>h7 was Black's original try. unclear) 2s ... .i::!.f6 26 h4 b6 27 g4 g s 28
Then 22 'ii'e2 d4 23 h4 .i::!.f6 24 .l:tel �g2 .l::!.c4 29 'ii'e s White kept some
'ii'xb4 2S a3 'ii'd6 26 h S .i::!.af8 27 'ii'e4 pressure in M.Boccia-S .Giuliani, corre­
ll:lc6 28 hxg6+ '>t>g7 29 �cl! 'ii'e 7 30 spondence 2004.
�xh6+ 'it>xh6 31 'tlfli4+ 'it>xg6 32 �xc6
bxc6 33 .l:te S leads to a further theoreti­
cal divide:
al) 33 ... 'ii'xe6? 34 .i::!.xe6 .i::!.xe6 3 S
'ii'g4+ '>t>f7 3 6 'ii'xd4 was winning for
White in A. Shirov-T.Radjabov, Linares
a2) 33 ... .i::txe6 is a better try: 34
'tlfli s+ '>t>g7 3 S .i:tgs+ 'ii'x gs 36 'ii'x gs+
.i::!. g 6 3 7 'ii'e s+ 'it>g8 38 'ii'xd4 .l::!.f7 39
'ii'd8+ (after 39 g3 .i::!.f g7! White will
have trouble advancing his pawns) 22 'ii'e 1
39 ....l:tf8? (39 ...'>t>h7 has to be a better This is an interesting try. White pro­
try, with the idea of 40 g3 .l:tfg7!, al­ tects the b4-pawn and may still con­
though White will still certainly try to tinue with 'ii'e s at some point.
probe with 40 'ii'd3 !) 40 'ii'd7 .l::!.f7 41 a) 22 'ii'e 2 'ii'xb4 23 'ii'e S allows
'ii'e 8+ '>t>g7 42 g3 and White won the Black to execute his idea with 23 ... d4
ending in M.Casella-1.Zenyuk, U S when he stands better.
Championship, S an Diego 2004. I can b) 22 'ii'd3 'ii'x e6 23 'ii'd4 transposes
still remember preparing this line with to variation 'd2', below.
Casella before the game ... c) 22 bS 'ii'x e6 23 'ii'd4 ll:lfs 24 'tlfli8+
b) 21 ...'ii'xb4 is Bologan's recom­ '>t>f7 2 S 'ii'h 7+ 'it>e8 26 �g4 .l::!.f 7?! gave
mendation. However, after 22 .l:tbl White excellent compensation after 27
.i::!.ac8 (instead 22 ...'ii'c s 23 �d4 'ii'c 7 24 'tlfli8+ 'it'd 7 28 'ii'c 3 .i::!. c 8 29 'ii'd2 .l::!.c4 30
�b2 'ii'c s 2s �d4 'ii'c 7 26 �b2 was �f3 in B.Lalith-A.Lahiri, Bhubaneswar
drawn by repetition in W.Fademrecht­ 2010, but Black can improve here with
F.Gerhardt, correspondence 2004, but 26 ... 'ii'f 7! 27 .l:tel+ ll:le7 28 'ii'xf7+ (28
White could also consider 24 'ii'd 2 'it>h7 'ii'xh6 'ii'xf2+) 28 ....i::txf7 when he is bet­
2 S h4 with compensation) 23 �g7 (23 ter.
�es 'ii'a 3 eyes cl and Bologan gives 24 d) 22 'ii'd2 has been White's most
'ii'd2 'it>h7 2S .i::!.xb7 'ii cl+ 26 �dl 'ii'xd2 common choice. Then following
27 .l:txe7+ 'it>g8 28 .l::!. g 7+ with perpetual 22 ... 'ii'x e6! (after 22 ... d4 23 .i::!. el .i::!.f6 24
check) 23 ...'ii'xbl 24 'ii'xbl 'it>xg7 2 S ..lli.g4 'it>h7 2S h4 h S 26 �h3 Black could

Th e M a r de/ Plata Va ria tio n : 9 b4 li'i h 5 1 o .i::!. e 1

hardly move in R.Buhmann-T.Nedev, d32) 25 .i::!.x e7 (forcing the draw)

Hersonissos 2007) we have: 2 5 ... 'ii'xe7 26 'ii'xg6+ 'ii'g7 27 'ii'e6+ 'ii'f7
dl) 23 'ii'x h6 d4 24 'ii'd 2 .i::!.xf3 ! 2 5 (27... .l::!.f7 28 �h5 does not help Black,
gxf3 li'if5 is given by Markos. This is a while 27 ... 'l;h8 28 �h3+ �7 29 'ii'g4
typical set-up for Black. White's king 'ii'b 1+ 30 �dl 'ii'xb2 31 'ii'h 5+ is also a
position is also loose and the strong d­ draw) 28 'ii'g4+ 'ii'g 7 29 'ii'e6+ with per­
pawn and knight give Black compensa­ petual.
tion for the pawn. 22 .i::!.d 6

d2) 23 'ii'd4 looks scary, but the Black could also try 22 ...d4 23 .i::!.d1
checks do not lead to anything: 23'if5 (after 23 �g4 h5 24 �h3 .i::!.de8! Black
24 'tlfli8+ 'l;f7 25 �7+ 'l;e8 26 'ii'xb7 intends'if 5) 23 ... h 5! ?.
.l::!.f7 27 'ii'h 5+ �8 when Black is able to 23 I!dl
get 'castled' again and his pieces coor­ Putting pressure on the d5-pawn
dinate well. Still, White retains com­ makes it more difficult for Black to play
pensation for the exchange. ... li'if5. Instead 23 'ii'c 3 d4 24 'ii'b3 .i::!.f6
d3) 23 I!el could lead to perpetual brings White's initiative to a standstill,
check after 23 ... 'ii'f7 (forced) 24 'ii'x h6 (if while 23 'ii' d2 .l:!.xe6 24 �d4 (after 24
24 'ii'd4 'l;h7 25 'ii'xa7 li'ic6 Black can 'ii'x h6 d4 25 'ii'd2 Black plays the usual
fight for the initiative after 26 'ii'b6 2 5 ... .i::txf3 ! 26 gxf3 - not 26 �xd4? .i::!.d6
.i::!.fe8 or 26 'ii'c 5 d4 27 b5 J:tfe8! ) 24 ... d4 winning - 26'if5 with thematic play)
and then: 24 ... 'ii'd6 25 'ii'xh6 is again not as scary
as it looks after 25'if5 26 �8+ 'l;f7
27 �7+ 'l;e8 28 �c3 J:te7 29 'ii'h4 d4!?
with unclear play.
23 .i::txe6 24 'ii'c3 .i::!.ef6 2 5 'ii'd 2

d32) 25 'ii'd2 li'if5 is unclear, but 26

�dl? (26 �e4) 26 ...'ii'xa2 27 h4 .i::!. de8
28 .i::!.xe8 .i::!.xe8 29 g4 'ii'e6 30 'l;h2 'ii'd6+
31 rt;g2 'ii' d5+ 32 �h2 li'ixh4 33 rt;g3
'ii'g 2+ and 0-1 was a significant upset 2 s....i::txt3
in A.Fier-K.Sai, Bhubaneswar 2010. This is certainly safer than 2 5 ... l!f4!?

A ttacking C h e s s : Th e Kin g 's I n dian, Vo l u m e 1

26 �es .i::!.xb4 27 'ii'xh6 when White has do not like this move order because it
the initiative. After 27 ...'l;f7 28 'ii'd2 allows the extra option of 14 exdS! ?
'l;e8 29 �d4 .i::!.xd4?! (29 ... 'ii'd6 is better, which scores rather well for White. In­
but White has compensation for the stead after the more common 14 cxds
exchange) 30 'ii'xd4 'ii'xd4 3 1 .i::!.xd4 h6 lS tlle6 �xe6 16 dxe6 fxe4 17 tllxe4
White had the better ending in tllxe4 18 �xe4 dS we just transpose
V.Golod-N.Huschenbeth, Hamburg back into the main line with 13 ...h6.
2008. A more serious option is 13 .. .f4!?.
26 gxf3 .i::!.xf3 27 .i::te1 .i::tf7 28 �d4 �d6 After 14 �cl it seems as though White
29 'ii'xh6 has just wasted time, but the centre
Or 29 'ii'b6 30 �d4 'ii'd6 with a has been stabilized and now the bishop
repetition. will head to a3. Following 14... h6 lS
29 ... .l:!.h7 30 �cs .i::!.xh6 31 �xd6 lllfs tll e6 �xe6 16 dxe6 tllc 8 17 bS 'ii' e 7
The endgame is equal (Golod). (worse is 17 ...'ii'e 8?! 18 bxc6 bxc6 19 cs
'ii'xe6 20 �a3 dxcs 21 ll'la4 tll b 6 22
A4) 13 �e3 tllxcs 'ii'f7 23 'ii' c l �c4 24 �dl! with
the better chances for White in
V.Kramnik-B.Gelfand, Belgrade 1997)
18 bxc6 bxc6 19 cs dxcs 20 �a3 tllb6
21 ll'la4 .i::tfd8 22 'ii'b 3 tllxa4 23 �xa4
�xe6 24 �xcs ll'ld7 2s .i::!.e cl tllxcs 26
I!xcs .i::!.d6 the position is equal.
14 tlle6 �xe6 15 dxe6

This move used to be the main line

and it has been played more than any
other move. Nowadays it is not consid­
ered so dangerous for Black.
13... h6
Instead 13 ... 'l;h8 14 exfs! was
pointed out by Markos and this looks
quite good for White. 1s ...fxe4
Bologan states that Black should Instead 1S .. .f4 16 �cl takes us back
play 13 ... cxds immediately, although to 13 .. .f4 in the previous note.
he gives no explanation why this is so. I Black could also consider l S ... gs.

Th e M a r de/ Plata Va ria tio n : 9 b4 li'i h 5 1 o .i::!. e 1

This has not had much success, but it apart Black's centre) 20 ... fllc8 (taking
does not look too bad. After 16 exfs immediate aim at the e6-pawn; 20... d4
li'ixfs 17 �d3 li'ixe3 18 .i::!.x e3 "ille7 19 21 .i::!.adl "illd6 is an alternative) 21 bxc6
.i::!.d1 .i::!.ad8 (19 ... flixe6 20 "illxd6 �xc4 21 (2 1 �b3 d4 22 �d2 "ilia 23 c s ! ?, main­
bS gives White the initiative) 20 li'ie4 taining the e6-pawn, is also possible)
Black has: 21 ... d4 22 �d2 fllxe6 23 fllxe6+ .i::!.xe6
a) 20... g4?! 21 li'ixf6+ flixf6 22 �xg4 (Black is likely to pick off the c6-pawn,
�xf2+ 23 '>t>hl gave White a big advan­ but White's bishop-pair will be fair
tage in V.Kramnik-F.Nijboer, Wijk aan compensation) 24 �e4 was Huang
Zee 1998. After this game nobody Qian-Ju Wenjun, Hefei (rapid) 2010.
wanted to play 1s ... gs anymore. Now Black could play 24....i::!.c8 because
b) 20 ... ds 21 li'ixf6+ flixf6 22 cxds 2s .i.ds? fails to 2s'ixds 26 cxds .i::!.d6
cxds 23 e7 (23 �xds? .i::!.xds) 23 ... �xe7 when the pawns begin to fall.
24 �xds+ 'it>h8 is given as slightly bet­ 18...cxds
ter for White by Kramnik. This certainly
looks pretty miserable for Black.
c) 20 ... li'ixe4 21 �xe4 fllxe6 22 bS
gives White good compensation for the
d) 20 ... �xe6 and now Kramnik
claimed that White has the upper hand
after 21 li'ixd6, but 21 ... li'ie8 22 cs li'ixd6
23 cxd6 I!f6 looks tenable, since Black
will win the d-pawn and have an extra
pawn. White certainly has some com­
pensation, but Black does not really 19 �C2
appear to stand worse. This leads to a complicated middle­
16 li'ixe4 li'ixe4 17 �xe4 dS 18 cxdS game. White can also head for an end­
White can also refrain from this ex­ ing with 19 �cs dxe4 20 flixd8 .i::!.fxd8
change with 18 �c2, but this does not 21 �xe7 .i::!.e 8 22 �cs. The position is
look too dangerous: 18 ... b6 (Black plays identical to the main line of Line A2,
as if the exchange had already oc­ but here White's pawn is still on b4. If
curred on dS, but 18 ... d4 19 �d2 and anything, this should help Black be­
now 19 ... .i::!.f6 or 19 ... "illd6 could seek to cause his queenside will not be under
punish White's omission as the a2-g8 any pressure. Both 22 ... b6 23 �e3 I!xe6
diagonal is not yet open) 19 "illg 4 (19 24 a4 �f8 2S bS (N.lbrayev­
cxds would transpose to the main line) S.Mamedyarov, Khanty Mansiysk 200S)
19 ....i::!.f6 20 bS!? (White tries to break 2s ...a6! and 22 ...I!xe6 23 I!xe4 .i::!.d8 24

A ttacking C h e s s : Th e K i n g 's I n dian, Vo l u m e 1

h4 a6 2 s a4 .l:!.ds 26 ii.e3 .l::!.c6 27 .l:tb1 bS ing the g6-pawn after 'lli'g4. After
(Z.Gyimesi-V.Baklan, Baile Tusnad 20 ... 'lli'd6 21 'lli'g4 h S 22 'lli'h 3 (E.Ovod­
2oos) are equal. V.Rajlich, Budapest 2002) 22 ... e4, in­
19 b6
... tending ....:tfs, the b4- and e6-pawns
are weak.
b) 20 ii.a4 'lli'd6 21 ii.d7 is mis­
guided. The e6-pawn is secure, but
White's bishop is totally out of play.
After 21 ...'l!Vxb4 22 .l::!.b1 'l!Vh4 23 f3 J:tfs
24 ii.f2 'lli'f6 2 S ii.g3 h S 26 h3 .l::!.f8 27
�hl? 'l!Vg s 28 ii.h2 e4! Black broke
through to White's king in
V.Malakhatko-F.Jenni, Istanbul Olym­
piad 2000.
c) 20 bS 1'6d6 21 is not an un­
This is a very important move. Black common idea, but here it is painfully
must prevent ii.e3-cS or his centre will slow. Now 21 ... .l:tad8 22 ii.b3 �h7 23 a4
fall apart. He has a strong centre and 'l!Vxe6 24 ii.a3 was V.Malakhatko­
White's e6-pawn is likely to fall, but S.Sivokho, Polanica Zdroj 1999, and
Black must be careful here because here 24 ... e4! 2S .:tel 26 ii.xe7 (26
White has the bishop-pair and the .l::!.x e4 ii.xh2+) 26 ...'l!Vxe7 is good for
safer king. Indeed, Black must try to Black because 27 ii.xds? loses to
keep his pieces well coordinated and 27 ...'l!Vg s, intending ....l::!.xd6 or ...ii.g7.
should not necessarily be in a hurry to
win the e6-pawn. Any misstep could
result in a serious tactical accident. As
long as Black is not too careless, his
chances are good, though, and in prac­
tice he has done pretty well from this
20 'lli'g4
White must play actively or the e6-
pawn will fall. Black is likely to win it
anyway, but he must not allow White
too much activity in return. Some other 20 e4

tries: The alternative 20....l:tf6 is not so bad

a) 20 ii.b3 immediately is inaccu­ theoretically, but Black must work for a
rate because White will not be attack- draw after 21 .l:tadl 'lli'd6 22 ii.b3 .l:td8

Th e M a r de/ Plata Varia tio n : 9 b4 l'i::i h 5 1 o .l:. e 1

(22 ... .l:r.xe6 can be met by 23 J:txds l'i::ixds bl) 27 ... .l:td7 28 ii.xe7 �xe7 29 J:txds
24 .l:tdl when White will recoup his ma­ .l:txds 30 ii.xds ii.c3 (also better for
terial with a slight edge) 23 bS (other­ White is 30 ....l::!.e s 31 ii.xe4 .l:txbs 32
wise, 23 .l:txds oversimplifies: 23 ...l'i::ix ds ii.xg6+ �6 33 ii.d3 .l:tb2 34 a4) 31 J:te2
24 .l:r.dl 'l!Vxe6 2s 'l!Vxe6+ J:txe6 26 .l:txds J:tes 32 ii.xe4 and White was much bet­
.l:txds 27 ii.xds �f7 and the ending is ter in V.Kramnik-A.Shirov, Linares 1998.
drawn) 23 ... 'l!Vxe6 24 'l!Vxe6+ .l:txe6 2S b2) 27....:tes ! 28 .l::!.d4 with a final di­! �7 26 ii.a3 e4. vide:
b21) 28 ....l:te6 29 ii.xe7 �xe7 30
.l:txds .!:txds 31 ii.xds J:tes 32 ii.xe4 �f6
33 a4 was V.Mikhalevski-A.Melekhina,
Philadelphia 2008. White can certainly
play on for a fair while, although Black
did manage to draw this game.
b22) 28 ... t'i::ifs ! is best. Then Mik­
halevski gives 29 .l:r.dxe4 l'i::id4 30 .l::!.x es
ii.xes 31 �g2 ii.f6 with equality.
21 .l:r.ad1
White threatens ii.xe4, so Black
Here White can try to improve his must move his queen off the d-file.
position or he can win back the pawn 21 'l!Vc7 22 ii.b3

immediately: Instead 22 ii.a4 still looks like the

a) 27 ii.xe7 seems to be the most ac­ wrong idea. After 22 ....l::!.fs (Black could
curate: 27 ... �xe7 28 .l::!.x ds .l:txds 29 look into 22 ... 'l!Vc3 or 22 ...'l!Vc4) 23 ii.d7
ii.xds .l::!.e s (instead 29 ... ii.c3 30 .l:te2 .l::!.e s .l:taf8 24 bS, thematic is 24 ... hS when
31 ii.xe4 .l::!.xbs 32 ii.d3+ .l:tes 33 .l:r.xes+ White must commit his queen: 2S 'lli'h4
ii.xes 34 ii.xg6 is Kramnik-Shirov, be­ (2S 'l!Ve2 could be met by 2s ... or
low) 30 ii.xe4 (White must play this, 2s ... ii.c3 26 .l:r.c1 'l!Ves) 2s ...ii.f6 26 'lli'g 3
because 30 .l:txe4 �d6 31 .l:txes �xes 32 (26 'l!Vh3 is possible, but the queen does
ii.f7 gs 33 a4 was already agreed not seem well placed here) 26 ... 27
drawn in V.Mikhalevski-M.Klinova, Gi­ 'lli'h4 ii.f6 repeats. Instead 24 ..'lli'c3 2 S
braltar 2008). Now 30 ... .l:txbs 31 ii.d3+ 'l!Ve2 'l!Ves 26 'lli'd2 'lli'd6 (now 26 ...h s can
.l::!. e s 32 J:txes+ ii.xes 33 ii.xg6 is Kram­ be met with 27 'l!Vb4! ii.f6 28 'lli'a4 'lli'b8
nik-Shirov again, so Black should play 29 .l::!.e 2!, preparing to double rooks on
30...�6 31 a4 which is Mikhalevski­ the d-file) 27 h3?! (not 27 ii.xh6 ii.xh6
Melekhina below, but with a couple of 28 'lli'xh6 .:ths, but 27 a4 gs 28 'l!Vc1 is
tempi less for each side. possible) 27...g S ! 28 a4 was B.Lalith­
b) 27 g3 and here: V.A.Rajesh, Delhi 2010, and here Black

A tt a c k i n g C h e s s : Th e King 's I n d i a n, Vo l u m e 1

should play 28 ...d4! 29 ii.xd4 (not 29 ter is 24...'l!Vxb4 2 S .l::!.c7 h S , leading to

'l!Vc2 dxe3 30 .l::!.xd6 exf2+) 29 ....l::!.ds 30 variation 'e2' below with Black having
.l:!.xe4 .l:tf4! when White has some prob­ avoided the note to White's 24th move
lems. there) was L.Gerzhoy-V.Rajlich, Buda­
22 .l::!.fs!
... pest 2002, and White could play 2S
'l!Vxe4! here.
d) 23 bS is logical. The pawn is not
hanging anymore, c6 is controlled and
ii.c1-a3 becomes a possibility.

This move protects dS, cuts off the

e6-pawn and Black may also double
rooks on the f-file or play ..J::t e s. White
has tried many things here, but has
been unable to prove much. Here Black has:
23 .l:td2 dl) 23 ...'l!Ves?! 24! .l::!.af8 2s ii.a3
White prepares to double rooks on hS 26 'lli'g 3 'l!Vxe6 27 ii.xe7 'l!Vxe7 28
the d-file and leaves the dl-hS diagonal ii.xds+ �h8 29 ii.xe4 .l:tg s 30 'lli'd3 and
open for his queen. There are many White was up a pawn with better posi­
other moves: tion in L.Lenic-A.Jankovic, Sibenik 2006.
a) 23 'l!Ve2 .l::!.af8 24 'lli'a6 (24 .l::!.fl and d2) 23 ....l:taf8 24 .l:te2 'l!Vc8 2S
1/2 -Yi was T.Radjabov-J.Moreno Carnero, (White should consider 2S .l:tc2 'l!Vxe6 26
Pamplona 2002) 24 ... ii.c3 2S .!:tel 'l!Ves .!::tc7) 2s ... hs 26 'i!Vh4 ii.f6 27 'l!Vh3 �h7
26 .l:tedl ii.xb4 27 ii.d4 (27 'l!Vxa7!?) 28 ii.a3 'l!Vxe6 29 g3 .l::td8 30 f3 'lli'c8 gave
27 ... 'lli'b8 28 'lli'a4 ii.d6 29 g3 hS gave Black a good position in M.Bosiocic-
Black a good position in A.Janusonis­ 1.Saric, Split 20os.
D.Chocenka, correspondence 2oos. d3) 23 ...�h7 24 .l:te2 (after 24
b) 23 ii.d4 ii.xd4 (23 ... .l:taf8!?) 24 .!::taf8 2S ii.a3 .l:txf2 26 ii.xe7 'l!Vxe7 Black
l:i'.xd4 'l!Vc3 2S 'lli'dl .!::taf8,, 26 .!::t e 2 'l!Vc7 27 is much better following 27 .l:txds e3 or
g3 'l!Vc6 28 'l!Vc2 1/2-Yi was S.l skusnyh­ 27 ii.xds 'l!Vcs) 24 ... .l:td8 2s 26
A.Motylev, Moscow 1999. g3 hS 27 'l!Vh3 'l!Vcs is comfortable for
c) 23 .:tel 'Ii' d6 24 .l::!. e dl .l::!.af8?! (bet- Black, who is well coordinated.

Th e M a r de/ Plata Varia t i o n : 9 b4 l'i::i h s 1 0 .l:. e 1

e ) 23 .l:te2 is similar t o the main line, retreat along the dl-hS diagonal. After
but here White may bring the rook to 24 'l!Vh4 ii.f6 2 s 'lli'g3 (2s ...'lli'c 6!?
c2 as well. could be tried) 26 'lli'h4 .�f6 27 'lli'g 3 28 'lli'h4 ii.f6 the game was drawn
by repetition in Xu Jun-R.Ponomariov,
Shenyang 2000.
Returning to 23 .l:td2:

Black has:
el) 23 ....l:td8 24 bS hS 2S 'lli'h 3 �h7
26 'l!Vcs 27 .l:tc2 'l!Vxbs? (better is
21...'l!Vd6) 28 .l:te7 'l!Vb4 29 J:txa7 'l!Vcs 30
'l!Ve3 gave White a winning position in 23 'l!Vc3!?

N.Alfred-S.Nagle, Budapest 2004. This is the most enterprising move.

e2) 23 ...'lli'c 3 24 .l:tc2 (it is probably Others:
better to play 24 bS .l:taf8 2S .l:tc2 with a) 23 ... ii.c3? 24 .:tel 'l!Ves loses to 2S
some initiative for White, V.Mikha­ .l:txds ! l'i::ixds 26 'l!Vxg6+ �h8 2 7 'lli'x h6+
levski-A.David, Vlissingen 2000) �g8 28 'lli'g 6+ �h8 29 'l!Vh6+ �g8 30
24 ... 'l!Vxb4 2S l:te7 h S 26 'l!Vh3 (and here e7!, R.Sherbakov-S.lskusnyh, Tula 1999.
better is 26 'l!Ve2 'lli'd6 27 l:txe7 'l!Vxe7 28 b) 23 ...J:taf8 24 .l:tedl 'lli'c6 was
.l:txds .l:txds 29 ii.xds .l:tc8 30 g3 ii.f6, S.Savchen ko-R.Ponomariov, Alushta
which is pretty even) 26 ... as! 27 g4! 2000. Now 2S .l:tc2 'lli'd6 (worse is
(forced, because 27 a4 is good for Black 2s. . .'l!Vxe6 26 l:te7 - Ponomariov) 26
and after 27 .l:tb7? a4 28 ii.c2, as in 'l!Vxe4 ! 27 .l:txds ii.xh2+ 28 �hl
J.Sales-Lim Chuing Hoong, Kuala Lum­ .l:txds 29 ii.xds ii.f4 was level in
pur 20os, Black has 28 ... .l::!.c 8! 29 d4 A.Koz lowicz-F.Schwarz, correspondence
winning, as pointed out by Mik­ 2003.
halevski) 27 ...hxg4 28 'lli'h4 .l:te8 29 c) 23 ...'l!Vc6 24 bS!? 'l!Vxe6 2S f3 exf3
e3! and Black has good counterplay 26 .�f2 'lli'f6 27 .l:txe7 'l!Vxe7 28 ii.xds+
according to Bologan. .l::!.x ds 29 .l:txds 'l!Ve2 30 gxf3 .l:tf8 31
e3) 23 ...h S ! takes immediate advan­ 'l!Vxg6 'lli'xf3 led t o a draw in Z.Straka­
tage of the fact that the queen cannot D.Fridman, correspondence 2004.

A tt a c k i n g C h e s s : Th e King 's I n d i a n, Vo l u m e 1

d) 23 ... �h7 24 .l:tedl 'Ii'c6 25 .l:tc2 (25 tally equal.

'l!Vh3 .l:td8) 2s ...'l!Vxe6 (2s ...'lli'd6 26 b) 24 ...'l!Vxb4 25 g4 .l:tes (instead
'l!Vxe4!) 26 .l:tc7 was B.Lalith-A.Lahiri, 2s ... .l:tf6 26 .l:td4 'l!Vd6 27 .l:txds lllxds 28
New Delhi 2009. Now 26 ... hs with the 'lli'xds 'lli'xds 29 ii.xds .l:te8 30 .l::!. c 1 gives
idea of ...'lli'd6 would be unclear. White compensation for the exchange)
e) 23 ... .l:td8!? is untried but looks 26 .l:td4 'l!Vc3 ! (not 26 ... 'l!Vas 27 ii.d2 'lli'bs
sensible. 28 a4 'lli'c6 29 ii.b4 with a big advan­
tage for White - Tsesarsky) 27 ii.f4 .l:tf8
28 ii.xes ii.xes 29 .l::!.xds (forced)
29 ... ii.xh2+ 30 �g2 .l:txf2+ 31 �xf2
lllxds 32 ii.xds 'lli'g 3+ 33 �1 (not 33
�e2? 'IJVf3+ 34 �d2 'IJVd3+ 35 �Cl Ji.f4+
36 �b2 37 �cl 'lli'c3+ 38 'l!Vc2
'l!Vxel+) 33 ...'lli'h 3+ with a draw.
2s bs 'lli'b4
Black can also play 2s ... 'lli'c8 when
26 ii.d4 ii.xd4 27 .l:txd4 'l!Vcs transposes
to the main text, and in fact this was
24 'l!Vd1 the actual move order of Xu Jun-Ye Ji­
Instead after 24 .l:tedl 'l!Vxb4 25 angchuan, below. However, White has
ii.d4?! (better is Panczyk and llczuk's 25 an extra option here in 26 g4!? .l:tff8 27
'l!Ve2 when 2s ....l:td8 26 'l!Va6 would give ii.xds ii.c3 28 ii.xe4 ii.xd2 29 ii.xd2
White counterplay) 2s .. Jk8 26 'l!Ve2 'l!Vxe6 30 'l!Ve2 with good compensation.
ii.xd4 27 .l:txd4 'l!Vcs 28 g4 .:tes 29 'lli'd2 26 ii.d4
.l:tf8! 30 'lli'x h6 .l:txe6 Black stood well in But not 26 a4? ii.c3.
J.Werle-A.David, Amsterdam 2000. 26....:tdfB 27 a4 ii.xd4 28 .l:txd4 'l!Vcs 29
24 ....:tds .l:te2
Black has some other options:
a) 24... .l:taf8 25 bS (25 g4 .l:tes)
2s ...�h 7 (2s ...'lli'b 4!? has been proposed
by Panczyk and llczuk) 26 J:tc2 'l!Ves 27 d4 28 ii.a3 d3 29 ii.xe7 (29 J:td2
ii.f6 30 .l::!.xd3 .l:txf2 31 .l:td7 ii.h4! is un­
clear) 29 ... dxc2 30 ii.xc2 .l:te8 31 .l:txe4
'l!Vxbs 32 ii.a3 'lli'ds 33 'lli'xds 1/2-1/2,
A.Eriksson-G.Glatt, correspondence
2002. After 33 ... .l:txds 34 e7 ii.f6 35 ii.b3
.l:td7 36 ii.a4 .l::!.e xe7 the position is to-

Th e M a r de/ Pla ta Va r i a t i o n : 9 b4 tb h s 1 0 .l:. e 1

White covers the vulnerable f2- is not looking very attractive these days
pawn. because of Bareev's idea 13 cs tl:if4 14
29 .l::!. es 30 h3
... ii.c4 �h8 lS .l:tbl!.
Instead 30 .!:tdd2 'lli'd6 31 .l:tc2 .l::!.xe6
32 'lli'c1 .l:tf7 33 .l:te7 (not 33 'lli'xh6? .l:th7) Bl) 12 c6

33 34 .l:txf7 �xf7 3S 'lli'd2 �f6!? 36

ii.xds .:tes 37 ii.b3 'lli'xd2 38 .l:txd2 .:tcs
gave Black a slightly better ending due
to his active pieces in J.Skeels-S.Boyd,
correspondence 2006.
After 30 h3, the game Xu Jun-Ye Ji­
angchuan, Shanghai 2001, was agreed
drawn, but Black could play on after
30 ...J:tffS OT 30...�g7.

B) 12 f3
Black increases the tension as in
Line A while also threatening ... 'lli'b6+.
This choice is somewhat more risky
than Line B2, but it is quite playable
and leads to a double-edged game.
White has two fundamental lines:

811: 13 'ili>h1
812: 13 .i.eJ

Other moves do not mind the g1-a7

This move is less forcing than 12 diagonal and are rare: for example, 13
ii.f3 and it is for that very reason that it bS h6 14 tl:ie6 ii.xe6 lS dxe6 fxe4 16
has become very popular. Indeed, a lot fxe4 'lli'b6+ 17 �hl ex bS 18 ex bS .l:tac8
of tension remains in the position, al­ when White must be careful, because
though Black has a broader choice 19 tl:ids loses to 19 20 exds
here: tl:ie4 and 19 ii.b2 fails to 19 ....l::!.xc3 20
ii.xc3 tl:ixe4.
B1: '12 c6
•.. Instead 13 'l!Vb3 is reasonable, al­
82: 12 ¢>hl
••• though it takes pressure off the d-file
and 13 ...h6 14 tl:ie6 ii.xe6 lS dxe6 'lli'c 8
Black's other main move, 12 ... tl:ihs, 16 .l:tdl .l:td8 looks fine for Black.

A ttacking C h e s s : T h e King 's I n d i a n , Vo l u m e 1

811) 13 �h1

17 c5
White safeguards his king by re­ This is almost always played, but
moving it from the g1-a7 diagonal. Black should also be ready for 17 .l:td1! ?
13 ... h6 which looks natural. Here he can con­
Black forces the play. sider:
14 l2Je6 ii.xe6 15 dxe6 lbe8 a) 17 ... l2Jxe6 18 cs dS 19 exds l2Jd4
Threatening to play ... lbe7xe6-d4. 20 .l:Ixd4 exd4 21 dxc6+ (not 21 d6+
Bologan suggest the rarer 1s ... 'l!Vc7 �h7 22 dxe7 'l!Vxe7 and Black wins ma­
when 16 bS (instead 16 ii.b2 .l:tad8 17 terial) 21 ... �h7 22 cxb7 .l:tb8 23 ii.f4 is
'l!Vb3 .l:tfe8 18 .l:tadl �h7 19 ii.d3 fxe4 20 very messy, but looks roughly balanced.
lbxe4 lbxe4 21 J:txe4 l2Jfs 22 f4 J:txe6 23 b) 17 ...�h7 18 cs dS 19 exds cxds 20
fxes was V.Epishin-1.Nemet, Biel 1996, lbxds lbexds (worse is 20...lbcxds 21
and here Epishin suggests 23 ...ii.xes) ii.c4 e4 22 .l:tb1) 21 ii.c4 e4 22 ii.xds
16 ....l:tfd8 17 bxc6 bxc6 18 .l:tbl 'lli'c 8 19 (after 22 .l:tb1 'lli'h 4! 23 g3 'l!Vh3 24 ii.xds
'lli'a4 'l!Vxe6 20 .l:tb7 as 21 ii.e3 .l:td7 22 lbxds 2s 'lli'xds .l:tad8 26 'l!Vb3 J:td3 ! 27
.l:tb6 .l:te7 23 .l:tebl lbd7 is fairly level in 'l!Vc2 exf3 Black has good play)
his view. 22 ...lbxds (not 22 ... ii.xal 23 ii.xb7 with
16 'l!Vb3 a big advantage) 23 .l::!.x ds 'lli'f6 24 e7
After 16 bS Black should play 'l!Vxe7 2S ii.b2 ii.xb2 26 'l!Vxb2 .l:tad8 and
16 ... 'lli'c8, rather than 16 ... cs?! 17 lbds the position is level.
lbc7 18 l2Jxc7 'lli'xe7 19 exf s gxfs 11... ds 18 exds cxds
(19 ... lbxfs may be better, but even if After 18...lbexds 19 ii.c4 White
Black wins the e6-pawn White will al­ maintains pressure.
ways have light-square compensation) 19 ii.b2
20 g4 when White had the initiative in Black has a tremendous centre, but
D.Lima-A.Romero Holmes, Leon 1996. it is not easy to hold it together if
16...l2Jc7 White plays aggressively.

Th e M a r de/ Plata Varia tio n : 9 b4 l'i::i h s 1 0 .l:. e 1

stronger, since after 27...'lli'f7 28 c6!

l'i::i x a3 29 l'i::i x d5 l'i::i xd5 30 'lli'xf7 .l::!.xf7 31
.l:txd5 White is much better.
b2) 20 a4 a6 21 .l:tadl (the sacrifice
21 l'i::i xd5 does not work as well here:
21 ...l'i::ic xd5 22 l:!.ad1 'lli'c6 23 ilc4 l:!.fd8
24 ii.xe5 ii.xe5 2 5 .l::!.x e5 t'i::if6 26 .l:teel b5
27 'l!Vc3 l'i::ifd5 28 ii.xd5 .l:txd5 29 axb5
axb5 30 'lli'f6 'lli'e8 gave Black good
chances in Van Wely-Kotronias, Euro­
pean Championship, Warsaw 2005)
19 b6!?
... 21 ....l:td8 22 l'i::i x d5 .l:txd5 (worse is
Black seeks to break up White's 22 ...l'i::ic xd5 23 ii.c4 'lli'c 6 24 b5 axb5 2 5
queenside pawn chain. Other moves axb5 'l!Vxe6 26 ii.xe5 ii.xe5 27 f4) 23
are risky: ii.c4 .l:txdl 24 .l:txdl 'l!Vb8 25 l:!.d7 and
a) 19 ... 'lli'c8 20 t'i::ib 5! l'i::ixb5 (this is even though White has only a pawn, he
probably a mistake; instead 20 ... a6 21 has good compensation for the piece,
l'i::id6 'l!Vxe6 22 l'i::i xb7 was E.Bareev­ L.Van Wely-D.Stellwagen, Dutch Cham­
R.Polzin, Rethymnon 2003, and here pionship, Leeuwarden 2005.
22 ... e4! ? would be unclear, while b3) 2o l'i::ix d5!
20 ... a5! ? is another option) 21 ii.xb5
'l!Vxe6 22 ii.xe5! ii.xe5 23 f4 gives White
a comfortable advantage.
b) 19 ...'l!Ve8 is the most popular
move. Black removes his queen from
the d-file and keeps an eye on the b5-
square as well. Markos stops here,
pointing out that Black has scored well,
but the position is actually not so easy
for him. White has:
bl) 20 .l:tadl .l::!. d 8 21 a4 a6 22 b5
'it>h7 23 ii.a3 l'i::i xe6 24 bxa6 l'i::id4 This immediate sacrifice is more
(probably better is 24 ... bxa6 2 5 ii.xa6 dangerous than in Van Wely-Kotronias
l'i::id4 26 'l!Va2 'lli'd7 with an unclear posi­ above. Leaving the respective a-pawns
tion) 25 'l!Vxb7 .l::!.b8 26 'lli'e7 l'i::ic2 27 l'i::ib5 on their home squares helps White.
was E.Bareev-V.Topalov, Dortmund After 20 ...l'i::i c xd5 21 .l:tadl there is:
2002, and now 27 ... .l:tc8 is probably b31) 21 ...'l!Vc6 22 ii.c4 .l:tfd8 23 b5!
okay for Black. However, 27 ii.b5 looks 'l!Vxc5 24 ii.a3 ! is the difference because

A tt a c k i n g C h e s s : The King 's I n d i a n, Vo l u m e 1

the a-file is not open for Black's rook Worse is 21 ...'lli'd6 because of Mik­
and 24 ... 'l!Vc7 2 5 ii.xd5 seems better for halevski's suggestion 22 lbb 5! lbxb5 23
White, although I suspect Black can ii.xb5 d4 (23 ...'l!Vxe6 allows the familiar
hold after 25 ... .l::!.xd5 26 .l:txd5 l2Jxd5 27 24 ii.xe5 ii.xe5 25 f4) 24 a4 .l::!.fc8 25
'lli'xd5 'l!Vc3! 28 e7+ �h7 29 .l::!.g l 'l!Vxa3 ii.d7 .l:tc7 26 b5 with the idea of ii.a3
30 'lli'd8 .l::!.xd8 31 exd8'1li' 'l!Vxa2. when White has pressure.
b32) 21 ... �h7 22 ii.c4 l2Jf6 23 ii.xe5 22 lbxds
b5 24 ii.fl gave White good compensa­ After 22 lbb5 lbxe6 23 ii.c4!? dxc4
tion for the piece in L.Van Wely­ (Black should avoid 23 ...l2Jf4 24 ii.xd5+
V .Kotronias, European Team Champi­ l2Jfxd5 25 l2Jc3) 24 'l!Vxc4 'lli'c8 Mik­
onship, Gothenburg 2005. After 24... a6, halevski gives 25 lbc7 'l!Vxc7 26 'l!Vxe6+
Bologan suggests 25 �gl (instead of .l::!.f 7 27 .l::!.d 7 'lli'c6 28 .l:txe7 'l!Vxe6 29 .l:txe6
the game's 25 g3) 2S ....l::!.a7 26 a3 with .l:txa2 30 ii.xe5 ii.xe5 31 .l:t6xe5 �g7 32
some advantage. .l:te6 l:!.f6 33 .l:te7+ .l:tf7 with a draw.
20 cxb6 22 lbcxds 23 .l:txds lbxds 24 'lli'xds

Worse is 20 c6 �h7 21 .l::!.adl l2Jxc6 .:tds 2 s 'lli'b3 �h7 26

22 l2Jxd5 lbxe6 23 lbxb6 'l!Vxb6 24 'l!Vxe6 White's advanced d-pawn gives him
.l::!.fe8, which was drawn here in L.Van enough compensation for the ex­
Wely-F.Nijboer, Dutch Championship, change, but no more. Black soon sacri­
Hilversum 2006. However, instead of fices back the exchange to equalise.
24....l:tfe8 Black could play 24 ...'l!Vxb4 25 26 'lli' d 6 27 ii.d7 .l:!.a7 28 .l:td1 'lli' e 7 29

.l:t bl (25 'l!Vxc6 'l!Vxb2 leaves Black a a4 .l:taxd7 30 exd7

pawn up) 25 ... .l:tf6 26 'Ii'c4 'l!Vxc4 27 Here a draw was agreed in E.L'Ami­
ii.xc4 .l:td6 and despite White's bishop­ F.Nijboer, Dutch Championship, Hilver­
pair, the extra pawn gives Black reason sum 2008.
to play on.
20 ...axb6 21 .:tad1 'lli'b8 812) 13 ii.e3

The M a r de/ Plata Va r i a t i o n : 9 b4 lb h s 1 0 .l:. e 1

White covers the diagonal towards Instead 14 cs seems to ignore

his king by developing. Black's threat, but after 14 ...f4
13 ...ii.h6 (14 ...cxdSl? l S cxd6 cxd6 could be an
This is the main line. Black tries to improvement) l S cxd6 fxe3 16 dxe7
exploit the pin along the c1-h6 diago­ -.xe7 17 d6 -.ds 1S ii.c4+ �g7 19 lbe6+
nal and threatens .. .f4. Instead 13 ... h6 is ii.xe6 20 ii.xe6 �6 (or 20 ... lbh s l ? with
possible as well. After 14 lbe6 ii.xe6 lS the idea of ...lbf4 and ....l:tf6) 21 lbe2
dxe6 -.a (Markos does not mention .:tads it was unclear in E.Bareev­
this possibility, giving only 1s ...lbeS 16 J.Balcerak, German League 2001.
-.b3 lbc7 17 .l::!.adl when White has the 14...cxds 15 cxds
upper hand) 16 .l:!.b1 (16 �3 .l:tfdS 17 White has also tried lS exds. This
bS -.cs is equal according to Bologan) gives him access to the e4-square, but
16 ... .l:!.fdS 17 bS -.cs 1S bxc6 bxc6 19 it hands fs to Black. After 1s _.f4 16 ii.f2
-.a4 �h7 (19 ...-.xe6 and 19 ...f4 are also ii.xgs 17 hxgs lb hs 1s cs lLifs 19 lbe4
possible) 20 .l:tedl -.xe6 21 .l:tb7 .l:td7 22 (following 19 lbhg3 20 lbe4 lbxe4
-.a6, as in D.Komljenovic-A.Kuzmin, 21 .l:txe4 -.xg s 22 -.d2 the position is
Benasque 1999, then 22 ... hs 23 .l:txd7 unclear according to Mikhalevski)
-.xd7 24 ii.cs lbcS 2s ii.xd6 lbxd6 26 cs 19 ... lbfg3 20 ii.xg3 (after the 20 ii.c4
..tfS is equal (Bologan). lbxe4 21 .l:txe4 of M.Bosboom-F.Nijboer,
Wijk aan Zee 1997, Black should simply
play 21 ...-.xgs) 20 .. .fxg3 21 ii.c4 ii.fs 22
lbxd6 -.xgs 23 lbxfs .l:txfs 24 d6+ �g7
Black had a strong attack in
T.Kotanjian-V.Kotronias, Thessaloniki
1s ...ii.d1
Black will have to play this move at
some point, both to cover the bS­
square and to allow the e7-knight to
move off the queen's path to gs. Black
14 h4 has also tried the immediate 1 S .. .f4 16
Too tame is 14 ii.d2. After 14.. .f4 lS ii.f2 ii.xgs 17 hxgs lbhs 1s .l:tc1 and
lbh3 (1s lbe6 ii.xe6 16 dxe6 -.cs wins then :
the e6-pawn right away) 1s ...ii.xh3 (not a) 1S ... lbg3 19 -.d2 ii.d7 (no better
1s ... gS?l 16 lbf2 when White was better is 19 ...lbxe2+ 20 -.xe2 ii.d7 21 lbbs) 20
in E.Lobron-R.Barcenilla, Yerevan and White was better in
Olympiad 1996) 16 gxh3 Black has T.Radjabov-F.Nijboer, Wijk aan Zee
a good game. 2001.

A ttacking C h e s s : Th e King 's I n dian, V o l u m e 1

b) 18 ... ii.d7 19 ! gave White hxg s .l:ta4! was suddenly much better
some advantage in J.Berkvens-P.Smir­ for Black in S.Brynell-V.Kotronias,
nov, Patras 2001, but 19 lbbs ii.xbs 20 Stockholm 2006. Correct, though, is
ii.xbs l2Jc8 gave Black counterplay in 20...ii.xgs 21 hxgs l2Jxe4 22 l2Jc7 l2Jfs 23
A.Huzman-1.Caspi, Israeli League 2008. l2Jxa8 'l!Vxa8 with some compensation
for the exchange in an unclear posi­
16 f4

Instead 16 ... a6 is a sneaky move be­

cause Black will have to play this at
some point anyway. After 17 a4?! (a
waste of a tempo; it would be much
more testing to just bring a rook to cl)
11...f4 18 ii.f2 ii.xgs 19 hxgs lbh s 20 bs
l2Jc8 21 b6 'l!Vxgs 22 .l:!.ecl 'lli' d 8 23 'l!Ve1
lbg3 ! 24 ii.xg3 fxg3 2S 'l!Vxg3 'l!Vxb6+ 26
16 'lli' d2 'lli'f2 'l!Vas left Black a pawn up in
This move was recommended by L.Murzin-V.Kotronias, Linares 2003.
Markos and it is White's latest try. 11 ii.f2 ii.xgs 18 hxgs lbhs
White connects his rooks and there is
also a subtle point in lining up White's
queen against the gs-square, as we
shall see. The alternative is 16 .l:!.cl a6
17 bS and then:
a) 11 ...f4 18 ii.f2 ii.xgs 19 hxgs lbhs
20 'l!Vb3 as (not 20 ... l2Jc8 21 bxa6 bxa6
22 'l!Vb7) 21 b6 l2Jc8 22 .l:tf7 23
ii.xd7 .l:txd7 24 lbbs was S.lvanov­
H. Rudolf, Panormo 2001, and here
Black should take the pawn with
24 ... lbxb6, although White has a strong This is the typical structure for the
initiative after the simple 2S .l:tc2, in­ 13 ii.e3 line. Black will win the g s-pawn
tending to double rooks on the c-file. and have some kingside pressure, but
b) 11... ii.xbs! ? 18 ii.xbs axbs 19 White will obtain a strong initiative on
lbxbs fxe4 20 fxe4 l2Jxe4?! 21 'lli'g 4?! the other wing.
(White could play 21 lbxe4 ii.xe3+ 22 19 .l:tec1
.l:txe3 'lli'b6 23 'l!Vb3! lbfs 24 l:tcel with Instead 19 looks premature.
an obvious advantage) 21 ...ii.xgs 22 After 19 ... ii.xbs 20 lbxbs a6 21 l2Jc3

Th e M a r de/ Plata Varia tio n : 9 b4 lb h s 1 0 .l:. e 1

�h8 ( 21 . .lbc8 looks better) 22 .l::!. ecl Instead 24 .l:tcbl lbg3 (24 ...'lli'd8 2 S
lbg8 23 lbe2, as in l.lkonnikov­ 'l!Ve1 g s 26 l:i'.b8 .l:txb8 27 .l:txb8 was bet­
V.Kotronias, Gausdal 2008, Black ter for White in Xu Jun-S.Safin, Bled
should play 23 ....l:tf7 to cover C7, with Olympiad 2002, because the a6-pawn
the idea of ...'l!Vxgs with an unclear po­ is so weak) 2 S ii.xg3 'l!Vxg3 26 ii.d3 is
sition. unclear according to Markos.
19 ...a6 24 ...'lli'dB
Now this is necessary because after Black wants to play ... lbg3 and able
19 ... lbc8 both 20 lbbs and 20 look to meet ii.xg3 with .. .fxg3. Still, the
strong. immediate 24 ... lbg3 2S ii.xg3 'l!Vxg3
2o bs looks playable as well.
White consistently plays to open the 25 'l!Ve1 gs 26 lbd1 lbf6 27 J:tcc7 .l:tb8!?
queenside. Markos also mentions the Black gives back the pawn to relieve
strange 20 'l!Ve1?!, but this is playing on some of White's queenside pressure.
the wrong flank and Black gets good 27 ... ii.e8 was an alternative.
counterplay after his 20 ...lbc8 21 ii.h4 28 ii.xa6 .l::!. x b7 29 .l:txb7 g4
h6! 22 gxh6 gs 23 ii.f2 .l:tf6. Here Black had counterplay in L.Van
20 ... lbcB 21 bxa6 bxa6 22 .l:tab1 'l!Vxgs Wely-F.Nijboer, Dutch Championship,
23 .l:tb7 .l:tf7 Leeuwarden 2001.

82) 12 ...�hB

Now Black threatens ... ii.h3. Note

that Black has trouble creating direct
threats with his queen and knight be­ This is considered the most reliable
cause ...lbg3 will be met by ii.xg3 when move nowadays, mostly due to the ef­
Black cannot recapture with the pawn forts of Radjabov. Because 12 f3 is a bit
- this is one point of White's queen slow, Black makes a useful waiting
being on the d2-square. move without committing his pawn
24 ii.fl structure. White has tried many ideas

A ttacking Chess: Th e King 's I n dian, Vo l u m e 1

here but Black's defences continue to


B21: 13 .i.e3
B22: 13 lUe6
823: 13 cs
824: 13 .:.b1

White has also tried 13 bS and here:

a) 13 ... lbhs 14 ii.fl (14 l2Je6 ii.xe6 1 S
dxe6 l2Jf4 16 hf4 exf4 17 .:tel c 6 18
ii.d3 fxe4 19 ii.xe4 was about This has often been played, but is
equal in P.Ricardi-D.Lemos, Berazategui not very popular nowadays.
2007) 14 ...ii.f6 (14 ... l2Jf4!?) lS lbe6 13 ...ii.h6
ii.xe6 16 dxe6 lbg7 (16 ... l2Jf4!?) 17 ii.h6 Black plays along the lines of 'B12',
.l:te8 18 l2Jds l2Jg8 19 ii.xg7+ ii.xg7 20 cs but 13 ...lbhs ! ? is also interesting. After
(20 exfs looks like a better try) 20.. .fxe4 14 cs (perhaps White will look for im­
21 fxe4 dxcs 22 'l!Vb3 c6 23 l2Je3 'lli'e7 24 provements here) 14...l2Jf4 practice has
bxc6 bxc6 2S ii.c4 l2Jf6 and Black took seen:
over the initiative in F.Abbasov­ a) lS ii.c4 fxe4 (Wells gives the line
N.Mamedov, Baku 2009. lS ... lbexds! 16 lbxds 'l!Vxgs 17 hf4
b) 13 ... l2Je8 14 ii.e3 ii.f6 (after 14.. .f4 exf4 18 lbxe7 fxe4 19 l2Jxa8 ii.h3 20
lS ii.f2 lbxds 16 lbxds 'l!Vxgs 17 cs ii.fl exf3 21 'lli'x f3 ii.g4 22 'l!Vxb7 f3
White has some initiative for the when Black indeed has a strong attack)
pawn) lS l2Je6 ii.xe6 16 dxe6 lbg7 16 lbgxe4 lbfs 17 ii.f2 l2Jd4 has scored
(16 .. .f4 17 ii.f2 lbg7 18 cs is given as well for Black; one example being 18
unclear by Carlsen, although after .:tel ii.fs 19 �hl 'lli'd 7 when Black had a
18 ... l2Jxe6 19 cxd6 'lli'xd6 20 'lli'xd6 cxd6 reasonable position in A.Rychagov­
2 1 ii.c4 lbd4 22 .l::!. a cl .l:tfc8 Black can N.Pokazanjev, Novokuznetsk 2008.
hardly complain with the result of the b) lS .:tel fxe4 (Wells's idea works
opening) 17 ii.h6 l2Jxe6! 18 ii.xf8 'lli'xf8 here too, although not quite as well:
19 cs!? lbxcs (19... dxcs!? 20 'lli'd7 'lli'c8) lS ... lbexds!? 16 lbxds 'l!Vxgs 17 hf4
20 ii.c4 21 'l!Ve2 'l!Vh6 22 .l:tad1 .l:tf8 exf4 18 lbxe7 fxe4 19 l2Jxa8 ii.h3 20
gave Black sufficient compensation for ii.fl exf3 21 'lli'xf3 ii.g4 22 'l!Vxb7 f3 and
the exchange in M.Carlsen-V.lvanchuk, here, with White's rook on cl instead of
Foros 2008. al, he can hold: for example, 23 .l:tc2
f2+ 24 .l:txf2 ii.d4 2S �hl ii.xf2 26 .l::!.e 7
821) 13 ii.e 3 'l!Vh6 27 .l:tf7 dxcs 28 'l!Ve7 .l:txf7 29 'lli'xf7

Th e M a r de/ Plata Va riat io n : 9 b4 lb h s 1 o .l:.e 1

id4 30 'lli'e 8+ �g7 31 'l!Ve7+ with a squared bishop. Instead 16 cs makes a

draw) 16 lbgxe4 lLifs 17 ii.f2 lbd4 18 lot of sense because 16 ... lbeg8 is rather
ifl was slightly better for White in forced. Then 17 ii.g s h6 18 ii.h4 gs 19
E.Bareev-1.Cheparinov, Amsterdam ii.f2 .l:tg7 20 g4 was R. Pogorelov-1.Nataf,
2008. Reykjavik 2004, and here Black should
14 lbf7+ probably just continue with 20 ... hs.
The slow 14 ii.d2 should not give White could also play 16 g4 imme­
Black any trouble after 14.. .f4 lS lbe6 diately. After 16 ...lbeg8 17 hS
(1s h4 ii.xg s 16 hxgs lbh s) 1s ... ii.xe6 (Black can also play 17 ...h6 18 ii.h4 gs
16 dxe6 c6 with ideas like ...'lli'b 6+, 19 ii.f2 hS, heading into the main line)
... 'lli'c8 and maybe ... lbe8-c7. 18 gxhs (18 h3 hxg4 19 hxg4 .l:th7 20
Instead 14 h4 f4 lS ii.f2 ii.xg S 16 �g2 fbh6 21 :J:!.h l fbf7 22 J:txh 7+ �Xh7
hxg s lbh s 17 cs lbg8 looks like a better 23 ii.h4 �g7 should give Black enough
version of Line B12 for Black because play) 18 ... .l:th 7 we have:
... �h8 is a useful move as it clears g8 a) 19 hxg6 .:th S 20 h4 was given as
for his knight. After 18 'lli'd2 much better for White by Gofshtein,
(M.Thejkumar-K.Ramu, Hyderabad but Black has 20 ....i::!.x gs+ 21 hxgs lbxds
2006), Black can just play 18 ... 'l!Vxgs 22 �f2 lbxc3 23 .l:th1+ �g7 24 l:th7+
(18 ... ii.d7 and 18 ... a6 are reasonable �xg 6 2s 'l!Vh1 'l!Vxgs 26 .l:tg1 'l!Vxgl+ 27
too) 19 lbbs 'lli'd8 with the idea of �xgl lbxe2+ 28 �2 lbg 3. This looks
... lbg3 and ... 'lli'h4. quite risky, but there is no mate: 29
14....:txf7 15 ii.xh6 f4 'lli'h4 ii.e6 30 .l:txc7 lbf6 just looks very
b) 19 h4 .!:txhs 20 �2 was
Z.Gyimesi-P.Acs, Budapest 2004. Now
Black has 20 ....l:txh4! 21 ii.xh4 lbg4+ 22
fxg4 'lli'xh4+ 23 �g2 'lli'g 3+ 24 �1 wh en
he can take the draw or play 24...'l!Vxc3
with good compensation for the ex­
16 ...h6!? looks even stronger be­
cause 17 ii.xh6?! lbh7 18 h4 lbg8 19
White has obtained the bishop-pair, ii.g s lbxgs 20 hxgs 'l!Vxg s is nice for
but it has taken time and Black has a Black. Instead 17 ii.h4 gs 18 ii.f2 .l:tg7 is
clear plan on the kingside. an improved version of the main line
16 because Black has not spent a tempo
White wants to keep his dark- on ... lbeg8. Indeed, after 19 cs g4! 20

Attackin g C h e s s : Th e King 's I n dian, Vo l u m e 1

ii.h4 l'Llg6 21 ii.xf6 'lli'xf6 22 fxg4 h S ! Van Wely won the first one, but then
Black had good play i n A.Vlasov­ lost the next three, at which point he
V.Tukhvatullin, Sterlitamak 2008. headed towards other variations ...
11 g4 h6 1s ii.h4 gs 19 ii.f2 hs 20 h3 13...ii.xe6 14 dxe6

20...hxg4! 14 ...l'Llhs
This looks better than 20 ....:th7 21 cs This has been played the most, but
l'Lle7. In this position Black has lost two it may not be best. Others:
tempi compared to 16 ...h6 and after 22 a) 14.. .fxe4 lS fxe4 l'Llc6 is a typical
'it>g2 l'Llg6 23 .l:thl White was somewhat idea, but it does not fare well here: 16
better in B.Gelfand-T.Radjabov, Sochi l'LldS! l'Llxe4 (compared to the position
2008, although Black managed to win where .l:tbl and ... h6 are thrown in,
anyway. 16 ...l'Llg8 is not so good because after
21 hxg4 l'Llxg4 22 fxg4 f3! 2 3 ii.xf3 'lli'f6 17 bS l'Lld4 18 e7 l'Llxe7 White has 19
24 .l:tf1!! .l:te8 20 ii.g4 h6 21 ii.xe7 .l:txe7 22
After 24 'it>g2? ii.xg4! 2S J:te3 .l:taf8 l'Llxe7 'l!Vxe7 23 'lli'd 3 and Black does not
White's position falls apart - Mik­ have enough) 17 ii.f3 l'Llf6 18 bS! was
halevski. very good for White in L.Van Wely­
24...'l!Vxf3 25 'lli'xf3 .l:txf3 26 l'Llbs! ii.xg4 T.Radjabov, Khanty-Mansiysk 200S.
21 l'Llxc1 .l:tc8 28 l'Llbs .l:txc4 b) Van Wely suggests that 14 ....l::!.e 8!?
Black has equal chances in this is Black's best in New in Chess.
double-edged endgame.
(seefollowing diagram)
822) 13 t'Lie6
The most direct move. White does The game A.Grischuk-T.Radjabov,
not even wait for ...h6. Van Wely liked Elista 2008, continued lS l'Llds fxe4 16
this approach for a while, playing it in l'Llxf6 (after 16 fxe4 Black could try
four of his games against Radjabov. 16 ...l'Lleg8!?; the alternative is 16 ...l'Llxe4

The M a r de/ Plata Varia tio n : 9 b4 lb h s 1 0 .l:.e1

1 7 ii.f3 lbf6 when both 1 8 lbeg8 2 2 e 7 .l:txe7 23 lbxg6+ hxg6 2 4 .l:th4
and 18 lbxf6 ii.xf6 19 ii.xb7 .l:tb8 seem mate) 21 e7 .l::!.xe7 22 lbe6 'l!Vb8 (after
satisfactory for Black) 22 ...'lli'd7 23 lbxg7 .l:txe4 24 fxe4 lbe3 2S
'l!Ve2 'l!Vxg7 26 .:tel lbxc4 27 'l!Vxc4 dxcs
28 bxcs White is a little better accord­
ing to Mikhalevski) 23 lbxg7 .l:txg7 24
cxd6 (Mikhalevski suggests 24 1i.e6 !?
dS 2s .l:te2 'lli'f4 26 ii.xfs gxfs 27 'l!Ve1
.l:tag8 28 'l!Vc3) 24 ...lbxd6 2s .l::!. d 4 lLifs
(playing for a win; 2s ... lbxc4 26 .l:txc4 is
just equal) 26 J:td3 .l:te7 27 'lli'd2 'l!Ve S 28
.l:tdl .l::!. a e8 was level in L.Van Wely­
T.Radjabov, Monaco (blindfold) 2007,
although Black did indeed manage to
16 ... ii.xf6 17 fxe4 lbc6 18 ii.g4 lbd4 win.
19 1i.e3 c6 (Black should also consider b) 18 g3 lLih 3+ 19 'it>g2 lLigs 20 .:tg4
both 19 ... lbxe6 20 ii.xe6 .l:txe6 21 'lli'ds (Radjabov gives both 20 ii.xg s 'l!Vxgs 21
ii"c8 and 19 ...'l!Ve7 20 1i.xd4 exd4 21 .l:tfl lbds c6 22 cxd6 lbxd6 23 .l:tg4 'l!Vh6 24
:ts) 20 1i.xd4 exd4 21 J:tf1 .l:tf8 22 'lli'd3 e1 cxds 2s exf8'1li'+ .l:txf8 26 'lli' x ds lLifs
-W-e7 when Black was better, but the 27 .l:te4 lbe3+ 28 J:txe3 'l!Vxe3 29 .l:tdl b6
position was drawish. Indeed, after 23 and 20 e1 'l!Vxe1 21 lbds 'lli'd8 22 ii.xgs
:n 1i.g7 24 .l::!.af1 .l:txf 3 2s .l:txf3 .l:tf8 26 'l!Vxg s 23 lbxc7 lbe3+ 24 .l:txe3 'l!Vxe3 2S
:xf8+ 'lli'xf8 27 'lli'f3 hS 28 'lli'xf8+ hf8 lbxa8 e4! 26 f4 .l:txa8 27 cxd6 .l:td8 as
29 1i.e2 'itig 7 30 bS �6 31 bxc6 bxc6 32 unclear) 20...lbxf3 ! 21 'it>xf3 e4+ 22
.i.d3 'it>xe6 the game was drawn. This lbxe4 1i.xa1 23 lbe1+ 24 'it>g2
method of play looks like a simple solu­ was rather unclear in L.Van Wely­
tion to 13 lbe6, which will probably not T.Radjabov, Biel 2007.
be seen so much anymore. Neverthe­
less, we will still examine the sharp
14 ...lbh s.
15 g3
White has a serious alternative in
lS cs. After 1s ...lbf4 16 ii.c4 fxe4 17
.l:txe4 (17 fxe4 lbc6 with the idea of
... lbd4 is at least equal for Black)
11 ... lbfs and here:
a) 18 ii.xf4 exf4 19 lbds .l::!.e8 20
lbxf4 c6 (not 20 ... ii.xal? 21 'l!Vxal+ lbg 7

A ttacking C h e s s : Th e Kin g 's I n dian, Vo l u m e 1

1s ...ii.f6 L'Ami) 30....l:txe4! 31 'l!Vxe4 gxfs 32 'lli'd3

A typical idea. Black wants to play a6 33 a4 'lli'e6 34 ii.xg7 �xg7 3S gxfs
...lbg7xe6. 'l!Ves is level according to L'Ami.
16 c5 16 ...f4
This does not lead to much for
White. More interesting is 16 exfs lbxfs
17 g4 (else ... l2Jd4) 17 ... e4 18 l2Jxe4
ii.xa1 19 and here:
a) 19 ...ii.f6 20 l2Jxf6 l2Jxf6 21 gxfs
(Mikhalevski prefers 21 ii.d3, but after
21 ... l2Je7 22 'l!Va1 �g7 23 ii.xf6+ .l:txf6 24
gs 'lli'f8 the position is unclear) 21...gxfs
22 f4 .l:tg 8 23 'lli'd4 cs 24 'l!Va1 'lli'e7 2S
�2 .l:taf8 26 bxcs dxcs 27 ii.f3 .l:tg6 28
'l!Ves when White has excellent com­
pensation and went on to win in 17 �g2
V.Mamonovas-C.Gatto, correspondence Worse is 17 g4?! lbg7 18 ii.c4 lbc6
2007. 19 cxd6 cxd6 20 lbe2 (after 20 lbbs
b) 19 ..id4+ 20 �hl 'lli'e8 21 e1 lbxe1 lbxe6 21 'lli'xd6, as in T.Bakre-A.Kuzmin,
(21 ..Af7 22 gxfs 23 f4 l2Jxf4 24 f6 Dubai 2002, Black has 21 ...lbg s! 22
lbe6 2S ii.h4 lbd4 is messy, but looks 'lli'xd8 .:taxd8 23 l:tfl a6 24 l2Jc3 l2Jd4
better for White) 22 'lli'xd4+ lbg7 23 bS with the threats of ... .l:tc8 and ...l2Jgxf3)
l2Jg8 (L'Ami relates how at this point 20....l:!.c8 21 ii.ds lbxb4 22 .l::!.b1 l2Jc2 23
both players were playing very quickly!) .l:tfl b6 was much better for Black in
24 ii.d2 'lli'f7 (after 24 ... 'l!Ves White has L.Van Wely-T.Radjabov, Wijk aan Zee
the simple 2s 'l!Vxes dxes 26 ii.c3 or can 2007.
play 2S 'l!Ve3 .l:tae8 26 'l!Vxa7! .l::!.a8 27 17 lbc6 18 cxd6 cxd6 19 lbds

'l!Vxb7 .l::!.xa2 28 'lli'xe7 which looks very

strong) 2s ii.c3 h6 (2s ... .l:tae8 26 'l!Vxa7
b6 27 'l!Va3 h6 28 ii.b2 .l:ta8 29 'l!Vb3 �h7
30 ii.d1 .l:tae8 31 'l!Vc3 also looked good
for White in V.Mamonovas-G.Staf, cor­
respondence 2007) 26 f4 .l:!.ae8 27 .:tfl
l2Jf6 28 ii.f3 was 1.Cheparinov-T.Radja­
bov, Sofia 2008. Here 28...l2Jxe4 29 ii.xe4
�h7! 30 fs (not 30 ii.xb7?! lbe6 31 'l!Vxa7
lbxf4 32 'lli'd4 cs 33 bxc6 'lli'e7 34 �xf4
.l:Ixf4 3S 'lli'xf4 .l:tf8 with a strong attack -

Th e M a r de/ Plata Vari a t i o n : 9 b4 lb h s 1 0 .l::!. e 1

19 ...lbd4 16 exds lbfxds 11 lbxds 'lli'x ds!

19 ... lbg7 also looks good. Worse is 17...lbxd5 18 'l!Vb3 (both 18
20 ii.b2 lbxe6 21 g4 lbhg7 ii.b2 and 18 ii.c4 lbxb4 19 .l:tbl look
Chances were fairly balanced in this dangerous too) 18 ... c6 19 ii.c4 with
murky position in A.Shirov-T.Radjabov, pressure.
Wijk aan Zee 2001.

823) 13 cs

18 1'6b3
White's hopes for an edge have be­
come associated with this finesse. By
This obvious and direct move was threatening ii.c4 he induces Black to
considered harmless for a long time, exchange queens, after which the a-file
but lately White has been trying to will be opened for the al-rook.
squeeze something out of the end­ The immediate 18 'l!Vxd5 should also
game that arises. not be taken too lightly. After 18 ... lbxd5
13...h6 19 ii.c4 lbxb4 20 .l:tb1 lbc6! White has:
Black has also tried 13 ...dxc5 14 a) 21 J:txb7 lba5 22 .l:tb4 lbxc4 23
bxc5 h6 1 5 lbe6?! ii.xe6 16 dxe6 'lli'd4+! J:txc4 .!:tfe8 is no problem for Black.
17 'lli'xd4 exd4 18 lbb5 fxe4 with good b) 21 ii.d5 .!:tfd8 22 ii.xc6 bxc6 23
counterplay in Z.Gyimesi-T.Radjabov, .l:tb7 .l:te8 24 .l:txe7 .l:txe6 25 .l:td1 ii.f8 was
Moscow 2005, but Avrukh's suggestion fine for Black in G.Rechlis-B.Avrukh,
15 lbh3 ! fxe4 16 fxe4 ii.xh3 17 gxh3 Israeli League 2003. White has com­
indeed loo ks good for White. pensation for the pawn, but no more
Black could check out 13 ... a5!?. than that.
14 lbe6 ii.xe6 15 dxe6 ds c) 21 ii.e3 .l:tab8 22 .l:tbdl .!:tfd8 23 .l:td7
Black cannot sneak back into Gy­ .l:txd7 24 exd7 ii.f8 25 f4 .l:td8! (worse is
imesi-Radjabov with 15 ... dxc5, because 25 ... e4 26 .l:tbl!, A.Korobov-N.Huschen­
of 16 'l!Vxd8! (not 16 bxc5 'lli'd4+) beth, Pardubice 2008) 26 fxe5 .l:txd7 27 e6
16 ....l:tfxd8 17 bxc5 with a big plus. .l:td8 and if anything Black is better.

A t ta c k i ng C h e s s : Th e K i n g 's Indian, Vo l u m e 1

d) 21 e7! ? looks like White's best try,

intending 21 ....l:tfe8 (not 21 ...l2Jxe7 22
.l:txb7) 22 i..f7 .l::!. x e7 23 i..x g6.

20 .l:tfeB!

Radjabov's move looks like Black's

best road to equality. Black has also
Black has: pushed his a-pawn:
dl) 23 ... e4 is considered Black's best a) 20 ... as?! is loosening. Instead of
by Avrukh, but Black will suffer here: 24 21 bS?! l2Jd4 22 i..c4 .l:tfe8 23 i..e 3?! l2Jc2
i..xfs!? (24 fxe4 fxe4 25 .l:txe4 i.. d4+ 26 24 J:te2 l2Jxe3 25 .l:txe3 e4 26 i..d s .l::!.ad8
.l:txd4 l2Jxd4 27 i..b 2 .l:td8 28 .l:tdl �g7 27 fxe4 fxe4 28 i..xe4 .l:!.xe6 when Black
29 i..c2 .l:te2 30 .l:txd4 .l:txc2 is given as was better in Z.Gyimesi-A.Volokitin,
equal by Avrukh) 24 ...exf3 25 .l:txe7 German League 2007, the simple 21
lbxe7 26 i..e4 i.. d4+ 27 �fl fxg2+ 28 bxas! just looks much better for White.
�xg2 .l::!. g 8+ 29 �h3 b6 30 cxb6 cxb6 31 b) 20... a6 is a little bit better: 21
i..x h6 should be a draw, but White can i..b2 (21 bS axbs 22 i..xbs .l::!.fe8 23
certainly play on for a while with his J:txa8 .l:txa8 24 i..b2 .l:te8 25 i..xe 5 i..x es
bishop-pair. 26 .l:txes �g7 27 J:tds .l:te7 should hold)
d2) 23 ... .l:tf8 24 .l:txb7 is given as 21 ....l:tae8 22 i..c4 .l:te7 23 i..d s .l:td8? 24
slightly better for White by Avrukh , but i..xc6 bxc6 25 i..xe s i..x es 26 .l:!.xes with
24... e4 gives Black counterplay. a healthy extra pawn for White in
d3) 23 ... l2Jd4 also looks okay after 24 Z.Gyimesi-E.Hagara, Jenbach 2009. A
.l:txb7 .l:te6 or 24 i..b 2 e4!. better try is 23 ...l2Jd8 24 .:ta2 (24 i..x es
18 'lli'x b3 19 axb3 lbc6 20 .l:ta4
... i..x es 25 .l:txes c6 26 i..c4 �g7) 24 ... c6
White protects the b4-pawn and 25 i..c4 lbxe6 26 i..xe 5 l2Jc7 27 i..x g7+
may double on the a-file. Instead 20 .l:txg7, as given by Golubev, but White is .l::!.fe8 21 i..xc6 bxc6 22 bS .l:txe6 somewhat better here as well.
(22 ...cxbs 23 .l:ta6 gives White compen­ 21 i..c4
sation) 23 b6 cxb6 24 cxb6 cs 25 bxa7 This does not lead to much, but
.l:te7 is equal according to Giri. White has trouble getting anywhere:

1 38
Th e M a r de/ Pla ta Va ria tio n : 9 b4 lb h s 1 0 .l:. e 1

a) 21 ii.b2 .l:txe6 22 bS lbd4 23 ii.c4 so he activates his rook.

J:tee8 24 ii.f7 .l:tf8 25 ii.xg6 lbxb3 26 2s ....l:txe6 26 .l:td7 ii.f8! 27 .l:txc7 ii.xcs+
ii.xes lbxcs is level. 28 �fl .l:te7 29 .l:txe7
b) 21 ii.e3 has been used success­ Perhaps simpler was 29 .l:tc8+ �g7
fully by Arun Prasad. Now Black has: 30 ii.xc6 bxc6 31 J:txc6 with equality.
bl) 21 ...lbd8 22 .:teal lbxe6 23 .l::!.x a7 29 ... ii.xe7 30 ii.b2
.l:tab8 24 ii.c4 (or 24 .l:tdl lbd4 25 ii.c4) After 30 ii.xh6 gS! White will have
24.. .f4 2 5 ii.d2 e4 26 fxe4 ii.xal 27 J:txal to give up a pawn to save his bishop.
.l:tbd8 28 ii.c3+ �h7 29 h4 and White 30...lbd4 31 ii.xd4 exd4
had good compensation in 5.Arun Despite Black's extra pawn, the
Prasad-A.Lahiri, Gurgaon 2010. endgame is a draw, A.Grischuk­
b2) 21 ... a6 22 bS axbs 23 .l:txa8 .l:txa8 T.Radjabov, Linares 2009.
24 ii.xbs .l:te8 25 .l:td1 .l::!.xe6 26 .l:td7 was
clearly better for White in 5.Arun 824) 13 l:tb1
Prasad-V. Rajesh, Chennai 2010, as
26 ....l:te7? loses to 27 ii.xc6.
b3) 21 ... .l:txe6 is the simplest. Now
22 ii.c4 J:tee8 23 bS lbd4 transposes to
variation 'c'.
c) 21 bS lbd4 22 ii.c4 lbxe6 23 ii.e3
lbd4 and now:
cl) 24 ii.xd4 exd4 25 J:txe8+ .l:txe8 26
.l::!.xa7 d3 27 ii.xd3 ii.d4+ 28 �fl ii.xcs
29 l:txb7 ii.b6 and with White's rook
trapped on b7, Black cannot be worse.
c2) 24 ii.f7 lbxb S ! ? (Black could also This flexible move has become
play 24.. 1H8 25 ii.xd4 exd4 26 ii.xg6 White's latest attempt to get some­
d3) 25 ii.xe8 (25 ii.xg6 J:tf8 is fine for thing from the position. The rook gets
Black) 2s ....l:txe8 26 ii.d4 .l:te7 27 ii.b2 c6 off the long diagonal and takes a far­
28 .l:taal �g8 29 .l::!.adl �f7 and Black sighted look at the b-file which, as we
had two pawns for the exchange in have seen, can open up in some varia­
M.Kozakov-R.Vidonyak, Lvov 2009. tions. This line was also recommended
21... a6! 22 bs axbs 23 .l:txa8 .l::!. xa8 24 by Markos in Beat the KID. Moreover,
ii.xbs .l::!.e8! after a few reversals with 12 lbe6, Van
Black finally rounds up the e6- Wely turned to this variation to beat
pawn. Radjabov. He was not going to aban­
2 s .l::!. d 1 don the Bayonet and he praised Rad­
25 ii.c4 lbas ! does not help White, jabov for sticking to his guns as well,

A ttacking C h e s s : Th e King 's I n dian, V o l u m e 1

expressing that he would rather 'lose a position favourably for White after
game than lose face'. It will certainly be both 16 ... tllxfs 17 il.d3 and 16 ... gxfs 17
interesting to see what their future g3, intending f4 when Black is weak on
battles bring. the light-squares.
13 ... h6 16 fxe4
Forcing the pace. After 13 ... tllh s 14 Exchanging pieces helps Black. After
cs tllf4 lS il.c4 we have transposed to 16 tllxe4 tllxe4 17 fxe4 tllc6 Black is fine
Bareev's 12 f3 tllh s 13 �bl! which was because White has no knight coming to
discussed at the beginning of Line B. ds.
The main alternative is 13 ... c6 14 16...tllc6 17 tlld s
il.e3 and now:
a) 14 ... il.h6 lS h4 cxds 16 cxds f4 17
il.f2 tllh s 18 'lli'd2 il.xgs 19 hxgs is very
similar to Line B21. As Markos points
out, the additional moves �bl and
... 'iii>h 8 help both sides to some degree,
because White will often open the b­
file with b4-bS (Black usually has to
play ... a6 to prevent a white piece from
hopping to bS), while ...'iii>h 8 could be
useful too because Black can play a
quick ... tll g 8 to open his queen's path 17 tllg S!

towards capturing the gs-pawn. Black cannot play 17 ...tlld4? 18 e7

b) Black can also play 14 ... h6 lS tlle6 and 17...tllx e4? 18 'lli'd3 tllf6 19 'l!Vxg6
il.xe6 16 dxe6, but this should compare must also be avoided.
unfavourably for him to the line 12 ... c6 18 il.d3
13 il.e3 h6, because of the extra moves Opening a path for the white queen.
�bl (which White plays anyway) and Other moves do not lead to much:
...'iii>h8 (which Black may or may not a) 18 l:tfl �xfl+ 19 il.xfl tlld 4 20
play). However, maybe it is not so bad 'lli'g 4 c6 (with the idea of ... 'lli'f6) virtu­
after 16 ...tll e 8 17 bS tll c 7 18 bxc6 bxc6 ally forces White to sacrifice a piece
19 �b7 tllx e6 20 'lli' a4 (not 20 il.xa7?? with 21 'l!Vxg6 cxds 22 exds, but this
�xa7 0-1, S.Kojima-G.Timoshenko, looks insufficient after 22 ...'l!Vh4!.
Cappelle la Grande 2009) 20 ...tll d4 with b) 18 cs tlld4 19 il.c4 tllxe6 20 'lli'g4
an unclear position in R.Markus­ 'lli'e 8 21 cxd6 cxd6 (with the idea ... tlld4)
Y.Dembo, Budapest 2002. 22 ! ? 'l!Vxbs 23 'l!Vxe6 'l!Vd3 24 il.e3
14 tlle6 il.xe6 15 dxe6 fxe4 tllf6 2S �bdl 'lli'c 4! is also satisfactory
Instead lS ... tllhs 16 exfs! opens the for Black - Markos.

The M a r de/ Plata Va ria tio n : 9 b4 ll'i h s 1 0 .U. e 1

c ) 18 'lli'd3 ll'id4 19 'l!Vh3 .U.e8! 20 20 h4

�g4!? (an interesting pawn sacrifice; This move was Van Wely's choice in
instead 20 e7 ll'ixe2+ 21 l:txe2 'lli'c8 22 his defeat of Radjabov. There are some
il.e3 ll'ixe7 23 'l!Vxc8 .U.axc8 24 ll'ixe7 othertries, though:
:.Xe7 2S il.xa7 is equal according to a) 20 il.e3 c6 21 .U.fl! ? is speculative,
Markos) 20 ... c6 21 ll'ic3 'lli'e7 22 bS ll'ixe6 but White was . also successful after
(similar is 22 ... 'iii>h7 23 bxc6 bxc6 24 ll'ie2 21 ... cxds 22 exds ll'if6 23 'l!Vh3 e4? 24
!Lixe6 2S il.xe6 'l!Vxe6 26 'l!Vxe6 .U.xe6, il.xd4 exd3 2S 'lli'xd3 'lli'e7 26 cs in
which has led to draws in some corre­ S.Arun Prasad-J.Ashwin, Kavala 2009.
spondence games) 23 il.xe6 'l!Vxe6 24 The calm 23 ...'iii> g 8 would have been
'l!Vxe6 .U.xe6 2S bxc6 bxc6 26 il.e3 l:te7 27 more testing, but the idea is interest­
.U.edl �d7 gives White compensation for ing.
the pawn, although probably no advan­ b) 20 bS hopes to open the b-file in
tage. Then 28 a4 ll'if6 29 as �c8 30 a6 the event of ... c6 by Black, but Markos
dS!? (30 ... �ce7 is solid, but obviously points out that Black can play 20 ... 'lli'c 8!
rather passive) 31 exds cxds 32 ll'ixds 21 h4 'l!Vxe6 22 'lli'hs 'lli'f7 23 'lli'xf7 .U.xf7
was A.Korobov-K.Maslak, Pardubice 24 hxgs hxgs 2 s il.xgs il.h6 with good
2010, and now 32 ... ll'ig4! would keep the prospects.
game within the bounds of equality. c) 20 'l!Vh3 was Ponomariov's choice
18 ... ll'id4 19 'lli'g4 gs when this position first occurred and
This move is forced. 19 ... 'iii>h 7?! 20 this move is considered best by Markos.
�e3 ll'if6 21 'l!Vh3 with the idea of e6-e7 After 20 ... c6 White can try:
is a problem, and 19 ... c6?! 20 'l!Vxg6! is cl) 21 e7 ll'ixe7 22 il.xgs cxds 23
now very strong: for example, 20...'lli'e 8 il.xh6 is a sharp piece sacrifice, but
(20 ...cxds 21 exds ll'if6 22 il.xh6 �g8 23 Markos illustrates how Black can neu­
�gs is crushing) 21 'l!Vxe8 �axe8 22 tralize it: 23 ... 'lli'c 8! (not 23 ... 'iii>g8 24
!Lie3 �xe6 23 bS gives a good endgame. il.xg7 'iii>x g7 2s 'lli'g4+ ll'ig6 26 exds �f6
27 l:tfl with a strong initiative) 24 'lli'h4
dxc4! (again holding on to the piece is
dangerous after 24 ... 'lli'd7 2 S il.e3+ 'iii>g 8
26 il.xd4 exd4 27 exds) 2 s il.xg7+ 'iii>xg7
26 'l!Vxe7+ �f7 27 'l!Vgs+ 'iii>h8 28 'l!Vh6+
�h7 29 'lli'xd6 ifa 30 'lli'xe7 �xe7 and
Black has excellent compensation for
the pawn in the endgame.
c2) 21 ll'ie3 'lli'f6 and now:
c21) 22 ll'ifs ll'ixe6 23 �fl ll'if4 is al­
ready better for Black.

A ttacking C h e s s : Th e King 's I n d i a n, Vo l u m e 1

c22) 22 tllg4 'l!Ve7! 23 il.e3 (better is continuing initiative after 24 il.xg7+

23 tll e 3 'l!Vxe6 24 'l!Vxe6 tll xe6 2s tllfs 'iii>x g7 2S 'lli'g4+ tllh gS 26 l:tfl OT 26 tll e 3)
.U.ad8 26 il.e3 with some compensation 24 'l!Vh3 il.xh6 2S 'lli'xh6 c6 26 J:te3 ! with
for the pawn) 23 ... tllxe6 24 g3 'lli'd7! ? a winning attack in L.Van Wely­
(perhaps even better is 24...tll d 4!? 2 S T.Radjabov, Wijk aan Zee 2009.
�fl 'lli' e 6) 2S �fl was R.Ponomariov­ 22 hxgs tllx ds
T.Radjabov, Wijk aan Zee 2003. Here Worse is 22 ...tllxg S 23 il.xg S hxgs 24
the simplest is 2s ... tll d4 with the idea 'l!Vxgs when White has all the play.
of 26 il.xg s? tll f 3+. 23 gxh6
c23) 22 �fl 'l!Vxe6 23 �xf8 (Markos After 23 exds tllxg s 24 'lli'g4 'iii>g8
suggests 23 tllf s tllx fs 24 exfs 'lli'f6 2s (24... �f6 is also possible) 2S il.e3 'lli'c8
il.b2 dS 26 bS with compensation for 26 'lli'h4 'lli'd8 27 'lli'g4 'lli'c8 28 'lli'h4 'lli'd8
the pawn) 23 ...�xf8 24 'l!Vxe6 tllxe6 2S 29 'lli'g4 the game was drawn in
tllfs .U.d8 26 il.e3 il.f8 27 g3 tll g 7 28 K.Landa-A.Shomoev, Ulan Ude 2009.
tllx g7 'iii>xg7 29 bS cs 30 �b3 il.e7 31 23 ... il.f6
.U.a3 .U.a8 32 'iii>f2 and White had enough This is fine, but 23 ...tlldf4 24 hxg7+
compensation for the pawn in tll x g7 2S 'l!Vh2+ 'iii>g 8 26 hf4 �xf4 27 cs
A.Klimov-A.Shomoev, Kemerovo 2007. .U.h4 28 'lli'g 3 'lli'f6 also looks okay for
20...tllf6 21 'lli'g3 Black.
After 21 'l!Vh 3 g4 Black keeps the 24 exds il.h4
kingside closed.

An important resource. The position

21...tllxe6! remains very complicated.
This is much better than 21 ... gxh4?! 25 'lli' h 3
22 'lli'xh4 tllxe6 23 il.xh6 'iii>g 8? (this After 2S 'lli' g 6? il.f2+ 26 Wfl il.e3+ 27
loses, although after 23 ...tll h7 White il.fs l:txfs+! 28 'lli'xfs tll d 4 29 'lli'f7 'lli'f8
has a choice between an endgame Black wins, while 2S 'lli'g4 il.xel 26 dxe6
edge, with 24 'l!Vxd8 �axd8 2S il.e3, or a (26 il.e3 'lli'h4 27 'lli'x h4 il.xh4 28 dxe6

Th e M a r de/ Plata Va ria t i o n : 9 b4 ll'i h s 1 0 .U. e 1

:ae8 is much better for Black) 28 il.d2 l:tae8 29 .U.f1 -.xe6 30 -.xe6
26 leads to: �xe6 31 .il.e4
a) 27 �1 il.h4+ 28 'iii>e 2 (28 'iii> g l
..if2+ is a draw) 28.. ltf2+ 29 'iii> d l -.e7
intending ....U.g8 is unclear.
b) 27 'iii>h 2 .U.g8 28 -.e4 -.e7!
(28 ... �4+ 29 -.xh4 il.xh4 was given as
much better for Black by Velickovic and
Sasa in the Inform ant, but White has
two dangerous passed pawns and
looks to be doing well after 30 'iii>h 3 ! ) 29
i.d2 .U.af8 with counterplay.
25 ... il.xel 26 dxe6 il.f2+ 27 'iii>h 1
White looks to have sufficient com­
pensation for the exchange, especially
if the kingside pawns get moving, but
Black finds a clever resource to take
over the initiative.
31 ...1If4!?
Instead 3 1 .. ltef6 32 g4 il.g3 33 �xf6
.U.xf6 34 is not so easy for Black.
32 il.xf4 exf4 33 il.f5
Better was 33 J:txf2 l:txf2 34 g4!.
33 ...1Ixh6+ 34 iLh3 il.g3
21 ...-.f6I Black had a big advantage in Shen
This is more ambitious than the sim­ Yang-Ju Wenjun, Xinghua Jiangsu
plifying 27 ...°ii'h4. 2009.

Chapter 7
The Mar del Plata Variation
White's other Ninth Moves

1 d4 tllf6 2 c4 g6 3 tllc3 il.g7 4 e4 d6 5

tllf3 o-o 6 il.e2 es 7 o-o tllc6 8 ds tlle7 A: 9 �gS
B: 9 .il.d2
C: 9 34
D: 9 �h1

There are a few other moves to con­

sider, although they are rather experi­
a) 9 'l!Vc2 looks a bit pointless, but it
cannot be too bad: 9 ... tllh s (9 ... tlld7 and
9 ... tlle8 are possible as well) 10 g3 (10
b4 is Epishin's line considered at the
In this chapter we look at moves beginning of Line B in Chapter 5) 10 ...fs
other than 9 tlle l, 9 tlld2 and 9 b4. The 11 tllgs was P.Eljanov-M.Al Sayed,
first of these, 9, seems illogical, Khanty-Mansiysk 2009. Here 11 ...tllf6
but it has had its bouts of popularity. 9 looks sensible and best.
il.d2 was once a main line, but it seem s b) 9 il.e3 rather invites 9 ...tllg4 (oth­
too passive for modern tastes. Finally, erwise White will play 10 tlld2), but it is
both 9 a4 and 9 'iii>h l look a bit indul­ not so bad: 10 (after 10 il.d2 fS 11
gent, but they should not be underes­ tllg s tllf6 the position is the same as the
timated. Although these various lines main line of the Bayonet, but White has
are not considered to be theoretically played il.d2 rather than b4 and .U.e1, and
dangerous, they all have their ideas after 12 exfs tllxfs! Black is fine) 10 .. .f6
and Black should still be well prepared. (worse is 10 ... h6 11 il.h4 fS 12 tlld2 tllf6

Th e M a r de/ Pla ta Va ria t i o n : White 's O t h e r N i n t h M o ves

13 f3 gs 14 il.f2 f4 lS cs with the initia­ following 12 .. ..i:!xfs 13 il.g4 or 12 ...gxfs

tive) 11 il.h4 (11 il.d2 fS is the same as 13 f4) 10 g3 (another point to Black's
10 il.d2) and now after 11 ... gs 12 il.g3 9th move is that it covers the Cl-pawn,
�h6 we reach a line of the Gligoric so 10 il.e3 can be met by 10 ... tllx dS!)
where White has castled rather prema­ and now, rather than 10...ilh3 11 �el
turely - see note 'c' to White's 12th fs 12 exfs tllxfs 13 tllf3 ! ? h6 14 tlle4
move in Line Bl of Chapter 9. when the h3-bishop was a bit mis­
c) 9 �bl is undoubtedly playable, placed in G.Hertneck-W.Watson, Ger­
but it should not be too scary for Black man League 1994, I would prefer 10.. .fs
theoretically. 11 exfs tllxfs 12 tllf3 tllf6 13 il.d3 h6
with an equal position in C.Bauer­
F.Libiszewski, Montpellier 2004.

A) 9 iLgS

When faced with a move like 9 �bl

Black must mainly rely on his under­
standing of the Mar del Plata in gen­
eral, as he did in E.Lobron-L.Brunner,
German League 1990: 9 ... tllh s (this is This move was brought to promi­
very logical; instead 9 ... tlld7 10 b4 as 11 nence by the creative Ukrainian Grand­
a3 'iii>h 8 12 il.e3 ! tllg 8 13 cs! axb4 14 master Vereslav Eingom and was
axb4 fS lS tllg s favoured White in favoured by the Swiss Grandmaster
l.Nikolaidis-A.Jankovic, Leros 2010) 10 Yannick Pelletier for a while. It does not
tt:Je1 tllf4 11 tlld 3 tllx e2+ 12 'l!Vxe2 fS seem logical for White to trade off his
and there was a more or less normal­ dark-squared bishop, but if Black forces
looking position similar to some found the exchange then White hopes to gain
in Line A, below. time for a queen side assault.
d) 9 tllh4 looks quite extravagant, 9 tll h s

but it is playable too: 9 ... tlle 8 (after Black has a decent alternative in the
9 ... tlld7 10 g3 fs 11 exfs tllxfs 12 tllxfs popular 9 ... h6. After 10 il.xf6 il.xf6 11
we can see some point to White's idea b4 Black has:

A ttacking C h e s s : Th e Kin g 's I n dian, Vo l u m e 1

il.xb7 or the untried 14... as!? with


a) 11 ...c6!? was recommended by

Bologan. We have seen this pre­
emptive strike before and after the 12 10 tlle1 (12 a4 as! 13 bxas cs is comfort­ White attacks the hS-knight and
able for Black) 12 ... as 13 a3 il.g7 14 prepares to bring his own knight to d3.
'l!Vb3 cxds l S exds!? b6 16 �fdl axb4 17 Other moves are less dangerous:
axb4 il.d7 18 bS of Y.Pelletier­ a) 10 .U.el h6 11 il.d2 tllf4 12 il.xf4
V.Bologan, Cap d'Agde 2002, Bologan exf4 gives Black a position similar to
suggests 18 ... g S ! ?. those in the Bayonet with 10 cs, but
b) 11 has been the most here Black is better placed. After 13
common choice: 12 cs (12 tlld2 fS �d2 gs 14 h3 tllg 6 lS tllh 2 tlle s 16
transposes) 12 .. .fs 13 tlld2 fxe4 (this is .U.acl a6 Black had no problems in
usually played, but it scores badly; still, J.Gonzalez Zamora-V.Bologan, Turin
13 .. .f4? is bad because of 14 il.g4 and Olympiad 2006.
other moves like 13 ... 'iii>h 7 look silly) 14 b) 10 g3 h6! (this move will prove
tlldxe4 tllfs looks like a better version useful because it controls gS; instead
of the 9 tlld 2 variation for White after 10 .. .f6 11 il.d2 fS transposes to a line
lS a4, 1 S �Cl OT 1 S il.g4. which can arise from 9 il.d2 too, but
c) 11 ...'iii> g 7 has always looked stur­ Black should avoid it as White has 12
diest to me: 12 cs tllg 8 13 a4 (13 cxd6?! tllg S!) 11 il.d2 (there is not much point
cxd6 releases the tension prematurely now in exchanging the bishop with 11
and 13 tlld 2 il.e7 14 c6 bxc6 lS dxc6 fS il.xe7?! 'l!Vxe7) and here:
gives Black good play) 13 ... il.e7 14 c6 bl) 11 ... il.h3 12 .U.el fS 13 tllh4 tllf6
(after 14 cxd6 both 14 ... il.xd6! ? and 14 exfs g s ! l S tllg 6 tllx g6 16 fxg6 il.fs
14...cxd6 lS as fS 16 tlld2 tllf6 give 17 il.e3 'lli'e8 18 cs a6 19 'l!Vxg6 20
Black counterplay), and now Black can cxd6 cxd6 was fine for Black in l.Bem­
play 14 ... bxc6 lS dxc6 fS, 14 .. .fS l S cxb7 W.Watson, Gausdal 1991.

Th e M a r de/ Plata Varia tio n : White 's O t h e r N i n t h M o v e s

b2) 11 .. .fs 12 tllh 4!? (after 12 exfs with variation 'c' in the note to Black's
Black should play 12 ... tllxfs, because 13 next move) 14... g S ! (14.. .f4 lS il.f2 g s
g4? fails to 13 ...tlld4 14 gxhs tll x e2+ lS 16 cs tll g6 i s also possible) lS c s tll g 6
'l!Vxe2 il.g4) 12...tllf6 13 exfs g s 14 tll g2 16 .U.fcl �f7 17 .U.C2 (better is 17 exfs
(14 tll g 6 tllx g6 lS fxg6 il.fs is similar to il.xfs according to Pelletier) 17 ... tllf4 18
variation 'bl', but White's rook is not 'lli'd2 tllxd3 19 'lli'xd3 g4 20 fxg4 f4 21
on el) 14 ... tllxfs l S il.d3 tlld4 was level il.f2 il.xg4 22 cxd6 cxd6 23 .U.acl a6 24
in M.Krasenkow-W.Watson, Cappelle la tl:Ja4 'l!Vg s 2s tllb6 .U.af8 and Black had
Grande 1990. an excellent attacking position along
10 tllf4 11 tll d3 tllxe2+
... with the bishop-pair in Y.Pelletier­
Black secures the bishop-pair. Prac­ A.Fedorov, Plovdiv 2003.
tice has show that White keeps some 13 gs

edge after 11 ... tllx d3 with both 12 Black clamps down on the kingside.
'l!Vxd3 and 12 il.xd3. The natural 13 ...fs leads to sharp play,
12 'l!Vxe2 but is risky after 14 f4! (instead 14 f3 is
well met by 14.. .f4 or 14...c6!?) 14... exf4
1s tllxf4 gs.

12 h6

Again this move will prove useful.

Instead 12 ...f6 13 il.d2 (less incisive is Here White has:
13 il.e3 fS 14 f3 f4) 13 .. .fs 14 f4! com­ a) 16 tlld3 f4 is fine for Black, as 17
pares favourably (for White!) with the es dxes 18 tllx es fails to 18....U.e8!.
note to Black's 13th move, below, be­ b ) 16 tlle6 il.xe6 17 dxe6 f4 (intend­
cause Black does not have control of gs. ing ... tllg6) 18 g3 tll g6 (but not 18.. .fxg3
13 il.d2 19 hxg3 tll g 6 20 tll d s! c6 21 'l!Vh s with
Instead 13 il.xe7 'l!Vxe7 is harmless, the initiative in A.Yermolinsky­
while 13 il.e3 fS 14 f3 (after 14 f4 exf4 V.Topalov, Wijk aan Zee 1999) 19 gxf4
1s tllxf4 g s 16 tllh s 17 exfs tllxfs il.xc3! 20 bxc3 gxf4 21 'iii>h l 'lli'f6 gave
Black hits the e3-bishop - compare Black promising counterplay in A.Blees-

A ttacking Chess: Th e Kin g 's I n d i a n, Vo l u m e 1

H.Klarenbeek, Heraklio 1993. c2) 1 9. ..fxg3 1 9 �xf8+ 'lli'xf8 20 tllb s!

c) 16 tllh s ! il.d4+ (worse is (after 20 �fl iLh3 ! 21 �xf8+ !Ixf8 Black
17 exfs tllxfs when, instead of 18 g4 will win back the queen with ...�f2, as
tllg 7 19 tllx g7 'iii>x g7 20 �xf8 'lli'xf8, as in the game G.Prakash-P.Konguvel,
seen in D.Berczes-L.Valdes, Budapest Nagpur 1999) 20 ...'lli'f2 21 'lli'xf2 il.xf2
2006, White can play 18 tlle 4! il.xb2 19 22 tllx c7! (worse is 22 hxg3 il.b6 23 il.c3
il.c3 il.xc3 20 tllx c3 with a strong initia­ a6 24 tlld4 'iii>f7 2S 'iii> g 2 'iii> g 6 26 �hl
tive for the pawn) 17 'iii>h l f4 sees Black il.d7 when Black was at least equal in
threatening to consolidate with ...tllg 6. G.Prakash-P.Konguvel, Chennai 2000)
22 ... il.h3 23 hxg3 �c8 was P. Lukacs­
V.Rajlich, Budapest 2001, and now 24
tllb s �xc4 2S 'iii>h 2! would give White a
comfortable plus.
14 g4
White blocks the kingside. Instead
14 f3 fS gives Black counterplay, while
14 h4?! g4 lS f4 gxf3 16 'lli'xf3 fS gave
him the initiative in L.Oll-A.Shirov, Til­
burg 1992.
14...tll g6
Thus White must act quickly with The knight heads for f4.
18 g3! and here: 15 f3 tllf4 16 tllxf4

cl) 18 ... il.h 3 19 gxf4! looks promis­ White could also try 16 il.xf4!? exf4
ing: 19 20 �xfl tllg 6 21 es! (after 17 .U.acl with the idea of opening the c­
21 'lli'd3 Black should avoid 21 22 file, but Black's strong bishops should
es 'lli'e8 23 e6 as in M.Galyas-S.Husari, certainly hold the balance.
Balatonlelle 2003, and play! 16...exf4
22 il.xc3 tllxf4 23 tllxf4 �xf4 24 �xf4
gxf4 2s 'lli'd4 'l!Vgs 26 'lli'h8+ 'iii>f7 when
White does not have more than a per­
petual check with 27 'lli'h 7+ because 27
'l!Vxa8? gets mated after 27.. .f3)
21...dxes 22 'lli' d 3! tllxf4 (White wins
after 22 ...'iii>h 7 23 fxg s hxgs 24 �fS!
!Ixfs 2s 'lli'xfs with a decisive attack) 23
l:.xf4 (not 23 il.xf4 'lli'e 8!, turning the
tables) 23 ...exf4 24 'lli'xd4 'l!Ve7 2S tlle4
and White has a huge advantage.

The M a r de/ Plata Va ria t i o n : White 's O t h e r N i n th M o ves

Black is already quite comfortable days, but it is not completely harmless.

with the bishop-pair and th e possibility White prepares to bring his rook to cl
of opening the h-file with ... hs. and remains very flexible.
17 �fdl 9 ...tlle8
More solid is 17 tlldl with the idea Black can also play 9 ...tlld7 when 10
of exch anging Black's dark-squared �cl fS 11 tll g s tllf6 leads us back to the
bishop immediately. However, after main line. However, White has an extra
17 ...c6 18 il.c3 cxds 19 il.xg7 'iii> x g7 20 option in 10 b4 fS 11 tllgs tllf6 12 f3, so
c:xds il.d7 21 tllf2 �c8 22 'Ii' d2 'lli'b6 9 ... tll e8 seems more accurate.
Black was still very comfortable in L.Oll­ Instead 9 ... tllh s was for a long time
S.Dolmatov, Rostov 1993. considered the main line, but after 10
11 hs! 18 h3 19 il.e1
... g3 fS 11 tllg s ! (more common is 11 exfs
Safer was 19 'iii>g2 with the idea of tllxfs 12 tlle4, but Black is fine after
defending with �hl at some point. 12 ...tllf6) 11 ...tllf6 12 f3 White has
19. 'iii>g7
.. scored well. This position is similar to
Strengthening the long diagonal the Bayonet with 12 f3, but here White
with 19 ...'lli'f6 was also possible. has played il.d2 instead of b4. This may
20 J:td3?! �h8 21 tlld 1 hxg4 22 fxg4 not favour him, but it could well intro­
After 22 hxg4 both 22 ...'lli'f6 and duce a new world of complications that
22 ... �h3 with the idea of ... 'lli'h 8 are very is not really worth delving into consid­
good for Black. ering how rare 9 il.d2 is nowadays.
22 iff6 23 il.c3 il.d7 24 tllf2 �ae8
... 1o �c1 fs
Black had a clear advantage in
V.Pelletier-T.Radjabov, Biel 2006.

B) 9 il.d2

After 10 tll el fS 11 tlld3 tllf6 we

have transposed back to Chapter 1. In­
stead 10 b4 fS 11 'lli'b 3 (11 tll g s makes
less sense here because the e6-square
This old move is rarely seen nowa- is defended) 11 ...tllf6 12 exfs should be

A ttacking Chess: Th e King 's I n dian, Vo l u m e 1

met by 12 ... tllx fs! when 13 tllg s? runs tlle6 Si.xe6 13 dxe6 'Ii'c8 14 'l!Vb3 c6, as
into 13 ...tlld4. in L.Ftacnik-Kr.Georgiev, Groningen
11 tll gs 1976. Here lS cs!? would give White
Others: the initiative.
a) 11 exfs gxfs (11 ...tllxfs! ?) 12 tll g s After the text, again we have a posi­
h6 13 tlle6 il.xe6 14 dxe6 'lli'c8 lS 'l!Vb3 tion resembling the Bayonet. Here
c6 16 il.hs 'l!Vxe6 17 'l!Vxb7 tllf6 18 il.e2 White has played the rather slow il.d2
�fb8 19 'lli'a6 .U.xb2 gave Black good and l:tcl, though, instead of b4 and
play in M.Taimanov-R.Fischer, 1st l:tel.
matchgame, Vancouver 1971. 12 f3
b) 11 'l!Vb3 b6!? 12 exfs gxfs 13 tll g s 12 exfs should be met by 12 ... tllxfs!,
tl\f6 (after 13 ...h6 14 tll e6 Si.xe6 lS dxe6 heading for the d4-square.
'lli'c8 16 tllds 'l!Vxe6 17 tllxe7+ 'l!Vxe7 18 12 ...c6 13 'l!Vb3 h6 14 tll e6 il.xe6 15
cs+ 'iii>h8 19 cxd6 tllxd6 20 .U.c6 White dxe6 'lli'c8 16 .il.e3 '1li'xe6
had decent compensation for the pawn This is fine, but Black has decent al­
in M.Taimanov-Ma.Tseitlin, USSR 1973) ternatives in 16 ... .U.d8 and 16 .. .f4.
14 f4 h6 1s fxes dxes 16 cs tllfxds 17 17 l:tfdl
tllxds tll x ds 18 cxb6 axb6 19 .U.c6 'iii>h8 Instead 17 'l!Vxb7 is met by 17 ... .U.fb8
20 tl\f3? (better is 20 'lli' h3 when Black 18 'lli'a6 �xb2 with an unclear position.
can play either 20.. J:tf6! ? or 20 ... tllf6 21 11 ... 'iii>h7 18 'l!Vxb7 .U.fb8 19 'lli'a6
il.c3 f4! ? 22 'lli'h4 il.b7) 20 ... il.b7 21 �g6
tl\f4 22 Si.xf4 exf4 23 .U.dl 'lli'e7 gave
Black a big advantage in M.Taimanov­
R.Fischer, 3rd matchgame, Vancouver

19 .. .fxe4!
An important move. Here 19 .. ..i:!xb2
is bad because of 20 'l!Va3! which also
hits the d6-pawn. Likewise, 19 ... hS?! 20
cs ds 21 exds tllfxds 22 tllx ds tllxds 23
il.c4 gave White the upper hand in
This looks better than 11 ... h6 12 J.Plachetka-V.Babula, Stare Mesto 1992.

Th e M a r de/ Pla ta Varia tio n : Wh i t e's O t h e r N i n t h M o ves

2o fxe4 11 ... fs 12 tll g s tllf6 13 cs!, but 11 ...b6

If 20 tllxe4 tllxe4 21 fxe4 then and 11 ...tllcs are sensible alternatives)
21 ... �xb2 hits the bishop on e2. 12 b4 axb4 13 il.xb4 fS was M.Carlsen­
20...�xb2 21 'l!Va3 �xe2 22 tllxe2 A.Morozevich, Biel 2006, and Black was
Not 22 �xd6 .U.xg2+! 23 'iii>xg2 'lli'g4+ a tempo up on a line of the Bayonet (9
24 'iii>f2 �f8 when Black has a strong b4 as 10 il.a3 axb4 11 il.xb4 b6).
attack. 10...tll d 7 11 tll d 3
22 ...tllxe4 This is more dangerous than 11 il.e3
With two pawn for the exchange fS 12 f3 when Black can keep the ten­
and squares for his pieces, Black has a sion with 12 ...tllc s!? (instead 12 .. .f4 13
good position. il.f2 gs leads to the main lines of Chap­
ter 3) 13 tlld3 b6 and now:
C) 9 a4 a) 14 b4 tllxd3 lS 'lli'xd3 axb4 16
tllb s 'iii>h8 17 'l!Vb3 (after 17 il.d2 Black
can consider 17 .. .fxe4 18 fxe4 .U.xfl+ 19
il.xfl tl:Jg8 20 il.xb4 tl:Jf6 21 aS il.h6 OT
17... cs!? 18 dxc6 tll x c6) 17 ... tll g 8 18
'l!Vxb4 and here 18 ... tllf6 19 exfs gxfs 20 h6 21 il.h4 il.d7 22 .U.a3 'l!Vb8 23
il.f2 .U.g8 24 �bl tllh s 2s 'iii>h l 'l!Vd8 gave
Black counterplay in V.Korchnoi­
G.Kasparov, Barcelona 1989, but later
Kasparov preferred 18 .. .fxe4! 19 fxe4
�xfl+ 20 �xfl il.h6 21 il.xh6 tllxh6 22
This move is rare but is hardly bad. 'l!Vd2 'iii> g 7 when Black is certainly bet­
Moreover, it has certain similarities ter.
with and may even transpose to 9 tll e l b) 14 �a3 ! ? 'iii>h 8 lS 'lli'd2 f4 16 il.f2
tt:ld7 10 il.e3 fs 11 f3 f4 12 il.f2 gs 13 a4 h S (after 16 ... g s Black must consider 17
(Line E of Chapter 3). g4! ?) 17 tllb s gs 18 b4 axb4 19 'l!Vxb4 g4
9 .as
.. 20 �fal tll g 6 21 'lli'el (or 21 as �xas 22
This is a sound reply. Black can also �xas bxas 23 �xas? gxf3 24 gxf3 tllx d3
play 9 ...tllh s as he would against 9 b4. 2S il.xd3 iLh3 when White cannot de­
White chooses between 10 as and 10 fend against 26 ... 'l!Vgs+), and now
g3 with rather unexplored play. 21 ...tllb7! ? was unusual but held up
10 tlle1 White long enough on the queenside
Instead 10 b3 looks slow, but White for Black to develop good kingside play
can still play this way: 10 ...tlld 7 in A.Evdokimov-D.Jakovenko, Russian
(10...tllh s ! ?) 11 il.a3 il.h6 (worse is Team Championship 2010.

A ttacking C h e s s : Th e Kin g 's I n dian, Vo l u m e 1

11...fs 12 f3 b3, il.a3, b4 and il.xb4, rather than

After 12 exfs tllxfs 13 tlle4 tllf6 il.e3-f2-e1, b4 and il.xb4.
Black is fine. 13 14 il.a3 b6
Instead 14...tllf 6?! allows the imme­
diate lS cs, as in S.Atalik-V.Bologan,
Manila Olympiad 1992.
15 b4 axb4 16 il.xb4 tllc s 17 as

12 ...f4
This is rather committal, but it is
thematic and logical. Instead 12 ...b6 13
il.d2 tllf6 14 b4 axb4 lS tllxb4 il.d7 16
tlld3 cs!? 17 �a3 �f7 was unclear in This is a critical position. Black is too
A.Evdokimov-L.Nisipeanu, European far behind to just play for mate:
Championship, Budva 2009. 17 ... il.d7?! 18 tllb s gave White the ini­
Another possibility is 12 ...'iii> h8 13 tiative in V.Popov-F.Amonatov, Sochi
il.e3 b6 14 b4 (if 14 'l!Vb1 il.a6! ? then lS 2007, and 17 ... il.a6 18 axb6 (better
b3 tllg 8 is equal according to Bologan than 18 tllbs c6!) 18 ... cxb6 19 tllx cs
and lS tll b s can be met by the reply bxcs 20 'lli'd7 21 'lli'a4 looks better
1 s ... tllf6, pressuring e4) 14 ... axb4 l S for White as well.
tllxb4 tllf6 16 tlld3 tlleg8!? 17 tllf2 tllh s Black's best looks like 17 ...c6!? 18
18 �el il.h6 19 il.xh6 tllx h6 20 iLfl f4 axb6 (instead 18 tll a4 tllxd3 19 'lli'xd3
21 tllb s g s gave Black good counterplay bxas 20 il.c3 cxds 21 cxds il.a6!? 22
in S.Khmelevsky-G.Beckhuis, Vienna 'lli'd2 il.xe2 23 'l!Vxe2 'lli'e7 24 �fbl �fb8
2006. was fine for Black in A.Klimov­
13 b3! E.Bondarenko, Tomsk 2008) 18 ... tll x d3!
This is the point of White's play - he (but not 18 ....U.xal? 19 'l!Vxal tllxd3 20
will be able to open the queenside. This!) 19 �xa8 (19 il.xd3? loses to
may look slow, but it is the right plan 19 ...�xal 20 'l!Vxal 'l!Vxb6+) 19 ... 'l!Vxb6+
and if we compare it to Line E of Chap­ 20 cs! tllxcs 21 'lli'b7 22 �xc8 .U.xc8
ter 3, White is actually getting his 23 dxc6 tllxc6, which looks okay for
bishop to b4 more quickly by playing him.

Th e M a r de/ Pla ta Va ria t i o n : Wh ite 's O t h e r N i n t h M o ves

D ) 9 �h1

A few possibilities:
Tony Miles's old favourite looks bl) 13 a3 ! ? looks promising. White
pointless, but it is not completely with­ will play b4 next unless Black sacrifices
out merit. a pawn with 13 ...a4 14 il.xcs dxcs lS
9 tlle 8!
... tllxa4.
One point of White's play is that b2) 13 l:tcl is a bit slow as the rook
9 ... tllh s can be met by 10 tllg 1 tllf4 11 does not have much purpose on this
l.f3, although i t is not clear that this is square:
so dangerous. Instead 9...�h8 can be
met by 10 a4 or 10 tll e l, when it is
hardly clear which player the king
moves will favour.
As I found out for myself, 9 ... tlld7 is
less accurate than 9 ...tlle 8, though, be­
cause it allows 10 g4!?. With this move
White discourages Black from opening
the kingside, after which he may tum
his attention back to the queen side.
a) 10...fs 11 gxfs gxfs 12 exfs
(White can use both the e4-square and 13 ... il.d7 14 b3 tllg 8 lS tlld2 f6
the g-file) 12 ...tllf6 13 �gl tllxfs 14 il.d3 (Black intends ... il.h6) 16 h4 fS! (now
�h8 lS tllg s left White with an edge in that White has weakened his kingside,
R.Kaufman-D.Vigorito, Chicago 2008. Black tries to open things up) 17 gs
b) 10...�h8 11 �gl! as 12 il.e3 tll cs tllxe4 18 tlldxe4 fxe4 19 tllxe4 tl\e7 20
gives White an active version of some il.g4 tllfs gave Black good counterplay
similar plans we have seen with the g4- in M.Carlsen-T.Radjabov, Biel 2006.
advance. b3) 13 'lli'd2 (with the idea of �g3

A ttacking Chess: Th e King 's I n d i a n, Vo l u m e 1

and .U.agl) 13 ...tll g 8 14 l:tg3 b6 lS l:tag l king move looks out of place here.
tllf6 16 'l!Vc2 il.d7 17 tlld2 c6 18 f3 gave White can also play 12 il.d3 tllf6 13
White an edge in M.Tosic-A.Kolev, tllf3 which transposes to the main line.
Vrnjacka Banja 1990. Instead 12 tllc2 tllf6 13 g4? just weak­
b4) 13 tlle l fS 14 f3 fxe4 lS fxe4 c6 ens the kingside too much. After
16 tlld3 cxds 17 cxds b6 18 tllxcs bxcs 13 ...tll d4 14 h3 tllx g4! ? lS hxg4 'lli'h4+
19 �d2 �b8 20 !Iafl and again White 16 'iii>g l tllx e2+ 17 'l!Vxe2 il.xg4 18 f3 a
could claim some advantage in A.Miles­ draw was agreed in A.Miles­
F.Gheorghiu, Cran s Montana 2001. D.Gormally, British Championship,
10 tlle1 Hove 1997, but Black should really have
Instead White can also play 10 a4 as played on with 18 ... il.xf3! 19 .U.xf3
11 tlle l fS 12 tll d3 tllf6 13 f3 b6, which 'lli'g4+ 20 'l!Vg2 'lli'xf3 21 'lli'xf3 .U.xf3 22
is similar to Line C, although here 'iii>g 2 l:taf8 when he is certainly better.
White has played 'iii>h l rather prema­ Here a clearer refutation is available
turely. even sooner with 14 ... tllx c2! lS 'l!Vxc2
It looks a bit late for 10 l:tgl because il.xg4 16 il.xg4 (White gets mated after
Black can play 10...fs 11 exfs when 16 hxg4 tllxg4 17 il.xg4 'lli'h4+ 18 'iii>g2
11 ...gxfs 12 tll g s tllf6 has been played 'l!Vxg4+ 19 'iii>h2 .U.f3) 16 ... tllx g4 with a
with some success, but I prefer winning position.
11...tllxfS ! when White's kingside shuf­ 12 ...tllf6 13 il.d3 tlld4 14 tllxd4 exd4 15
fling looks rather silly. tlle4 tll xe4 16 il.xe4
10 ... fs 11 exfs tllxfs!
Again I prefer activating the knight
rather than allow 11 ...gxfs 12 f4!?.

16...'lli' h 41
This is more aggressive than
16... il.d7 17 'l!Vd3 cs 18 dxc6 bxc6 19
12 tllf3 il.d2 .U.b8, which was also satisfactory
Covering the d4-square, but this is for Black in A.Miles-A.Beliavsky, Nova
obviously time-consuming. White's Gorica 1999.

1 54
Th e M a r de/ Plata Varia tio n : White 's O t h e r N i n t h M o ves

17 'l!Vc2 il.fs! This is forced, because 24 'lli'fl loses

Black develops very quickly. to 24 ... il.xf2+ 25 'lli'xf2 .U.el+.
18 il.xfs �xfs 19 'iii>g1
This is a bit humbling. White 'takes
back' his 9th move. However, he has
trouble developing his queenside and a
good alternative is hard to find, be­
cause 19 il.d2 loses to 19 ...�xf2.
19 �es

Black has a fantastic position. All of

his major pieces are active, he has a
passed d-pawn and White has no ac­
tive possibilities.
Now after 24 ... 'lli'xh3 ? White man­
aged to hold on with 25 il.f4 il.xf2+ 26
'iii>xf2 'lli'h2+ 27 'iii>t3 'l!Vh5+ 28 'iii>g 2 .U.e2+
29 'iii>fl .U.xb2 30 'lli'e4! (this covers the
hl-square and prepares to check the
black king) 30 ... 'lli'h 3+ 31 'iii> el �g2 32
'lli'e8+ 'iii>g 7 3 3 'lli'e7+ 'iii> g 8 34 'lli'e8+ and
V2-V2 in A.Miles-H.Tirard, Cappelle la
Grande 2000.
However, Black could have crowned
his excellent play with 24 ...'lli'f6! 25 'lli'fl
20 a4 (or 25 il.f4 il.xf2+ 26 'iii>xf2 'l!Vxb2+)
Black could also play 20...c5!?. 25 ... 'lli'f3 !, paralyzing White and threat­
21 h3 d3!? ening ...�e2: for example, 26 il.d2 �e2
This forces the pace. Again 21...c5 27 il.el 'l!Vxg3+ 28 'l!Vg2 (28 'iii>h l il.xf2
was possible too. 29 'l!Vxe2 'l!Vg1 mate) 28 ... .U.xel+ 29 J:txel
22 'l!Vxd3 .U.xf2 23 �xf2 il.d4 24 g3 il.xf2+.

Chapter 8
The Mar del Plata Va riation

White's Eighth Move Deviations

1 d4 tl:Jf6 2 C4 g6 3 tl:Jc3 J.g7 4 e4 d6 5 cause it i s not easy to create counter­

tl:Jf3 0-0 6 J.e2 es 7 0-0 tllc6 play unless one is familiar with certain

A: 8 dxes
B: 8 J.e3 .

Note th at 8 J.g 5 is not very good

because of 8... tllxd4 9 tllxd4 exd4 10
'lli'xd4 tllxe4! 11 J.xd8 J.xd4 12 J.xc7
J.xc3 ! (this is better than 12 ... tllxc3 13
bxc3 J.xc3 14 �abl which is just equal)
13 bxc3 J.e6 14 J.f3 l:tac8! when Black
Usually when White plays 7 o-o he is much better.
is willing to enter into a sharp fight.
However, that is not always the case A) S dxes
and in this chapter we look at moves White plays a kind of delayed ex­
other than 8 dS. It seems that with the change variation. This should not trou­
moves 8 dxes and 8 J.e3 White is ble Black, but it is not as innocent as
changing his mind about what kind of one might think and Black should not
game he wants and these variations go to sleep just yet.
are certainly less popular than 7 dxes 8 dxes 9

and 7 J.e3. While the lines considered Instead 9 'lli'xd8 tllxd8 is solid, but
here are not terribly dangerous, they Black can also just play 9...�xd8 10
should not be ignored, especially be- l:tf8! (White may keep a pull after

Th e M a r de/ Plata Varia tio n : W h i t e's Eig h t h M o ve D e viatio n s

1 1...l:te8 1 2 tlld s tllx ds 1 3 cxds tlld4 14

..'!Jxd4 exd4 lS il.d3 or 11 ...�d7 12
.i..d l!) 11 �fdl il.g4 which actually
transposes back into the main line.
9 'lli'xdl

Black has an interesting alternative

in 9 ... il.g4!? which immediately fights
for the d4-square. Then 10 'lli' x d8 and:
a) 10...�fxd8 11 tlld s tll x ds 12 exds
f6 13 dxc6 fxgs 14 cxb7 l:tab8 lS h3
.i..fs was l.Rausis-1.Nataf, Enghien les
Bains 1997, and here 16 cs! would give 11 h3
White some initiative. Black is certainly not troubled by 11
b) 10 ... �axd8 11 il.xf6 (better is 11 �acl h6 (11 ... tlld4!?) 12 il.e3 .U.fd8, but
..'!::l d s tllxe4 12 il.xd8 �xd8 13 �fdl 11 �d3 is interesting. Black could try
when Black probably lacks sufficient 11 ... hf3 12 il.xf3 tlld4 13 tlld s tllxds
compensation for the exchange) 14 cxds fS or even the immediate
11 ... il.xf6 12 tlld s il.xf3 13 il.xf3 il.g s 11 ... tllxe4!? 12 tllxe4 fS, but practice
was fine for Black in R.Ponomariov­ has focused on 11 ... h6 12 il.e3 tllxe4!
V.Topalov, Vitoria Gasteiz 2007; 14 (White is a little better after 12 ...�fd8
..'!::lxe7 l:td2 gives Black enough activity 13 l:tadl �xd3 14 �xd3) 13 tllxe4 fS 14
for the pawn. tll c s e4 lS �b3 and now:
c) 10...tllxd8 11 il.xf6 il.xf6 12 tlld s a) 1S ...f4 16 il.d4 (Black is certainly
.i..xf3 13 gxf3 ! (after 13 tllxf6+?! 'iii> g 7 14 okay after 16 tllxe4 fxe3 17 l:txe3 il.xf3
.i..xf3 'iii>xf6, with the idea of ...tllc 6-d4, 18 il.xf3 il.xb2 19 l:tbl il.d4) 16 ... tllxd4
Black is already better, while 13 il.xf3 17 tllxd4 il.xd4 18 il.xg4 il.xcs 19 �xb7
�g s 14 tllx c7 l:tc8 is certainly not worse il.d6! (but not 19 ... il.b6 20 b4!) is equal
for the second player) 13 ... il.g s 14 tllxe7 but unbalanced.
.:.c8 lS tll d s tlle6 gives Black some b) 1S ... exf3 16 gxf3 f4 17 fxg4 fxe3
compensation, but it is a bit specula­ 18 fxe3 tlld4!? 19 exd4 il.xd4+ 20 'iii>hl
tive - White's extra pawn is 'real' on il.xcs 21 �fl?! �xfl+ 22 il.xfl b6 was
the queenside. drawish but still a little bit better for
10 .U.fxd1 Black in B.ltkis-M.Golubev, Sovata 2000.
White can also play 10 �axdl, but However, 21 �xb7 il.b6 22 cs! is a bet­
then the fl-rook would be a bit stuck. ter try, so I would prefer to deviate ear­
10 il.g4
... lier with 18 ... tlla s 19 �bs b6 or just
Instead 10 ... h6 11 il.e3 il.g4 looks 18 ... �ae8. In both cases Black has good
fine too. compensation for a pawn.

A tt a c k i ng Chess: Th e King 's Indian, Vo l u m e 1 12 il.xf3 tlld4 13 tlld s tllx ds the lines of the Gligoric Variation with
14 cxds the sharper 8 ... tllg4.

Even though the d4-knight looks

nice, White has the bishop-pair and 81: 8...�e8
potential pressure on the c-file, so 82: 8 tllg4

Black must act quickly.

14 ...fs 81) 8 ...l:te8
Worse is 14...c6 1 5 dxc6 bxc6 16
.U.acl �fb8 17 b3 as 18 l:tc4 when White
has some pressure.
15 .U.acl .U.f7 16 il.e3
After 16 'iii>fl Black should avoid
16 ...f4?! 17 il.g4! and play 16 ... il.f8 or
16 ...l:taf8.
16 ...fxe4 17 il.g4
Wisely avoiding 17 il.xe4? tlle 2+ and
17 il.xd4?! exf3 18 il.e3 fxg 2. After 17
il.g4 tllfs 18 l:tc3 hs 19 il.e2 il.f8 20 l:tc4
tlld6 21 l:tc3 tllfs 22 l:tc4 tlld6 23 l:tc3 This move has long been known to
tllfs the game was drawn in l. Krush­ equalize.
A.Melekhina, US Women's Champion­ 9 dxes
ship, Saint Louis 2009. The point of Black's rook move is
seen after 9 dS tlld 4! 10 tllxd4 exd4 11
8) 8 .il.e3 il.xd4 tllxe4 12 il.xg 7 'iii>x g7 13 tll xe4 (or
White tries to keep the tension in 13 'lli'd4+ 'lli'f6) 13 ...�xe4 when the posi­
the centre. Black has the traditional tion is equal, although Black does have
equalizer 8 ... �e8 or he can play along slightly the better long-term chances

The M a r de/ Plata Varia tio n : W h i te's Eig h t h M o ve D e viations

because of his better bishop. l.Smirin, Tilburg 1992) 14 ... .U.b8 lS
9 dxes
... it.xf3 16 'lli'xf3 tlld7 (with the idea of
...tlld 7-e6) 17 b4 as 18 bS (18 a3 axb4
19 axb4 tllf8) 18 ...tllx cs 19 bxc6 tlle6
gave Black counterplay in L.Oll-1.Smirin,
Rostov on Don 1993.
b) 10 h3 prevents, but costs
White time. After 10 ... it.e6 11 cs Black

1o 'lli'xd8
This is not very ambitious, of course,
but other moves do not give White
much either:
a) 10 cs it.g4! (fighting for the d4-
square) 11 .tbs 'l!Vc8 (Smirin's idea;
11 ... tlld7 12 it.xc6 bxc6 13 h3 it.xf3 14
·tvxf3 'l!Vb8 is another option) and now: bl) 11 ...tllh s 12 tllg s (12 .tbs ! ?)
al) 12 h3 �d8 (it is probably better 12 ...tllf4 13 tllxe6 tllxe6 14 .tbs .U.f8!?
to just play 12! 13 it.xc6 bxc6, (14...'lli'xdl lS �fxdl .U.ed8 16 it.xc6
transposing to variation 'a2') 13 tlld s bxc6 is level) lS it.xc6 bxc6 16 'lli'a4 'lli'e 8
!Dxds 14 exds it.e6 (instead 17 .U.adl fS 18 exfs (after 18 f3 f4 19
lS 'lli'xf3 tlld4 16 it.xd4 exd4 17 �fel it.f2 gs Black has counterplay on the
·tvfs 18 'lli'xfs gxfs 19 it.c4 it.f8 20 b4 as kingside as well) 18 ...gxfs 19 f3 'iii>h 8 20
21 a3 d3 is possible, but Black is really 'iii>h l .U.g8 21 .U.d2 it.f6 and Black had an
just playing to draw here) lS dxe6! active position in P.Van der Sterren­
:xdl 16 exf7+ 'iii>xf7 17 it.c4+ �8 18 B.Gelfand, Biel lnterzonal 1993.
:taxdl gave White good compensation b2) 11...a6 kind of mimics White's
for the queen in P.Lukacs-A.Stummer, 10th move by preventing any .tbs
Budapest 1992. ideas: 12 'lli'a4 'lli'e7 13 .U.fdl .U.ad8 14
a2) 12 it.xc6 bxc6 13 h3 it.hs 14 it.c4 tlld4! lS tllxd4 (Black has no prob­
(otherwise, there is 14 g4? tll x g4 and 14 lems after lS it.xe6 tllxf3+ 16 gxf3
'l!Ve2 h6 1s 'iii>h2 gs 16 'lli'c4 �b8 17 b3 'l!Vxe6 17 'iii>g 2 c6) 1s ... exd4 16 �xd4 was
as 18 tlld2 g4 19 f3 gxh3 20 gxh3 .U.b4 1.Jelen-M.Tratar, Slovenian Team
was unclear in P.Van der Sterren- Championship 1993. Here Black should

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

play 16 ... ll'ld7! 17 ll'lds ii.xds 18 .l:f.xds Line A) 11 ... ii.g4 12 ll'lds ll'lxe4 13 ll'lxe7
c6 19 l:tdd1 ll'lxcs 20 'ti'b4 ii.f8 with a .l:f.xd1+ 14 l:txd1 .l:f.c8 1s ll'lds ll'lf6 16
good position. ll'lxf6+ ii.xf6 17 h3 ii.e6 18 b3 was
b3) 11 ... 'ti'e7!? 12 (this seems M.Dlugy-A.Fishbein, New York 1991,
logical as Black has omitted ... a6, but and here 18 ...l:td8 19 .l:f.xd8+ ii.xd8 looks
Black still obtains satisfactory play; in­ fairly level.
stead 12 'ti'c2 .l:f.ad8 13 .l:f.ad1 was 11 ll'lbs
J.Speelman-J.Nunn, Hastings 1987/88, After 11 ll'lds ll'le6 12 ll'lgs ll'lf4! ?
and now Black should play 13 ... a6 ac­ (12 ... ll'lxds 13 cxds ll'ld4 is also okay,
cording to Speelman) 12 ....l:f.ed8 13 'ti'a4 but not 13 ... ll'lxgs?! 14 ii.xgs with an
ll'ld4! 14 ll'lxd4 (14 ll'lxes a6! 1s ii.xd4 edge for White) 13 ii.xf4?! ll'lxds 14
axbs 16 't\Vb4 ll'lxe4! with the idea of 17 ii.d2 ll'lf6 1s f3 ll'lhs 16 .l:f.fd1 ll'lf4 11
ll'lxe4 .l:f.a4! is good for Black) 14... exd4 ii.fl ii.f6 18 ll'lh3 ii.xh3 19 gxh3 ll'le6
1S ii.xd4 c6 16 ii.e2 ll'lxe4! 17 ii.xg7 Black was winning in Z.Mijailovic­
ll'lxcs 18 't\Va3 ..t>xg7 (Black has won a V.Kotronias, Vmjacka Banja 2006.
pawn, but he will have to untangle his 11 ... ll'le6 12 ll'lgs l:te7 13 .l:f.fd1
pieces) 19 ll'le4 b6 20 .l:f.fc1 ii.xh 3 ! ? Alternatively:
(tempting, but 20....l:f.dS is simpler and a) 13 ll'lxe6 ii.xe6 14 f3 c6 (14 ...b6!?)
preparing for the opening of the b-file 1S ll'lc3 (1S ll'lxa7 looks risky but could
with 20....l:f.ab8!? is also possible) 21 be more testing) 1s ... .l:f.d7 16 .l:f.fd1 ii.f8
'ti'c3+ f6 22 ll'lxcs ii.xg2! 23 .l:f.e1! ii.ds was fine for Black in S.Reshevsky­
and Black had good compensation for R.Fischer, Santa Monica 1966.
the piece in D.Berczes-A.Jankovic, Sara­ b) 13 ll'lxa7 ll'lf4 (or 13 ...ll'ld4 14
jevo 2010. ii.xd4 exd4 1S ll'lxc8 .l:f.xc8 16 f3 ll'ld7
10...ll'lxdS with compensation) 14 ii.xf4 (14 ii.f3
ii.d7 1S ll'lbs h6 is pleasant for Black)
14 ...exf4 1S ll'lxc8 .l:f.xc8 16 f3 ll'ld7 gives
Black reasonable compensation for the
13 ... b6
13 ... c6 is also possible.
14 a4
Instead 14 ll'lxe6 ii.xe6 1S f3 c6 16
ll'lc3 ii.f8 is fine for Black, while 14 cs
ll'lxcs 1 S .l:f.d8+ ii.f8 16 ll'lxa7 nxa7 17
.l:f.xc8 is probably best met by 17 ....l:f.e8
Instead 10 ... .l:f.xd8 11 .l:f.fd1 (after 11 (17 ... ..t>g7 18 f3 ll'le8 19 a3 ll'ld6 20 .l:f.d8 .l:f.f8! 12 .l:f.fd1 ii.g4 we're back in h6 2 1 ll'lh3 ll'le6 was also okay for Black

The M a r de/ Pla ta Varia t i o n : White 's E ig h t h M o ve Deviatio ns

in B.Larsen-R. Fischer, Monaco 1967, but D.Polajzer-M.Tratar, Slovenian Team

White could have tried 18 ii.xcs!?) 18 Championship 199S.
.l:f.xe8 ll'lxe8 19 ll'lf3 f6 20 ii.c4+ was 11 .. l:tbS 18 axb6 axb6 19 l:txd4 ll'le8 20
agreed drawn in this even position in .l:f.ddl
J.Piket-J.Nunn, Wijk aan Zee 1991. 20 ll'lxc8 is more prudent.
14. ..c6 1s ll'ld6 20 ... ll'lxd6 21 l:txd6 ii.b7 22 .l:f.a2 h6 23
After 1S ll'lxe6 ii.xe6 16 ll'lc3 a clever ll'lf3 l:txe4
idea is 16 ... :tb7!. Black frees the f8- Here Black's bishops gave him an
bishop and discourages White from edge in L.Gonda-J.Gallagher, Olbia
playing a4-aS because then Black 2008.
would get play along the b-file. After 17
b4 ii.f8 18 .l:f.ab1 ll'ld7 (with the idea of 82) 8 ...ll'lg4
... as!) 19 bs .l:f.c8 20 ll'lds ll'lcs 21 bxc6
:txc6 22 as bxas 23 ll'lf6+ ..t>h8 24 J:txb7
lbxb7 2S ll'ld7 ii.d6 White faced an up­
hill battle to draw in L.Portisch-J.Nunn,
Am sterdam 1990.

This is a more enterprising move.

9 f6 10
In stead 10 ii.h4 is sometimes
played, but this combines the negative
aspects of the Gligoric Variation - the
1s ...ll'ld4 bishop could be stranded on the king­
This works tactically and secures side - while giving up the positive as­
Black equal chances. Also possible is pect - White's flexibility with his king.
1s ... ll'lf4!? 16 ii.fl h6 17 ll'lgxf7 l:txf7 18 Play will transpose to less critical lines
ll'lxf7 ..t>xf7 19 as bxas with unclear in Ch apters 9 and 10 after 10... gs
play in L.Gonda-V.Kotronias, Hungarian (10...'it'e8 and 10...ll'lh6 are also possible
League 2007. and could transpose too to the Gligoric)
16 ii.xd4 exd4 17 as 11 ii.g 3 ltJ h6.
Or 17 .l:f.xd4 ll'le8 18 ll'lxc8 .l:f.xc8 19 Now Black threatens ... g4 and
.U.d2 h6 20 ll'lf3 .l:f.xe4 21 l:tad1 ..t>f8 V2-V2, ...ll'lxd4, so White must react:

A ttacking C h e s s : Th e King 's I n dian, Vol u m e 1

11 ii.g s
The bishop continues to bounce
around. Other moves also do not lead
to much:
a) 11 exfs exd4!? (11 ...ii.xfs 12 dS
Ci:Jd4 13 Ci:Jxd4 exd4 14 ii.xg4 dxc3 1S
ii.xfs :xfs 16 bxc3 ii.xc3 17 .l:f.b1 b6 is
fine too for Black) 12 Ci:Jbs (after 12
'it'd7 13 Ci:Jbs I like 13 ....l:f.xfs! when
White has little better than 14 .l:f.f8,
which is certainly fine for Black)
a) 12 dS Ci:Je7 13 Ci:Jd2 and now both 12 ... .l:f.xfs 13 h3 Ci:Jges 14 Ci:Jfxd4 l:tf7 (or
13 ... Ci:Jg6 and 13 ...fs are covered in note 14 ... Ci:Jxd4 1S Ci:Jxd4 .l:f.f8) 1S ii.e3 a6 16
'c' to White's 12th move in Line B11 of Ci:Jxc6 bxc6 17 Ci:Jd4 cs 18 Ci:Jc2 ii.b7 with
Chapter 9. good counterplay in Bu Xiangzhi­
b) 12 dxes dxes also transposes to T.L.Petrosian, Tiayuan 20os.
the Gligoric after 13 cs, which is con­ b) 11 dS Ci:Je7 (Black can also play
sidered in note 'a' to White's 12th move 11...Ci:Jd4!? 12 Ci:Jxd4 exd4 13 Ci:Jbs Ci:Jes
in Line A of Chapter 10, or 13 h3, which 14 exfs d3 transposing to the Gligoric!)
is note 'b' to White's 12th move in Line 12 Ci:Jgs Ci:Jf6 looks really stupid for
B of Chapter 10. White.
10 fs

This is consistent. Black plays the­

matically and hopes to show that
White has been wasting time with his
bishop moves. I prefer this to the open
10...exd4 and the unprovoked 10...Ci:Jh6.

The position is the same as the

main lines of the Bayonet, but White
has not played b4 or .l:f.e1. Still, when
playing the white pieces one can some­
times get away with a few things and
maintain equality. Indeed, after 13

Th e M a r de/ Plata Va r i a t io n : Wh ite 's E ig h t h M o ve De viations

exfs! ? ll'lxfs! (better than 13 ... gxfs 14 f4 b) 12 dxes ll'lgxes (or 12 ...fxe4 13
or 14 'ifb3 !?) the position is fairly level. ll'lxe4 ll'lgxes 14 ll'lxes ll'lxes with the
c) 11 dxes ll'lgxes 12 exfs ii.xfs 13 idea of ...'it'f7) 13 exfs 'it'xfs 14 'it'd2
ii.e3 (after the hasty 13 ll'lds?! ll'lxf3+ ii.e6 1s ll'lds 'iff7 16 ll'lxes ll'lxes 17
14 ii.xf3 ll'ld4 Black has some initiative) .l:f.ac1 .l:f.ae8 was about equal in R.Vera­
13 ...'it'f6 (or 13 ... 'it'd7) 14 .U.c1 (14 ll'lds R.Slobodjan, Havana 1999.
ll'lxf3+ 1S ii.xf3 'it'f7 16 'it'd2?! would c) 12 dS ll'ld4!? 13 ll'lxd4 exd4 14
give Black the initiative after 16 ...ll'les ll'lbs fxe4 1s ll'lxd4 .l:f.xf2 16 ll'le6 .l:f.xf1+
17 ii.e2 c6) 14....l:f.ae8 1s b3 h6 16 'it'd2 17 ii.xf1 (17 'it'xf1 'it'xe6 18 dxe6 ii.d4+
gs 17 ll'ld4 ll'lxd4 18 ii.xd4 and now 19 ..t>h1 ll'lf2+ is a perpetual) 17 ...ll'lf6 18
18 ...g4 19 ..t>h1 'it'g6 gave Black good ll'lxg7 'it'xg7 19 'it'd4 ll'lg4 20 'it'xg7+
play in J.Granda Zuniga-J.Polgar, Aruba ..t>xg7 21 ii.d8 c6 reaches an unclear
1992. A promising alternative is ending.
18 ... ll'lf3+ 19 ii.xf3 'it'xd4 20 'it'xd4 ii.xd4
21 .l:f.fd1 (21 ii.xb7? ii.d3) 21 ...ii.cs! 821) 11...'it'eS
when 22 ii.xb7 ii.a3 win s the exchange.
Returning to 11 and here Black
has a pleasant choice:

Black keeps pieces on the board.

12 dxes
a) 12 exfs gxfs (12 ... e4!? is possible,
BU: 11 'ife8
... while after 12 ...ii.xfs 13 h3 ll'lf6 14 dS,
822: u. ..J:.f6
. as in B.Lalic-V.Zivkovic, Zagreb 2007,
14... ll'ld4!? could be investigated) 13
There is even a third option in dxes dxes 14 h3 ll'lf6 1s ll'lds 'iff7 16
11 ...'it'd7!?, which looks funny but is not ii.e3 h6 gave Black an active position in
bad: B.Kantsler-V.Bologan, Moscow 1991.
a) 12 exfs exd4 13 ll'lb s l:txfs ! is note b) 12 ll'lds 'it'f7 13 dxes (the creative
'a' to White's 11th move, above. 13 ii.e7 ll'lxe7 14 ll'lgs 'it'e8 1s ll'lxc7

1 63
A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

'it'd8 16 ll'lce6 - 16 ll'lxa8 exd4 is less 13 ...ll'lf6

clear - 16 ... ii.xe6 17 ll'lxe6 'it'd7 18 exfs
gxfs 19 ll'lxf8 .l:f.xf8 20 h3 was S.Lputian­
V.Bologan, Azov 1991, and here
20 ... ll'lf6 looks much better for Black)
13 ... ll'lgxes (or 13 ... dxes, leading to the
note to White's 13th move, below) 14
exfs ii.xfs 1s 'it'd2 l:tae8 was level in
P. S an Segundo Carrillo-V.Topalov, Ma­
drid 1997.
c) 12 dS ll'ld4? 13 ll'lxd4 exd4 14 ll'lbs
'it'es 1s f4! 'it'xe4 16 ii.xg4 fxg4 17 l:te1
'iffs 18 ll'lxd4 ii.xd4+ 19 'it'xd4 was better 14 ii.d3
for White in C.Toth-G.Hemandez, Merida This move has been criticized, but
1993, but 12 ... ll'le7 and 12 ... ll'ld8!? both White does not have any advantage:
look very playable for Black. a) 14 cs ll'lxe4 1s ll'lxe4 fxe4 16 ll'ld2
12 ...dxes (after 16 'it'ds+ 'iff7 both 17 'it'xe4 ii.ts
Preferring the dynamic pawn struc­ and 17 'it'xf7+ .l:f.xf7 18 ll'ld2 ii.e6 give
ture that arises to 12 ... ll'lgxes 13 exfs Black good play) 16 ... ii.e6 17 ll'lxe4 h6
ii.xfs 14 'it'd2 'iff7 with equality in gives Black a very active position.
R.Vera-J.Hebert, Montreal 1997. b) 14 ii.xf6 is the safest. After
13 h3 14...ii.xf6 1s ll'lds Black can try 1s...fxe4,
White can also play 13 ll'lds 'iff7 14 1s ...'it'd8 or the ambitious 1s ...ii.d8!?.
ii.d2 (instead 14 ll'ld2 h6 1s ii.h4 gs 16 14.. .ii.e6 15 �el 'iff7 16 cs
ii.xg4 fxg4 17 ii.g3 ll'ld4 gave Black a Instead 16 exfs gxfs 17 ll'lxes fails
good position in V.Burmakin-V.Bologan, to 17 ... ll'lxes 18 .l:f.xes ll'le4.
Sochi 2004) 14...ll'lxf2! ? (more enterpris­ 16...ll'ld7 1
ing than 14...ll'lf6 1s ll'lgs 'it'd7 16 exfs
gxfs) 1s l:txf2 fxe4 16 ii.e3 exf3 17 l:txf3
'it'd7 18 .l:f.xf8+ ..t>xf8 19 'iff1+ ..t>g8 20
.l:f.d1 (20 ll'lf6+ hf6 21 'it'xf6 looks like a
better try, although Black should be fine
here) 20 ...ll'ld4 21 ii.xd4 exd4 22 ll'lf6+
hf6 23 'it'xf6 'it'e8 24 ii.f3 'if e3+ 2S ..t>f1
ii.e6 26 'it'xd4 'it'xd4 27 .l:f.xd4 cs 28 .l:f.e4
'itf7 gave Black slightly the more pleas­
ant endgame in O.Csoli-G.Pirisi, Hungar­
ian League 2001.

1 64
The M a r de/ Plata Var i a t i o n : Wh ite 's Eig hth M o ve D e v i a t i o n s

Black already has some initiative. 17 ll'ld2 White had some edge follow­
11 ing both 17 ...ll'lhf7, E.Scarella­
After 17 exfs gxfs 18 ii.e3 h6 Black D.Valerga, Villa Martelli 2002, and
also stands very well. 11 ... ii.d7 18 .l:f.c1, P.Ricardi-G.Kasparov,
11...ll'ld4! Buenos Aires (simul) 1992 (a game
White is in trouble, since 18 ii.e3 which Kasparov lost). I think Black can
!Lixf3+ 19 'it'xf3 c6 20 ii.e2 f4, with the improve, however, with 14...ll'ld4!.
idea of ...ll'lxcs, is much better for Black.
Instead 18 ii.xd7 ii.xd7 19 ii.h4 ii.c6 20
!Ligs 'it'f6! 21 exfs gxfs 22 ll'le2 .l:f.ad8 23
�xd4 .l:f.xd4 24 �5? h6 and 0-1 was
l.Sokolov-A.Shirov, FIDE World Champi­
onship, Las Vegas 1999. Here 25 ll'lf3
loses to 2s ...ii.xf3.

822) 11 ii.f6

13 exfs
Instead 13 dS ll'le7 leaves Black with
the active ideas of .. .fxe4 and ...ll'lfs or
.. .f4, and 14 exfs ll'lxfs is certainly okay
for him.
Another possibility is 13 dxes dxes
14 'it'xd8 .l:f.xd8 15 ll'lds (15 exfs can be
met by 15 ... ii.xfs with equality or Kas­
parov's suggestion 1S ...e4) 1s ...ll'lxe4!
This move was Kasparov's choice 16 ll'lxc7 .l:f.b8 17 .l:f.fd1 (better is 17
and it was also played with success by .l:f.ad1 ii.d7 18 .l:f.fe1) 11 ... ii.d7 18 ii.d3
Anand... against Kasparov! l:tbc8 (Kasparov suggested 18 ... ll'lf6 and
12 ii.xf6 ll'lxf6 18 ... ll'lcs! looks good too) 19 ll'lds ll'lcs
Also possible is 12 ...'it'xf6 !? 13 h3 20 ii.fl ii.e6 , which is already some­
(the alternatives 13 dxes ll'lgxes, 13 what better for Black. Indeed, after 21
exfs ii.xfs and 13 dS ll'ld4!? do not look b4? ii.xds 22 cxds ll'lxb4 23 ll'lxes ll'le4
dangerous) 13 ...ll'lh6 14 dS (14 dxes 24 ii.c4 ll'lc3 25 .l:f.d2 ll'lbxds 26 g3 ..t>g7
dxes is relatively unexplored, but Black 27 ..t>g 2 ll'le3+ White resigned in
has done well in practice), but now af­ L.Portisch-G.Kasparov, Linares 1990.
ter 14 ...ll'ld8 15 cs ll'ldf7 16 cxd6 ll'lxd6 13 ...ii.xts

1 65
A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

This is more solid than 13 ... gxfs 14 'it'b6 18 h 3 ii.d7 19 dxc6 bxc6 20 .l:f.ae1
dxes dxes 1 5 'it'xd8 .U.xd8 16 ll'lds ll'le8 .l:f.ad8?! (20 ... l:tae8) 21 ll'la4 'ifC7 22 f4
17 l:tad1 when White had some pres­ gave White the initiative in D.Rajkovic­
sure in A.Mikhalchishin-K.Hulak, Por­ J.Nunn, German League 1990.
toroz/Rogaska Slatina 1993. Black has also tried 1s ... 'it'c8?!, but
14 d5 this is a strange choice of square. After
Instead 14 dxes dxes is fine for 16 ii.d3 ii.xd3 (16... c6 17 l:tc1!) 17 'it'xd3
Black. His active pieces and dark-square White has a slight edge. However, a
control compensate for the isolated es­ better version of this idea is 1 s ... 'it'd7!?
pawn. 16 ii.d3 c6 17 ii.xfs ll'lxfs 18 ll'le6 .l:f.fe8,
14...ll'le7 with the idea of ... ll'lg7 or perhaps even
capturing twice on dS to undermine
the e6-knight.
16 ll'le6 ii.xe6 17 dxe6 ll'lfs
Alternatives are 17 ... c6 and 17 .....t>g7.
18 ii.d3 ll'ld4 19 f4
19 ii.xg6 ll'lxe6 is unclear.
19 .....t>g1 20 'it'e1 ll'lxe6 21 'ifg3 gs

If White cannot generate any initia­

tive, the exchange of dark-squared
bishops will favour Black positionally.
1s ll'lgs
After 15 'it'd2 ..t>g7 16 ll'lh4 ii.d7 17
f4 exf4 18 'it'xf4 ll'lfs 19 ll'lxfs+ ii.xfs 20
g4 ii.d7 21 gs ll'lg4 22 'it'xf8+ 'it'xf8 23
.l:f.xf8 .U.xf8 24 ii.xg4 ii.xg4 25 ll'lbs l:tf4 The position is unclear. Now 22
Black was taking over the initiative in fxes?! dxes 23 'it'xes?? 'it'xd3 24 'it'xe6
G.Kasparov-V.Anand, Geneva (rapid) .U.ae8 25 'iffs ll'le4! 0-1 was the sudden
1996. end of P.Van der Sterren-G.Kamsky,
1s...h6 Wijk aan Zee 1994. Instead 22 fxgs
Instead 1 s ... c6 16 ii.d3 ii.g4 (proba­ hxg s with the idea of . . .ll'lhs is unclear.
bly better is 16 ...'it'b6 17 ii.xfs ll'lxfs 18 Black's kingside is rather loose, but his
ll'le6 .l:f.f7 with the idea ...ll'ld4) 17 'it'd2 knights are active.

1 66
Part II
The Classical· Variation
1 d4 �6 2 c:4 g6 3 lbc3 .ig7 4 e4 d6
s ttJf3 o o 6 .ie2 es without 7 o-o

1 d4 ltJf6 2 C4 g6 3 ltJc3 Ji.g7 4 e4 d6 5 Here we look at those lines where

0if3 o-o 6 ii.e2 es White does not play 7 o-o. There are
three different systems to consider and
they are vastly different. With the Gli­
goric Variation (7 ii.e3) White main­
tains the tension, which makes it more
difficult for Black to formulate a con­
crete plan. The Petrosian Variation (7
dS) does quite the opposite - White
fixes the centre immediately. Lastly
there is the Exchange Variation (7
dxes), where White resolves the ten­
sion in a rather extreme fashion by
trading queens.

Chapter 9
Gligoric Variation
7 i.. e 3 ttJg4 8 ..tgs f6 9 ..tc1 ttJc6 and 9 i.. h4 ltJc6

1 d4 ll'lf6 2 c4 g6 3 ll'lc3 ii.g7 4 e4 d6 5 has to be ready to handle the theory of

ll'lf3 0-0 6 ii.e2 es 7 ii.e3 several lines, whereas Black only really
needs to know one. White also has to
be prepared to play with several differ­
ent pawn structures. This makes it dif­
ficult to play compared to, for example,
the Petrosian Variation, where White
really only has to understand one
structure. Because of this demand on
the white player, the Gligoric is more
popular amongst titled players than it
is at other levels.
Before we move on to my principal
The Gligoric Variation is a pretty recommendation, 7 ...ll'lg4, here is a
flexible continuation for White and summary of Black's other options:
this makes it very dangerous. White a) 7 ...ll'lc6?! is played far too often.
keeps the tension in the centre and After 8 dS ll'le7 9 ll'ld2 White has the
does not disclose what he intends to do best of all worlds: an active bishop on
with his king. It is not easy for Black e3 that is typical of Chapter 1; a knight
because White can adjust his plans on d2 that is ready to jump to c4 after
depending on how Black chooses to c4-c5 as we saw in Chapter 4; and the
continue. The upside to all of this is flexibility to castle on either side of the
that the Gligoric is not easy to play for board (or not at all), as we shall see in
White because Black has eight (!) main the Petrosian Variation of Chapters 11
continuations. Because of this, White and 12.

Gligo ric Va ria t i o n : 7 ii.. e 3 tlJ g 4 8 ii.. g 5 f6 9 ii.. c 1 tlJ c 6 a n d 9 ii.. h 4 tlJ c 6

b) 7...tlJa6 i s perfectly respectable. but nowadays Black's exchange sacri­

After 8 o-o (not so dangerous is 8 dS fice 8 dxes (8 dS is also possible)
.'Llg4) we are in the main lines of 7 0-0 8 ... dxes 9 tLJds 'ikd8 (9 ... tLJxds 10 cxds
.'Lla6 8 ii..e 3 which is outside of our rep­ c6 11 d6 is also considered to favour
ertoire. This approach has its positive White) 10 ii.cs tLJxe4 11 ii.. e 7! 'ikd7 12
side, however, because Black only has ii..xf8 ..t>xf8 is considered insufficient
to know the 8 ii..e 3 line (8 .l:f.e1 is an­ after 13 'ikd3 !.
other major variation after 7 o-o tlJa6). e) 7 ...exd4 is one of Black's main
c) 1 ...tlJbd7 similarly is likely to continuations. Aft er 8 tlJxd4 l:te8 9 f3 c6
tran spose to the line 7 o-o tlJbd7 8 ii.. e 3 the fi nesse 10 ii..f2 ! is considered to
when 8....l:f.e8!? 9 dS tLJhs 10 g3 ii..f8 has favour White (but 10 o-o dS 11 cxds
recently become fashionable. This is tLJxds! equalizes instantly and 10 'ikd2
Bologan's DVD recommendation, but dS 11 exds cxds 12 cs tlJc6 13 o-o
he does not even consider 8 dS which .l:f.xe3! is another Kasparov exchange
leads to what I consider to be a good sacrifice).
variation of the Petrosian variation for f) 7...c6! ? is a fashionable line, al­
White after 8...tlJg4 (otherwise 9 tlJd2) 9 though here Black has to know a couple
�d2 ! (better than the more common 9 of different structures. One point of his
�g s) 9 .. .fs 10 tLJgs tLJdf6 (10 ... tLJcs 11 move order is that 8 o-o exd4 9 tlJxd4
b4) 11 exfs gxfs 12 h 3 tlJh6 13 g4 when can be met with 9 ... .l:f.e8 10 f3 dS when
White has the initiative. Black has snuck back into the equaliz­
ing line of 7...exd4. Instead White can
play 9 ii..xd4 with some chances of an
edge, but he often plays 8 dS instead.
The blocked position arising after
8...tlJg4 9 ii.g s f6 10 ii..h4 cs remains
g) 7 ...h6!? was a favourite of John
Nunn's. The point is that after 8 o-o
Black can play 8 ...tlJg4 9 ii..c1 tlJc6 10 dS
tlJe7 11 tLJe1 fS 12 ii..xg4 fxg4 with an
unusual and double-edged pawn struc­
There is a great trap concealed here ture. Instead 8 h3 is well met by 8 ... exd4!
too, as after 13 ... ..t>h 8? 14 tlJe6 ii..xe6 1S 9 tlJxd4 .l:f.e8 when 10 f3 would combine
gs! White is down a whole knight, but poorly with 8 h3 due to White's dark­
Black cannot avoid losing two pieces! square weaknesses on the kingside. 8 dS
d) 1 ...'ike7 was played by Kasparov is also considered harmless due to
against Karpov in their 1990 match, 8... tlJg4 9 ii..d2 fS (equalizing instantly

1 69
A ttacking C h e s s : Th e King 's I n d i a n , V o l u m e 1

according to Nunn), but after 10 h 3 ! A) 9

(not, though, 10 exfs ii.xfs! 11 h3?
ll'lxf2! 12 ..t>xf2 e4 when Black wins back
the piece with a strong attack) 10 ... ll'lf6
11 exfs gxfs (11 .. .ii.xfs 12 ll'lh4) 12 g3!
ll'la6 13 'ifcl ..t>h7 14 'it'c2 White's clever
manoeuvres gave him a pleasant ad­
vantage in E.Perelshteyn-B.Lopez, US
League 2009. Here Black's ...h6 move has
created weaknesses on the kingside and
hurt his position. We will discuss this
idea again in Chapter 11.
7 ...ll'lg4 This move makes a funny impres­
sion, undeveloping White's pieces.
Black has essentially played the moves
...ll'lg4 and ...f6 'for free' and now he
has the move as well. However, these
free moves do not automatically help
Black's position.
This is the most popular move and
fits in well with our repertoire. Even so,
some players are sceptical about its
value and in New in Chess the strong
This is the main line and it is has Georgian Grandmaster Mikheil
been Radjabov's frequent choice. It is Mchedlishvili wrote that '9 ... ll'lc6 just
both principled and forcing. White won't do for Black'. While I agree with
cannot very well allow his bishop to be the ideas presented by Mchedlishvili, I
captured, so he invariably plays: will dare to challenge his assessment of
8 �gS f6 9 ... ll'lc6 ...
Now White has a choice of retreats. Also possible is 9 ...ll'ld7 when 10 o-o
ll'lh6 transposes to the line 7 o-o ll'lbd7
k t · 8 ii.e3 ll'lg4 9 f6 10 ll'lh6. Black
B: 9 JUi4 has two other independent continua­
tions that are considered respectable in
Instead 9 ii.d2 cuts off White's con­ 9 ... exd4 and 9 ... fs, but we will not con­
trol of d4 and makes little sense. After sider them here.
9 ... ll'lc6 10 dS ll'ld4 Black is doing well. 1o d s

1 70
G lig o ric Va r i a t i o n : 7 ii.. e3 tlJ g 4 8 ii.. g 5 f6 9 ii.. c1 tlJc6 a n d 9 ii.. h 4 tlJc6

Instead 10 o-o transposes t o the line advantage as shown in several games,

7 o-o tlJc6 8 ii..e 3 tlJg4 9 f6 10 ii.. c 1 while 13 ...f4 14 g4!, with the idea of
which was covered in Line B2 of Chap­ 14_.fxg3 1S fxg3, is also much better
ter 8, while after 10 h3 Black should for White) 14 h S (or 14 ii..xg4 fxg4 1 S
avoid 10 ...tlJh6 11 dS! tlJe7 12 h4!, ii.. e 3) 14...tlJf6 1 S hxg6 hxg6 1 6 ii..f3 and
which transposes to the note to his White has an advantage across the
10th move, below, and play 10...exd4 whole board.
11 tlJxd4 tlJges 12 tLJxc6 (12 ii..e3 fS) b) 12 ...tlJf7 13 h s cs (after 13 ... gs 14
12 ...tLJxc6 intending 13 .. .fs. 'it'c2 h6 1s g4! c6 - 1s ... ii..xg4 16 tLJxgs!
10 tlJd4!
... ii..x e2 17 tlJe6 wins for White - 16 tLJh2
Instead 10...tlJe7 seems natural. It the battle can only be decided on the
looks as though Black is doing well - queenside, where White has a space
compared to lines in the Mar del Plata, advantage) 13 ... cs 14 hxg6 hxg6 1 S
Black is up a few tempi and ready to ii.. e 3 ii..d7 16 g 3 ! ? (with the idea of ..t>f1-
play .. .fs. However, White has not cas­ g2 and 'it'g1-h2) 16 ...a6 17 ..t>f1 fS 18
tled and because of that he has extra ..t>g2 fxe4 19 tlJxe4 tLJfs 20 tLJxgs
options. Another problem is the e7- 21 tlJfxgs 'ife7 22 .l:f.h7 tlJd4 23 ii..g4 .l:f.fs
knight. If this knight cannot contribute 24 �1 .l:f.xgs 2 s ii..xd7 .l:f.fs 26 ii..xfs
to Black's kingside play he often suf­ gxfs 27 tlJd2 and White was better in
fers. This is a difficult variation for M.Mchedlishvili-G.Kacheishvili, Tbilisi
Black and it is this line for which 2007.
Mchedlishvili condemned 9 ...tlJc6. The c) 13 .. .fs 14 hxg6 tlJxg6 1s 'it'c2
specific reason is that 11 h 3 ! tlJh6 12 (White pressures fS and prepares to
h4! causes Black major problems. castle queenside) 1S .. .f4 16 ii..d2 cs 17
Some examples: o-o-o a6 18 g3 ii..d7 19 l:LJM tlJxh4 20
l:txh4 fxg3 (a better try is!?) 21
fxg3 tLJgs 22 .l:f.dh1 bs 23 'it'd3 (23 .l:f.h s
also brings Black to the brink of defeat)
23 ...bxc4 24 'it'e3 h6 2S .l:f.xh6 ii..xh6 26
.l:f.xh6 tlJf7 27 .l:f.g6+ ..t>h7 28 .l:f.xd6! tlJxd6
29 �6+ ..t>g8 30 'it'g6+ ..t>h8 was
A.Maric-1.Gaponenko, Moscow 2001.
Here the simplest is 31 'it'xd6 when
White has tremendous compensation
for the two exchanges.
Now we return to 10...tlJd4!.
a) 12 ...fs 13 tLJgs! tlJg4 (13 .. .fxe4 14 This move is eminently logical.
g4! controls fS and gives White a big White has not been developing his

1 71
A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

pieces and, although this has granted fxg4 14 o-o (14 ll'lxd4 'ife7! 1s 'it'd3 ii.ts
him some flexibility, it could clearly gives Black the initiative), which re­
become a problem should the position mains untried. 14...'ife7 looks good
open up. This move was not considered enough after 1S 'it'd3 (1S .l:f.e1 'it'f6!)
by Mchedlishvili, but I believe it com­ 1 s...l:te8 (or 1s...cs!? 16 dxc6 bxc6 17
pletely solves Black's problems after 9 ll'lxd4 ii.b7, with ideas like ...cs and . ...l:f.ae8) when the e4-pawn gives White
problems. White will be forced to play
f3, either immediately or after 16 .l:f.e1
ii.ts, when ... gxf3 will give Black good
compensation for the falling d4-pawn.

11 ll'lxd4 exd4 12 ll'lbs

After 12 ii.xg4 dxc3 13 ii.xc8 'it'xc8
14 bxc3 Black is already better following
14 ...'if e7 or 14...fs, while 12 'it'xd4 fS is
very risky. One high-level example went 13 ll'les

13 'it'd1 fxe4! 14 ii.xg4 'ifh4 1S ii.e2 Black should not get carried away
ii.xc3+ 16 bxc3 'ifxf2+ 17 ..t>d2 ii.g4! 18 with 13 ... ll'lxh2? 14 ..t>xh2 fxe4 1 S ll'lxd4
cs .l:f.fs! 19 'it'e1 .l:f.xds+ 20 ..t>c2 'it'xg2 21 �4+ 16 ..t>g1 17 f4 exf3 18 ll'lxf3
l:tg1 'it'xe2+ 22 'it'xe2 ii.xe2 23 cxd6 .l:f.xf3 19 l:txf3 �2+ (or 19 ... ii.g4 20
ii.d3+ 0-1, E.L'Ami-L.Nisipeanu, Euro­ ii.e3) 20 �1 �1+ 21 �2 'ifh4+ 22
pean Championship, Budva 2009. ..t>e3 when the white king was safe
10...ll'ld4 must have been a shock to enough and the material advantage
L'Ami, who is usually very well prepared. counted in J.Ulko-D.Sokolov, Moscow
12 fs 13 0-0
... 200S.
Instead 13 h3 is crushed by 14 exfs
13 ... ll'lxf2 14 ..t>xf2 fxe4+, while 13 exfs White takes the pawn. Instead 14
also runs into 14 ... ll'lxf2! 14 ..t>xf2 and ll'lxd4 causes no problems at all for
now 14 ... a6! 1s ll'la3 ii.xfs 16 ii.f3 d3 Black. After 14_.fxe4 1 S ii.e3?! (safer is
gives Black a strong attack. 1 S ll'le6 ii.xe6 16 dxe6, but Black has
The only other option is 13 ii.xg4 several possibilities here, the surest

1 72
G lig o ric Va ria t i o n : 7 ii.. e3 tlJ g 4 8 ii.. g 5 f6 9 ii.. c1 tlJ c6 a n d 9 ii.. h4 tlJ c6

being 16 ... c6) 1 S ... �4 16 .l:f.c1 (it is al­ b) 17 'it'b3 a6 18 tlJc3 ii..d3 19 l:td1
ready too late for 16 tlJe6 ii..xe6 17 dxe6 �4 20 ii..e 3 ii..xc4 was good for Black
because of 17 ...tlJf3+! 18 gxf3 19 f4 in E.otero-R. Leitao, Cuba 1993. If 21
.l:f.xf4, as pointed out by Golubev) 'it'xb7 �ab8 22 'ifxe7 .l:f.xb2 and Black
16 ... tlJg4 17 ii..xg4 ii..xg4 18 'ifd2 ii..e S has the initiative.
Black already has a stron g attack: 19 f4 c) 17 'ifg3 prevents ...�4. but is an
(19 h 3 ii..x h3 and 19 g3 � 3 are no bet­ odd square for the queen.
ter) 19 ... exf3 20 tlJxf3 ii..xf3 21 gxf3
.l:f.ae8 and Black was clearly better in
A.Rychagov-M.Vachier Lagrave, Russian
Team Championship 2009.

Black has:
c1) 17...a6 18 tlJc3 ! ? 19 f4 ii..g 7
20 ii..d2 'it'f6 21 b3 bS! 22 cxbs 'it'd4+ 23
:tf2 axbs 24 .l:f.c1 was F.El Debs-D.Flores,
14 d3!
... Campinas 2010, and here 24...'it'cs! ?
Black cannot save the d4-pawn, so gives Black good play.
he uses to it to grab the bishop-pair c2) 17 ... 'it'd7 18 tlJc3 .l:f.ae8 19 ii..d2
and gain time for development. 20 f4 ii..d4+ 21 ..t>h1 a6 22 .l:f.ae1 bS
15 ii..x d3 tlJxd3 16 'it'xd3 ii..xfs and Black had counterplay in A.David­
Black's powerful bishops give him D.Stets, Fourmies 2010.
excellent compensation for the pawn. c3) 17...ii..e s immediately is also
White will also have some difficulties possible: 18 ii.g s!? 'it'd7 19 'it'b3 a6 20
developing his queenside, and the c4- tlJa3 (after 20 tlJc3 ii..d 3 21 .l:f.fd1? 'iffs
pawn may become vulnerable. threatening ...ii..c 2 is good for Black,
17 'it'dl while after 21 .l:f.fe1 Black should play
White has several alternatives, but 21 ...'it'fs with compensation, rather
Black always has enough play: than allow 21 ...'it'g4 22 .l:f.xes! when
a) 17 'it'e2 a6 18 tlJc3 'ifh4 19 ii..e 3 White gets counterplay) 20 ... 'ifg7 21
.l:f.ae8 gave Black good compensation in .l:f.ae1 l:tae8 22 tLJc2 (not 22 'it'xb7 ii..d 3)
R.Vidonyak-D.Stets, Lvov 2009. 22 ...ii..xb2 23 tlJe3 ii..c 8 (23 ...ii..d 4!?) 24

1 73
A ttacking C h e s s : The Kin g 's I n dian, Vo l u m e 1

ii.h6 'it'xh6 2 5 'it'xb2 .l:f.e4 was level in active and White's queenside is weak.
D.Yevseev-A.Khruschiov, Peterhof 2009. 21 ii.e3
11 'it'f6
••• This is sensible. White prioritizes
This active move leaves the Cl-pawn developing his pieces.
to its fate, but Black will gain sufficient 21 b6 22 ii.d4 ii.xd4 23 'it'xd4 .l:f.xc4

play. After 17...'it'M White should go for

18 ll'lxe7! as in E.Kanter-A.Savina, Dimi­
trov 2007. However, 17...a6 is plausible:
18 ll'ld4 �4 19 ii.e3 .l:f.ae8 20 'it'd2 ii.xd4
(Black basically plays for a draw against
his much higher-rated opponent with
opposite-coloured bishops, rather than
depending on the bishop-pair for long­
term compensation) 21 ii.xd4 .l:f.e4 22
ii.c3 .l:f.fe8 23 f3 .l:f.e2 24 'it'd4 'it'xd4+ 2 5
ii.xd4 .l:f.c2 2 6 .l:f.fc1 .l:f.ee2 27 .l:f.xc2 .l:f.xc2
28 b3 .l:f.d2 29 ii.c3 .l:f.c2 30 ii.f6 rt;f7 31 The endgame is equal. A couple of
ii.d8 .l:f.b2 32 .l:f.e1 .l:f.b1 33 rtlf2 l:txe1 34 examples are 24 'it'd3 l:te4 25 b3 .l:f.e2 26
..t>xe1 c6 and Black held in L.Fressinet­ 'it'd4 d5 27 a4 'ife4 28 .l:f.ad1 'it'xd4 29
D.Sharavdorj, Las Vegas 2009. .l:f.xd4 .l:f.d8 30 .l:f.fd1 .l:f.b2 with equality in
L.Andonovski-M.Jazadzievski, Struga
2008, and 24 'it'd2 .l:f.e4 25 .l:f.fe1 .l:f.xe1+
26 .l:f.xe1 'it'xa2 27 f3 d5 28 .l:f.e7 .l:f.f7 29
.l:f.e8+ .l:f.f8 30 .l:f.e7 with a draw in
A.Gavrilov-A.Zontakh, Lipetsk 2009.

B) 9 ii.h4

18 ll'lxc7 .l:f.ac8 19 ll'le6

After 19 ll'lb5 .l:f.xc4 Black has excel­
lent compensation for the pawn.
19 ii.xe6 20 dxe6 'it'xe6

White has managed to hold on to

the pawn and eliminate Black's bishop­
pair, but Black's major pieces are all

1 74
G l ig o ric Va riati o n : 7 ii.. e3 tlJ g 4 8 ii.. g 5 f6 9 ii.. c1 tlJ c 6 a n d 9 ii.. h4 tlJc6

This is much more challenging. transpose to variation 'b', below)

White keeps his bishop active and hin­ 11 ... tLJe7 and here:
ders the advance of Black's f-pawn. Be­ al) 12 'if c2 cs 13 g4 tlJf7 is fine for
cause this line is critical for 7...tlJg4, we Black. The plan is to play ...ii..d7, ...tlJc8
will consider two main lines. The move (to protect the f6-pawn), and then
9 ...tlJc6 is 'consistent', but most players ... ii..h 6, activating the bishop. If White
prefer the modem 9 ... g s, which we will castles queenside, then ... a6 and ... bs
consider in the next chapter. will be enticing. One example was 14
9 tlJc6
... o-o-o a6 1s ..t>b1 ii..d 7 16 tlJd2 'it'as 17
Black sometimes plays 9 ... tlJd7 10 l:tdf1 bS 18 ii..g 3 ii.. h6 with a good posi­
o-o tlJh6, which is a solid line generally tion in V.Gefenas-J.Van Oosterom, cor­
reached from the move order 7 o-o respondence 1996.
tlJbd7 8 ii..e3 tlJg4 (8 ... c6 is more com­ a2) 12 tlJd2 gs 13 ii.. g 3 fs 14 exfs (or
mon here, while 8 ... l:te8 is fashionable) 14 f3 tlJg6 with counterplay) 14 ...tlJhxfs
9 ii.g s f6 10 ii..h4 (and here 10 ii..d 2 and 1 S tlJde4 tlJg6 16 'it'd2 tLJf4 17 ii.fl tlJd4
10 are seen more frequently) 18 o-o-o a6 19 ..t>b1 ii..fs 20 f3 cs 21
10 ... tlJh6. The older 9 ... 'it'e8 is not so bad ii..d 3 bS was another good example of
either, but I have complete confidence Black's possibilities in F.Gheorghiu-
in my two recommendations ! 1 .Sokolov, Kava la 1990.
b) 10 dxes dxes (even simpler is
10 ... tLJgxeS; then 11 tLJxes dxes is equal
and 11 o-o can be met with 11 ... ii..e 6,
intending ...'it'd7 and ...fs, or by
11...tLJxf3+ 12 ii..xf3 ii..e 6, again intend­
ing ...'it'd7 and ...fs), and here 11 o-o
tlJh6 is again comfortable for Black.
bl) 11 'it'xd8 tlJxd8 is equal (but not
11 ... l:txd8? 12 h3 tlJh6 13 tLJds).
b2) 11 tLJds tlJh6 (intending to con­
1o d s quer d4 with ... ii..g4) 12 h3 and here
Only this move is critical. Instead 10 12 ...tlJd4?! is met by 13 tlJxd4 exd4 14
o-o leads us back to a position consid­ es!, while 12 ... ii..e 6 13 'ifb 3 ! 'it'c8 14
ered in the notes to White's 10th move 0-0-0 gave White some advantage in
in Line B2 of Chapter 8. others: V.Korchnoi-V.Nevednichy, Paks 2004.
a) 10 h 3 tlJh6 11 dS (11 dxes dxes However, Black can instead play 12 ... g S
makes little sense because Black plays 13 ii..g 3 f S with counterplay, V.Shish­
... tlJh6 anyway, but this could still kin-C.Carmaciu, Bucharest 2006.

1 75
A t tacking Ch ess: Th e K i n g 's I n d i a n, Vo l u m e 1

10 ll'le7
... After 11 ll'ld2 Black can retreat his
This retreat is more viable than it knight or maintain it on g4 for a move.
was in Line A, because it is not possible
for White to launch his h-pawn for- 81: 11 lbh6

ward with his bishop on h4. The main 82: 11 h s...

concern for Black i s still that the e7-

knight has trouble finding a useful Instead 11 .. .fs 12 ii.xg4 fxg4 13
role. :tf4! ? (perhaps Black could try 13 ...h6
It would be nice if the untried 14 ii.e3 gs 1s cs ll'lg6) is possible. Th is
10 ...ll'ld4 worked here too, but it looks a looks a bit too specul ative, though, af­
bit too speculative after 11 ll'lxd4 exd4 ter 14 o-o (there is no point in playing
in view of both 12 'it'xd4 g s 13 ii.g3 fS 14 ii.xf4 exf4 because Black will spend
14 'it'd1! ll'lf6 (14 .. .f4? 1 s ii.xg4) 1s exfs a tempo to force the capture anyway)
(1S f3 fxe4 16 fxe4 ll'ld7 gives Black 14...h 6 1s ii.xf4 exf4 16 ll'lb3 gs 17 ll'ld4
some play) 1 s...ii.xfs 16 o-o ll'le4 17 ii.xd4 18 'it'xd4 ll'lg6, as in Y.Shulman­
ll'lxe4 ii.xe4 18 'it'd2 'it'f6 19 l:tae1 'it'xb2 A.Poluljahov, Gausdal 1994, and here
20 'it'xg s and 12 ll'lbs .l:f.e8 13 h3 ll'lh6 19 cs would test Black's concept.
(13 ... ll'les 14 o-o cs 1s dxc6 bxc6 16
ll'lxd4 'ifb6 17 ll'lb3 also look insuffi­ Bl) 11 ll'lh6

cient) 14 f3 cs 1S dxc6 bxc6 16 ll'lxd4

'it'b6 17 'it'd2. Well, I tried.
11 ll'ld2
This natural move unleashes an at­
tack on the g4-knight. Instead 11 h 3
tran sposes back to 1 0 h 3 , while 1 1 o-o
will transpose to lines with 12 o-o after
11 ... ltJh6 12 ltJd2 OT 11 ...h S 12 ltJd2.

This is Black's main continuation.

12 f3
White stays flexible. other moves
show his hand too early:
a) 12 g4 ll'lf7 13 'it'c2 cs 14 f3 ii.d7 is
an easy-enough-to-handle set-up for
Black, especially with White committed
to g4.

1 76
G ligo ric Va ria tio n : 7 ii.. e 3 tlJ g 4 8 ii.. g 5 f6 9 ii.. c1 tlJ c 6 a n d 9 ii.. h 4 tlJ c 6

b ) 12 b4 does really not do that This is the sharpest continuation.

much, other than rule out castling Black initiates play on the kingside
queen side. Black can just begin his play immediately. There is a serious alterna­
on the other side of the board with tive in 12 ... cs. Black tries to block the
12 ... g s 13 ii.. g 3 fs (or 13 ... tlJg6) 14 f3 queenside before beginning his king­
tlJg6. side campaign. This has been popular
c) 12 o-o gs 13 ii..g 3 is committal amongst the top players, but I think it
too: actually gives White quite a pleasant
cl) 13 ... tlJg6 14 l:tel (after 14 l:tcl fS choice.
1S exfs ii..xfs 16 tlJde4 tLJf4 17 cs ii.. g6 Here 13 a3 prepares 14 b4, but
18 .l:f.e1 tLJxe2+ 19 'it'xe2 tLJfs 20 'it'bs g4! surely 13 :tb1 is a better way of doing
21 'it'xb7 h S ! Black had a kingside ini­ this. Instead 13 ii..f 2 can be met with
tiative in L.Van Wely-A.Shirov, Tunja 13 .. .fs, while after 13 g4 tlJf7 it is hard
1989) 14...tlJf4 1 S ii.fl (or 1S tlJf1 fS) for White to come up with a good plan.
1S ... fs 16 exfs tLJxfs (16 ...ii..x fs! ?) 17 If the bishop moves from h4, then
tlJde4 tlJd4 and Black has counterplay. ... ii..h6 is possible, but if the bishop
c2) 13 .. .fs 14 exfs (14 f3 f4 1S ii..f2 cs maintains its post, then White cannot
also gives Black sufficient play) play h4 and Black can play ...ii..d7,
14 ... tlJhxfs 1S ii..h s (to prevent ...l:tb8, ...a6, ...tlJc8, (now f6 is protected),
1S ... tlJg6) 1S ...tlJd4 16 f3 bS!? (a creative and ...ii..h6. So White should either ini­
idea; instead both 16 ... cs 17 tlJde4 h6 tiate his queenside play immediately or
and 16 ... ii..fs 17 tlJde4 tlJg6 look good else change nature of the position:
enough) 17 cxbs l:tb8 18 tlJde4 tLJxbs 19 a) 13 .l:f.b1 has been played fre­
'i'd2 (19 tLJxgs .l:f.fs) 19 ...h6 was unclear quently by Gelfand:
in A.Onischuk-V.Bologan, Poikovsky al) 13 ...tlJf7 14 o-o fS 1S b4 b6 16
2004. a4! ? ii..h6 17 ii..f 2 ..t>h8 18 as tlJg8 19
'it'c2 tlJf6 20 .l:f.b2 tlJhs 21 .l:f.a1 ii..d7 22
axb6 axb6 23 l:txa8 'it'xa8 24 .l:f.a2 'it'd8
2s bxcs bxcs 26 .l:f.a7 and White had
maintained the initiative in B.Gelfand-
1.Nataf, Cannes (rapid) 2002.
a2) 13 ... gs 14 ii..f2 fs 1S b4 b6 16
bxcs bxcs 17 h3 (17 o-o tlJg6) 17 ...tlJg6
18 g4 fxg4 19 hxg4 l:LJM and Black had
counterplay in B.Gelfand-L.Dominguez,
FIDE World Championship, Moscow
12 gs
... b) 13 dxc6 is also critical, changing

1 77
A ttacking C h e s s : Th e King 's I n d i a n, Vo l u m e 1

the contours of the position. Black's him) and now:

centre may look imposing, but White is a) 16 hxg4 ll'lh4 (or 16 ... ll'lf4 intend­
not really threatened by it and he will ing ... cs) looks unclear. Black can play
have good squares for his pieces after ... a6 and _J:tb8 and maybe ... l::tf4, even
13 ...bxc6 14 b4 ii.e6 and then : as an exchange sacrifice.
b) 16 fxg4 ll'lf4 17 l::th2 cs 18 ll'lfl a6
19 ll'le3 l::tb8! 20 ii.fl ii.d7 was unclear
in J.Gustafsson-V.Nevednichy, Sarajevo
2010. White has a space advantage,
but his king does not have an ideal

bl) 1S ll'lb3 dS 16 ll'lcs (instead 16

cxds cxds 17 exds ll'lxds 18 ll'lxds
ii.xds 19 o-o ll'lfs 20 ii.f2 ii.h6! was to­
tally fine for Black in B.Gelfand­
T.Radjabov, Astrakhan 2010) 16 ... ii.f7
17 0-0 d4 18 ll'l3a4 d4 18 ll'l3a4 ll'lc8 19
ltJ b3 ll'ld6 20 ll'las 'it'e8 21 ll'lcs was 14...ll'lg6
pleasant for White in G.Giorgadze­ This is more flexible than 14... g4 1S
T.Shaked, Linares 1997. fxg4 ll'lxg4 16 ii.xg4 fxg4 17 o-o ll'lg6
b2) 1S ii.f2 !? dS 16 ll'lb3 ! (16 cxds when 18 cxd6 �ess clear is 18 ll'lc4
cxds 17 exds ll'lxds 18 ll'lxds ii.xds 19 dxcs!? 19 ii.xcs l::txf1+ 20 'it'xf1 b6 21
o-o ll'lfs 20 ll'lb3 is Gelfand-Radjabov ii.e3 ii.a6) 18 ... cxd6 19 ll'lc4 looks a little
above) 16 ...d4 17 ll'la4 was V.lvanchuk­ better for White.
T.Radjabov, Monaco (rapid) 2007, when 15 g3
White's position is easier to play. White prevents the knight from
14 C5 coming to f4, but he can also allow it by
Instead 14 o-o ll'lg6 1S cs will trans­ playing 1S 0-0 ll'lf4 16 cxd6 cxd6 17
pose to 1S o-o below, but White does ll'lc4 .U.f6 !? (17 ... g4 18 fxg4 ll'lxe2+ 19
have another option in 14 h3 ll'lg6 1S 'it'xe2 ll'lxg4 is also possible) 18 a4 l::t g 6
g4 fxg4 (after 1S ... f4 White will have a 19 exfs ll'lxfs (not 19 ...ii.xfs 20 g4!,
free hand on the queenside, while crippling the h6-knight, S.Shipov­
1S .. .fxe4 1S ll'ldxe4 looks very good for D.Anagnostopoulos, Cappelle la Grande

1 78
G ligo ric Va ria tio n : 7 ii.. e3 tlJ g 4 8 ii.. g 5 f6 9 ii.. c1 tlJc6 a n d 9 ii.. h 4 tlJ c 6

1997) 20 ii..d3 .l:f.h6 21 ii..xfs ii..xfs 2 2 b) 18 c6! ? and now 18...a6 is cer­
�4 .l:f.g6 23 g4 and here 23 ...ii..xe4 24 tainly possible, but more enterprising
fxe4 was a little better for White in is 18 ... 'it'c8 when 19 ii..xg s ii.. g 2 20 .l:f.g1
S.Shipov-M. Pavlovic, Athens 1997. but �3 21 'it'c2 'it'xh2 22 o-o-o 'it'xg3 is
Black could consider 23 ...ii..c8! ? with the very murky and 19 tlJc4 tlJg4 20 'it'd2
idea of ...h s. tLJxe3 (20 ...h 6 20 ... ..t>h8) 21 tLJxe3 (21
1s ...g4 'it'xe3 'it'd8 regroups) 21 ... a6! ? is un­
Black has an interesting alternative clear.
in 1S ...fxe4 16 fxe4 (this has always 16 fxg4
been played; after 16 tlJdxe4! ? Black After 16 exfs gxf3 17 ii..xf3 tlJxfs
can play 16 ... tLJfs, which should be Black activates his pieces and has no
somewhat better for White, or else sac­ problems.
rifice a pawn with 16 ... ii..h 3 17 ii..e 3! 16...tLJxg4
0ifs 18 ii..xg s 'it'd7 when he has at least Black has also tried sacrificing a
activated his pieces, but it is not clear if pawn with 16 .. .f4, but I have trouble
it is enough for a pawn) 16 ... ii..h 3 17 believing that this can be sound after
�e3 b6 and here: 17 tlJc4.
17 ii..xg4 fxg4 18 o-o

a) 18 cxd6 cxd6 19 tlJc4 'it'd7 20 'ti'b3

(after 20 ii..xg s bS 21 tlJe3 b4 22 ii..x h6 White can also delay this move: 18
ii..x h6 23 ii..g4 't\Vb7! Black has serious tlJc4 (18 cxd6 cxd6 19 tlJc4 tlJh8 trans­
counterplay, and here 21 tlJd2 b4 also poses, but here too Black could con­
gives Black the initiative) 20 ...tlJg4 21 sider 19 ... b6) 18...tlJh8 (possibly not
ii.. x gs tlJf2 22 .l:f.g1 ii.. g 4 23 ii.. e 3 ii..xe2 24 best; again there is 18...b6!?, while an-
ii..xf2 was Y.Shulman-B.Socko, Saint other possibility is 18...ii..d7! ? with the
Petersburg 1997. Now 24...ii..d3 ! ? 2 5 idea of 19 cxd6 cxd6 20 tlJxd6?! .l:f.xf2 !)
.l:f.d1 ii..xc4 26 'it'xc4 ii..h 6 would give 19 cxd6 cxd6 20 o-o tlJf7 21 ii..e 3 ii..d7
Black good play for the pawn. 22 a4 when White's space advantage

1 79
A ttacking C h e s s : Th e King 's I n d i a n , Vo l u m e 1

brings him a slight but pleasant advan­ 26 a4 ii.h6 27 tllb s ii.xd2 28 tllx d2
tage. Indeed, he maintained an edge After 28 '1i'xd2 ii.xbs 29 axbs '1i'f6,
after 22 ... ii.f6 23 '1i'e2 24 'lth1 'ii' e7 with ideas like ... '1i'g6 and ...tllf7-g S,
2S tllb s ii.xe3 26 tllx e3 in J.Ehlvest­ Black has no problems.
L.Van Wely, Groningen 1993. 28... a6 29 tllc 3 bS 30 axbs axbs 31 tllf1
18 ... b6!? Or 31 tll x bs '1i'b8.

This move forces White to resolve 31 ... 'ii'cs 32 'ltg2 'ii'cs 33 tll e 3 hs
the queenside tension while preparing Black has gradually taken over the
a possible ... ii.a6. Instead 18 ... tllh 8 19 initiative and soon won in C.Csiszar­
cxd6 cxd6 20 tllc4 leads back to the P. S inkovics, Hungarian League 1996.
previous note.
19 cxd6 e2) 11 hs

White could also try 19 c6, although

this limits his queenside possibilities.
Black could continue with 19 ... ii.a6, but
this might tempt White into sacrificing
the exchange with 20 '1i'xg4!?. Instead
after 19 ... hs White could play 20 .lii . e3,
so perhaps the best move is 19 ...ii.h6 ! ?.
19 ... cxd6 20 ii.e3 ii.d7 21 '1i'e2 '1i'e7 22
tllc4 l:txfl+ 23 l:txfl l:tf8 24 l:txfB+ '1i'xf8
25 ii.d2

Fedorov's favourite move is ambi­

tious but quite risky. Black hopes to
gain some space on the kingside before
retreating his knight, but now White
can open the h-file, which can lead to
trouble for Black if he is not alert. On
the plus side, White's bishop blocks the
h-file for the time being and some of
White's dark squares on the kingside
will be weakened if he prises open the
2s ... tt:Jhs! kingside. Whether or not Black can
Now that he has developed his make use of this is a different matter.
pieces and consolidated his position a 12 h3 tllh 6 13 g4!
bit, Black begins a standard manoeu­ This is the only real test of Black's
vre; the knight is bound for gs. ambitious play.

G l ig o ric Va riatio n : 7 il e3 tlJg4 8 il g 5 f6 9 il c1 tlJ c 6 a n d 9 il h 4 tlJ c 6

13... hxg4 14 hxg4 tlJf7 1999) 17 ... l:te8 18 tlJe3 looked quite
threatening in J.Ehlvest-V.Spasov, Ko­
caeli 2002, but actually here Black
could try to create some counterplay
on the queenside, starting with
18...'ii' a s!?.
b) lS tiJfl cs 16 tlJe3 (Golubev also
mentions 16 '1i'd3 1'..d7 17 '1i'h3 l:te8 18
'1i'f3 tlJc8 19 tlJe3 a6 20 a4, but while
White may have some advantage, this
does not look fatal for Black; his king is
safe enough for now and he can try to
Black would love to play ... 1'.. h 6, prepare ... bs) 16 ... 1'..d7 17 1'..d3 was
...'ltg7, and ...�h8, but that is not so Z.Azmaiparashvili-A.Fedorov, Elis ta
easy because White's bishop attacks Olympiad 1998, and here Fedorov sug­
the f6-pawn. Therefore Black needs to gests 17 ...tlJc8 with the idea of ... il h6.
move his clumsy e7-knight, but this too 15 ...c5
is not that easy. The typical plan for After 1s ...1'..d7 16 o-o-o tlJc8 17 ilg3
Black is ... cs, ...1'..d7 and ...tlJc8, but il h6 White has a thematic device
while Black is making these manoeu­ available.
vres White can start feeding his pieces
towards the kingside.
15 '1i'c2
This move prepares castling queen­
side and is the most popular, but oth­
ers are dangerous too:
a) lS '1i'b3 also prepares castling,
but from here the white queen may
slide along the third rank to threaten
the black king: 1s ... cs 16 tiJd1 ! ? 1'..d7
(Black could also consider the immedi­
ate 16 ...l:te8 or an evacuation plan Indeed, 18 f4! (much stronger than
starting with 16 ... tiJh8) 17 '1i'h3 (17 18 l:th2 'ltg7 19 z:tdh1 l:th8 20 'ltb1 V2-V2,
'1i'xb7 l:tb8 18 '1i'a6 tlJc8 19 b3 il h6 20 E.Magerramov-A.Fedorov, Dubai 2001)
�b2 'ltg7 21 o-o-o '1i'C7 gave Black 18 ... 1'.. xf4 19 ilxf4 exf4 20 es! (the g6-
some compensation for the pawn due pawn is the target) 20... tZJxes 21 l:th6
to the clumsy position of the white '1i'e7 22 �dhl '1i'g7 23 CZJf3 tlJe7 24
queen in J.Ehlvest-A.Fedorov, Calcutta tZJxes fxes (24... dxes 2S cs!) 2S 1'.. d3

1 81
A ttacking C h ess: Th e King 's I n d i a n , Vo l u m e 1

ii.xg4 26 ii.xg6 gave White a winning g6-pawn, so Black must consider

attack in F. Peralta-V.Spasov, Lorca 17 ... a6, which could transpose to the
2004. note to White's next move) 18 ii.g3
16 o-o-o ii.h6 19 f4! (instead 19 lldhl 'ltg7 20
g S ! ? fxgs 21 'ltb1 gives White compen­
sation, but this is less challenging), and
now Black has problems:
bl) With 19 ... exf4 Black wants to
keep the bishops on the board, but this
fails tactically: 20 es 'ltg7 21 exf6+
'1i'xf6 22 tllce4 '1i'e7 (or 22 ...'ii' es 23 ii.h4
with a winning advantage in F.Peralta­
A.Suarez Real, Leon 2008) 23 ilh4 gs 24
ii.xg s ii.xgs 2s �h7+ 'ltxh7 26 tllxg s+
with a mating attack in T.Enkhbat­
16 ... a6 Y.Shulman, Connecticut 2002. An im­
By playing this immediately, Black is pressive attack especially considering
ready to play at any time. The al­ that Shulman is a strong proponent of
ternative is to play 16 ... ii.d7 first, fol­ the Gligoric with White!
lowed by either ...a6 or ... tllc8. Some b2) 19 ...ii.xf4 20 Lf4 exf4 21 es
examples: tllx es 22 l:i.h6 '1i'e7 23 �dhl '1i'g7 24 tl:Jf3
a) 17 �dfl and here: gives White a strong attack similar to
al) 17 ... a6 18 �h 2 tllc8 19 ii.g3 ii.h6 Peralta-Spasov above.
20 f4 exf4?! (an exchange of mistakes;
Black should play 20 ... ..txf4) 21 ii.xf4?!
(21 es!) 21 ... .Lf4 22 �xf4 was
B.Avrukh-E.lnarkiev, Moscow 2002.
Here 22 ... tlle s looks solid enough.
a2) 17 ... tllc8 18 ii.g 3 ii.h 6 19 f4
..txf4! 20 ii.xf4 exf4 21 l:i.xf4 tlle s (but
not 2 L .'ltg7? 22 �hf1 tlles 23 g s ! fs 24
exfs ii.xfs 2s tllce4 tllb6 26 '1i'c3 with a
big advantage in A.Huzman-V.Sara­
vanan, Biel 2000) 22 l:i.hfl gs 23 �4f2
'ltg7 and Black is okay. If White protects 17 �dfl
the g4-pawn with a rook, Black will be Instead 17 l:th2 can be met by the
able to play ... tllc 8-e7-g6. immediate 17 After 18 �dhl Black
b) 17 �h2 tllc8?! (this abandons the might try 18 ...b4 or 18 ... ii.d7, with the

1 82
G l ig o ric Va ria tio n : 7 il e3 tlJ g 4 8 il g 5 f6 9 il c 1 tlJ c6 a n d 9 il h 4 tlJ c6

idea of 19 ilg3 '1i'c8 when the pressure Black is very solid.

on g4 will deter White from playing f4 18 ...1'.. h 6 19 f4 b4 20 tiJd1 exf4 21 1'..xf4
for a while. If then 20 f3 b4 21 tiJb1 as 1'..xf4 22 l:txf4 'ltg7
with counterplay. Instead, long after I
had first drafted this section, 19 'ltbl
was seen in B.Gelfand-H.Nakamura,
Amsterdam 2010 (Nakamura had
opted for the move order 17 ...1'.. d7 18 bS), and after 19 .. �b8 20 1'.. g 3
'1i'c8! (again we see this instructive
idea) 21 f3 �b7 22 'ital '1i'e7 23 f4 one
way to obtain counterplay was
23 ... exf4!? 24 1'..xf4 b4 2S tiJdl as, but
Nakamura preferred 23 .. �fb8 24 fxes
dxes!?, maintaining a tough, unbal­ Black has achieved everything he
anced and approximately level middle­ could hope for and stands well. After
game. the further moves 23 �f2 l:th8 24 l:tfh2
11 bs 1s 1'..g3
... �xh2 2 S �xh2 1'..d7 26 '1i'd3 tlJg8 27
Fedorov suggests 18 tiJd1 ! ? with un­ CZJf2 tlJgh6 28 '1i'g3 '1i'e7 29 tiJf3 �e8 30
clear play. One possibility is 18 ... b4 19 '1i'f4 as 31 �hl a4 32 1'..d 3 a3 33 b3 �h8
ilg3 1'.. h6 20 f4 ilxf4 21 1'..xf4 exf4 22 Black's initiative had become quite se­
l:i.xf4 'ltg7 (but not 22 ...tZJes 23 'llt'b 3 ! rious in B.Gelfand-A.Fedorov, Batumi
with the idea of 'ii'h 3) 23 �hfl tZJgs and 1999.

1 83
Cha pter 10
Gligoric Variation

1 i.e3 lt:Jg4 s f6 g i.. h 4 gs

1 d4 tllf6 2 c4 g6 3 tll c 3 ii.g7 4 e4 d6 S

tllf3 0-0 6 ii.e2 es 7 ii.e3 tllg4 8 f6
9 ..th4 gs

A: 11 cs
B: 11 h3
This somewhat ugly move really C: 11 dxe5
makes the g7-bishop look bad, but its D: 11 dS
counterpart on g3 will have rather lim­
ited scope as well and it may even be­ A) 11 cs
come a target. This is currently the This aggressive move is quite tricky.
main line of the Gligoric Variation. 11...tllc 6!?
10 ii.g3 tll h6 More common is 11 ...g4 12 tllh4
It is too late for 10 ... tll c6, because af­ tllc6 (White is clearly better after
ter 11 dS tlle7 12 tllxg s! tllxf2 13 ii.xf2 12 ... exd4 13 '1i'xd4 and 12 .. .fs 13 tllxfs,
fxg s 14 ilg4 White is clearly better. but 12 ... dxcs with the idea of 13 dxes
Now White has a broad choice: ii.e6!? could be considered) 13 cxd6 (13

G lig o ric Va r i a t i o n : 7 il e 3 tiJ g 4 8 ilg5 f6 9 il h 4 g5

dxes dxes 1 4 o-o reaches note 'a' to White has tried several moves in
White's 12th move, below) 13 ... cxd6 14 this position:
dxes (of course after 14 ds Black can bl) 15 tiJfs tiJxfs 16 exfs 1'..xfs 17
play 14...tiJd4) and now Black has a ilxg4 tiJd4 with equality in Z.Ksieski­
choice: B.Socko, Glogow 2001.
a) 14 ... dxes looks fairly solid, but b2) 15 tiJbs a6 16 tiJxd6 'ii'a s+ 17
with the c-pawn s exchanged Black 'lt>f1 as played in 5.Slugin-1. Belov, Rus­
cannot control the ds-square. Indeed, sian Team Championship 2009, could
White maintains a pull while Black is be met with 17 ...tiJd4 with compensa­
devoid of counterplay after 15 1'..c4+ tion for the pawn.
ltih8 16 '1i'xd8 l:txd8 17 o-o tiJd4 and: b3) 15 tiJds 1'..e6 16 o-o tiJd4 17 tiJe3
al) 18 f4 gxf3 19 tiJxf3 tiJxf3+ 20 �c8 18 b3 'ii' d 7 19 ilc4 ilxc4 20 tiJxc4
:.xf3 looks great for White. After 20.. .fs 1'..f6 21 f4 '1i'e7 22 fxes dxes 23 tiJfs
21 exfs tiJxfs 22 �afl tiJd6 23 1'..b3 1'..g 4 tiJhxfs 24 exfs h s left Black with the
24 l:i.f6! he was winning in M.Roiz­ initiative in A.Poluljahov-R.Anton­
B.Socko, Austrian League 2009. iewski, Koszalin 1999.
a2) 18 �adl is also good. After b4) 15 o-o!? tiJd4 (worse is 1s ...1'..f 6?!
18 ... 1'..d7 19 f4 l:i.ac8 20 1'..d3 tiJf7 21 16 tiJfs tiJxfs 17 exfs 1'..xfs 18 1'.. x g4,
fxes tiJxes 22 1'..b1 1'..e6 23 1'..f2 tiJec6 24 which was pleasant for White in
�fs White maintained the initiative L.Fressinet-D.Stellwagen, European
deep into the endgame in A.Onischuk- Team Championship, Novi Sad 2009,
1.Smirin, World Team Championship, but Black should consider 1s ... 1'.. e6) 16
Bursa 2010. .�c4+ 'lth8 17 tiJe2 1'.. e6 (White also re­
b) 14.. .fxes is more combative, but tains an edge after 17 ... tiJxe2+ 18 1'.. x e2
it's certainly ugly. Black's justification 1'..f6 19 tiJfs tiJxfs 20 exfs 1'..xfs 21
for his strange pawn structure is the 1'..xg4) 18 1'.. x e6 tiJxe6 19 '1i'd2 tiJf4 20
floating knight on h4. l:i.adl l:tc8 21 f3! and White had the
initiative in A.Giri-M.Vachier Lagrave,
Biel 2010.
bS) 15 1'..c4+ is the most obvious
move and is probably best: 1s ... 'lth8 16
'1i'd2 tiJd4 17 o-o (White also kept an
edge after 17 tiJe2 tiJxe2 18 1'..xe2 1'.. e6
19 o-o '1i'e7 20 b3 �ad8 21 l:i.adl in
A.Onischuk-E.lnarkiev, Poikovsky 2009)
17 ... 1'.. d7 (this is Black's latest try;
11 ...'1i'e7 18 b3 'ii'as 19 �fd1 1'..d7 20
tiJe2 '1i'xd2 21 l:i.xd2 also left White with

1 85
A ttacking Ch ess: Th e King 's I n dia n , Vol u m e 1

a slight edge in E.L'Ami-G.Agamaliev, a) 13 o-o g4 14 tllh4 ii.e6 lS '1i'a4

Hoogeveen 2008) 18 tlle2 �c8 19 l:tacl '1i'd4! gave Black good play in
�xc4 (Black tries to solve his problems V.lvanchuk-T.Radjabov, Odessa (rapid)
tactically, but White's positional ad­ 2007.
vantage persists) 20 �xc4 tllxe2+ 21 b) 13 ilc4+ 'lth8 14 '1i'xd8 tllxd8 lS
'1i'xe2 .tbs and Black will win back the o-o-o (lS tllb s ii.e6 was also satisfac­
exchange. tory for Black in the game Y.Shulman-
However, White is still better: 22 1.Umanskaya, St Petersburg 1994)
�dl (more complicated but still favour­ 1S ... tlle6 16 b4 as 17 a3 axb4 18 axb4
ing White is 22 '1i'c2 ii.f6 23 tllfs tllxfs c6 19 'ltb2 tllf4 20 ii.xf4 exf4 21 'ltb3
24 exfs h S 2 s f3 gxf3 26 l:txf3 ii.xc4 27 1/2-1/2, P.San Segundo Carrillo-A.Romero
'1i'xc4 dS 28 '1i'e2 e4 29 �fl, S.Slugin­ Holmes, Vendrell 1996.
Lyaskovsky, Russian Team Champion­ c) 13 '1i'xd8 is the most dangerous
ship 2010) 22 ...'1i'd7 23 �dcl '1i'e6 24 a4 move according to Bologan, who now
Lc4 2S '1i'xc4 '1i'xc4 26 J:1xc4 ii.f6 27 gives 13 ... �xd8! (rather than 13 ... tllxd8
'ltfl and White eventually won the 14 tllb s �f7 lS o-o-o ii.e6 16 h4 g4 17
ending in S.Volkov-M.Krylov, Moscow tllg S! when White was much better in
2010. V.Grabliauskas-C.Krogh, Ringsted 1992;
12 cxd6 here 1S ... tlle6 16 ilc4 is unpleasant but
After 12 dxes dxes we reach a posi­ may be a better try) 14 tlld s g4 l S tlld2
tion that is often reached via the move tlld4 16 tllxe7 tllc2+ 17 'ltdl tllxal 18
order 11 dxes dxes 12 cs tllc6, al­ ilc4+ (no better is 18 tllxa8 ii.e6)
though in that case 12 ... ii.e6 is another 18 ... tllf7 19 tllx a8 ii.h6 20 ii.ds ii.d7 21
viable option for Black. Here Black's tllc7 llc8 22 tlle6 ii.xe6 23 ii.xe6 �xcs
knight is already on c6, but he can still when Black has equal chances in a
hold the balance. complicated endgame. Note that 24
ii.xg4? fails to 24...�c2 with the initia­
12 cxd6 13 dxes

Instead 13 dS just helps Black for­

mulate a plan:
a) 13 ... tlle7 14 o-o tllg 6 lS �el fS 16
exfs tllxfs 17 tll d 2 tllf4 18 ii.fl g4 19
�cl h S ! and Black had already seized
the initiative in E.Levin-S.Klimov, St Pe­
tersburg 2010.
b) 13 ... tlld4!? is also possible: 14
White has: tllxd4 exd4 l S tllb s (not l S '1i'xd4 fS)

Glig o ric Varia tio n : 7 il e 3 tlJ g 4 8 ilg5 f6 9 il h 4 g5

lS.. .fS! 16 .i.xd6 fxe4 (worse is 1 6. . .l:i.e8 stead 20 'ltxg2 looks critical, but Black
17 es!) 17 1'.. xf8 '1i'xf8 with some com­ has enough play after 20 .. .fs. One pos­
pensation. Here 18 tlJxd4? fails to sible line is 21 es f4 22 Lf4 1'.. e 6 23
18 ...'llt'b 4+. ilg3 CZJfs 24 '1i'e4 tZJxg3 2s fxg3 'ii'g s 26
13 ...dxes �xf8+ �xf8 27 �el bS! 28 il b3 (not 28
1'.. x bs '1i'd2+ 29 �e2 .i.xds) 28 ...'1i'd2+ 29
�e2 il fS! 30 '1i'e3 (30 �xd2 1'..xe4+ with
the idea of ...1'.. x es is equal) 30 ...'1i'xe3
31 �xe3 �e8 32 tiJC7 �xes 33 �xes
1'.. x es 34 tlJxbs and with such limited
material on the board, Black's good
bishops should ensure a draw.
20 ...1'..g4 21 'ltxg2
White sacrifices an exchange to
fight for the initiative. After 21 �d3 or
21 l:td2 Black has counterplay with
14 1'..c4+ 21 ... fs.
White should seize the diagonal 21... .i.xdl 22 l:i.xdl fS 23 es
while he can. Black had no probl ems
after 14 o-o 1'.. e 6 in P.H.Nielsen­
E.Mortensen, Festuge 1991.
This is better than 14... CZJf7 lS h3!
when Black is in a pin and has commit­
ted his knight too early (but lS o-o ilg4
successfully fights for the d4-square).
15 h3
White prevents ...1'..g4, so Black
must find a way to create some coun­
terplay. The position is unclear, but fairly
1s ... tlJd4 16 o-o g4! 17 tlJxd4 exd4 18 balanced. White has a strong central
tZJds presence, but his king is not completely
Instead 18 tlJe2 gxh 3 19 '1i'xd4 was comfortable and Black is up an ex­
P.San Segundo Carrillo-M.Pavlovic, change for a pawn.
Ubeda 1996, and now the simplest is 2 3 ...f4!?
19 ...'1i'xd4 20 tlJxd4 fS with counterplay. other interesting possibilities are
18 ... gxh3 19 '1i'xd4 hxg2 20 �fdl 23 ... �g8, 23 ...�c8 and 23 ...tlJg4.
The position is very complicated. In- 24 CZJxf4

A ttacking C h e s s : Th e King 's I n d i a n , Vo l u m e 1

Instead 24 ii.xf4 '1i'h4 2 S ii.g3 '1i'xd4 dxes (I do not like 12 .. .fxes 13 cs!: for
26 �xd4 tllfs is similar. example, 13 ... dxcs 14 '1i'xd8 followed by
24...'1i'xd4 2 s �xd4 tllfs 26 �e4 tllxg3 lS tllx gs) and now:
Black could also play 26 ... �ac8. a) 13 '1i'xd8 tllxd8 14 tllds tlle6 lS h4
21 'ltxg3 z:tfs 28 tlld3 �gs+ 29 'lth2 �fB (otherwise, lS tlle 7+ 'ltf7 16 tllxc8
30 f4 l:tg6 31 .tbs �dB 32 ii.c4 l:tfs 33 l:taxc8 is level, while lS o-o-o l:i.e8 16 z:tds 34 ii.c4 h4, as in G.Timoscenko-E.Hagara, Slo­
Now Black declined a repetition of vakian League 2008, could be met with
moves against his higher-rated oppo­ 16 ... g4, intending ... c6 and then either
nent by playing 34... a6! ? in ... tlld4 OT... tllc s) 1S ... g4 16 tllh 2 tlld4
J.Gustafsson-A.Shimanov, European (16 .. .fs also gives Black counterplay) 17
Championship, Rijeka 2010, and even­ tllxo �b8 18 ii.d1 fs 19 exfs tllhxfs 20
tually went on to win. tllxg4 tllx g3 21 fxg3 e4! and Black had
the initiative in S.Emst-D.Stellwagen,
B) 11 h3 Dutch Championship, Leeuwarden
b) 13 o-o ii.e6 (the solid 13 ...'ii'xdl
14 l:i.fxdl ii.e6 also looks playable) 14
'1i'c1 '1i'e8 lS tllds l:i.c8 16 b4 (Yer­
molinsky instead suggests 16 !tel g4
17 hxg4 ii.xg4 18 tlle 3 ii.xf3 19 ii.xf3
tlld4 20 ii.dl '1i'g6 21 ii.c2 which looks
unclear) 16 ... g4! 17 hxg4 ii.xg4 18 bS
ii.xf3 19 ii.xf3 tlld 4 20 '1i'a3 'lth8 21
l:i.adl (or 21 '1i'xa7 fS 22 exfs tllhxfs
with counterplay) 21 -.fS and Black had
This move looks a little funny be­ kingside counterplay in A.Yermolinsky­
cause Black usually voluntarily retreats J.Fedorowicz, US Championship, Los
his knight in any case, but it prevents Angeles 1991 .
... ilg4 and the advance ... g4 is discour­ 12 ... tlle 7
aged because it would open the h-file The speculative 12 ...tlld4 13 tllxd4
for White. exd4 14 '1i'xd4 fS lS '1i'd2 f4 16 ii.h2
11 ...tllc6 tllf7 has even been tried by Kasparov
Both 11 ... exd4 and 11 ... tt:Jd7 are pos­ (in a rapid game), but it is probably
sible as well, but the text move fits in objectively unsound.
best with our repertoire.
12 dS (seefollowing diagram)
White has an alternative in 12 dxes

1 88
Gligoric Varia t i o n : 7 11.. e 3 tlJ g 4 8 11.. g s f6 9 1'.. h4 g5

White managed to win. Instead the

flexible 17 ... 1'..d7 is natural, while
17 ...l:tb8 and 17 ...'ii'a s!? are also possible.
14 ...f5 15 exf5
After 1S f3 Black could force the
pace with 1S .. .f4 or maintain the ten­
sion with 1s ... tlJg6.
15 ...tZJxf5 16 o-o tiJh8
Black repositions his knight. Similar
is 16 ...tlJd4 17 1'.. g 4 tiJh8 when 18 tlJe2
CZJfs 19 11..h s was S.Shipov-V.Golod,
13 '1i'd2!? Minsk 1993, and here Black could con­
An independent try. After 13 tiJd2 sider 19 ... tlJg6 20 1'.. xg6 hxg6, intending
we tran spose to 9 ... tlJc6 10 h3 tiJh6 11 ... e4! ? with counterplay.
dS tlJe7 12 tiJd2 gs 13 1'..g3 which was 11 1'.. h 5 CZJd4
discussed in note 'a2' to White's 10th Black does not really have to rush
move in Line B of the previous chapter. this move and might prefer 17 ... 1'.. d7 ! ?.
13 ... tiJf7 18 tlJe4
After 13 ...tlJg6 ! ? 14 o-o-o White has White could also consider 18 tZJe 2.
scored well, but following 14... a6 1S Then 18 ...tZJfs would leave Black a
'iit b 1 1'.. d 7 the position is not clear. tempo down on the note to his 16th
14 tiJh2 move, or he could play 18 ... tZJxe2+ 19
Or 14 1'..h 2 cs 1S 1'..g 1 a6 16 tiJh 2 fS '1i'xe2 11..fs, intending ... tlJg6 and maybe
17 CZJf1 which looks rather comical. even ... e4. White has a grip on the light
squares, but his g3-bishop is not much
better than its g7-counterpart.
18 ... h6 19 �fe1 1'..f5 20 �acl tlJg6

Now 17 ...fxe4 18 1'..g 4 CZJfs 19 tZJxe4

bS looked okay for Black in S.Volkov­
A.Shomoev, Ulan Ude 2009, although

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

Now White played 21 cs? dxcs 22 12th in Line A, but here Black can also
l:i.xcs b6 and Black won a pawn in consider 12 ...ii.e6 ! ? covering the a2-g8
U.Boensch-B.Gelfand, Dortmund 1990. diagonal.
Instead 21 ii.xg6 ii.xg6 leaves Black A sharp try is 12 h4!?, but Black has
better with his bishop-pair, so White enough resources: 12 ... g4 13 tllh 2 tllc6
should settle for 21 ii.g4 with approxi­ 14 '1i'xd8 �xd8 1s tlld s tlld4! (Black has
mate equality. serious counterplay) 16 l:i.c1 (instead 16
ii.d1 fS was Li Wenliang-F.Nijboer,
C) 11 dxes Groningen 2002, but 16 ...ii.d7 ! ? looks
White exchanges in the centre, simple and good) 16 .. .fS! ? (or just
which often leads to an exchange of 16 ... c6 17 tlle7+ 'ltf7 18 tllxc8 tllxe2 19
queens. 'ltxe2 .!:i.axc8 with equality) 17 exfs (17
11 ...dxes tllxa �b8 18 exfs tllhxfs also gives
Black good play) 17 ...tllhxfs 18 ii.xg4 c6
19 tlle3 tllx g3 20 fxg3 ii.h6 and Black
had the initiative in l.Gulkov­
S.Dyachkov, Voronezh 1999.

Cl) 12 '1i'xd8 �xd8

Now White has two main continua­

tions. He can exchange queens imme­
diately with 12 '1i'xd8 or play a little
finesse first by going 12 'ii'd s+.

C1: U �8
cz� u trds+ 13 tlld s
This allows White to gain the
Instead 12 h3 tllc6 transposes to the bishop-pair, but that does not trouble
notes to White's 12th in Line B, while Black Instead 13 tlld2 ii.e6 14 o-o-o
12 o-o c6 is level and 12 '1i'b3 does not tlla6 1S f3 c6 16 tllb3 tllf7 17 ii.f2 ii.f8
achieve much After 12 ... c6 White can­ 18 �xd8 tllxd8 19 l:i.d1 b6 is equal and
not castle long. Note too that 12 cs tllc6 after 20 tll a1 tllb7 21 tll c2 �d8 22 l:i.xd8
was covered in the notes to White's tllxd8 the players agreed to a draw in

G l ig o ri c Va ria tion : 7 ile3 tlJ g4 8 ilg5 f6 9 il h 4 g5

V.lvanchuk-V.Bologan, Foros 2006. advantage in G.Timoscenko-E.Pinter,

13 ...tlJa6 Hlohovec 1998. The point is th at
More ambitious is 13 ... �d7, but this 19 ...l:i.xd2 20 1'..x cs+ is check.
is obviously riskier. Black clogs up his 15 tZJxc8 �axes 16 �cl
own development and the g4-pawn is White wants to safeguard his
weakened. Moreover, he should not queenside. The point of Black's 14th
fear the loss of his light-squared bishop move i s seen in the line 16 a3?! tZJcs 17
because he will get counterplay on the tiJd2 l:i.d4 18 f3 �cd8 19 �dl (after 19
dark squares. 1'..f2 �xd2 20 1'..x cs it's not check and
14 0R7+ Black can just play 20 ... �xb2) 19 ... tlJa4!
Instead 14 o-o-o 1'..e6 is similar to when White's position falls apart.
lvanchuk-Bologan, above. White tried 16 �dl tiJcS 17 tiJd2 as
18 b3 1'..f8 19 o-o in R.Akesson-G.Jones,
Klaksvik 2008, and now 19 ... �d4 20 f3
�cd8 2 1 tiJbl tlJe6 would be good for
Black because of his grip on the dark
16 ... tZJcs
The immediate 16 ...1'..f8 also looks
very good. The bishop threatens a dis­
ruptive check on b4. Now 17 0-0 would
be met by 17 ... g4 18 tiJel �d2, while 17
a3 tZJcs also favours Black.
14 '>tf7!
.•• 11 tiJd2 as 18 f3 .i.fB 19 1'..f2 �d7
Black's king will either have to take
a square away from his knight (on f7)
or his bishop (on f8). From a positional
standpoint, it makes more sense to go
to f7 because the h6-knight will not
necessarily be in a hurry to go to f7, but
the g7-bishop will almost certainly
want to go to f8. There is also a tactical
point, as we shall see.
That said, 14...'lt>f8 is also playable:
15 tZJxc8 �axc8 16 a3 tZJcs 17 tiJd2 �d4
(instead Black should play 17 ... �d7 Black is ready to double rooks on the
with equality) 18 f3 l:i.cd8 19 1'..f2 ! (but d-file and to increase his grip on the
not 19 �dl? tlJa4!) gave White a slight dark squares with ... b6, ... tlJe6 and

A ttacking C h e s s : The King 's I n dian, Vo l u m e 1

... ii.cs. White cannot even castle, so he 12 'lthB


tries to stir up some counterplay. Another possibility is 12 ... tllf7 13 h4

20 h4 g4! 21 ..11i. e3 (or 13 o-o-o 'ii' e7 14 h4 c6 lS 'ii'a s tlla6
Black has a strong initiative after 21 16 hxg s fxgs, reaching the same posi­
fxg4 l:i.cd8 22 �c2 tlld 3+ 23 ..11i.xd3 l:i.xd3 tion) 13 ... '1i'e7! (13 ... g4 14 tllh 2 hS lS f3
or 22 ..txcs ..txcs. wins a pawn) 14 hxgs fxgs lS o-o-o c6
21 �cdB 22 �c2 gxf3 23 gxf3 b6 24
•.. 16 'ii'a s tlla6 17 tlle1 tllcs 18 '1i'a3 �e8
�gl with unclear play.
Now Black erred with 24 ... tlle 6? 2S 13 0-0-0
cs! which unleashed White's light­ Instead 13 cs?! c6 14 '1i'xd8 �xd8 lS
squared bishop in M.Roiz-1.Smirin, Tel tlld2 ..11i.f8 16 tlla4 was V.Korchnoi­
Aviv 2002. J. Fedorowicz, World Team Champion­
Instead 24 ... tlld 3+ 2S ..11i. x d3 �xd3 26 ship, Lucerne 1989, and now 16 ...�d4!
tllfl and here either 26...tllg 8 or 17 b3 tlld7 18 .!:!.cl bS! 19 cxb6 ..11i.b4! 20
26 ...'lte6 are comfortable for Black, l:tc2 axb6 is much better for Black, as
while another idea is 24... tllg8, aiming pointed out by Mikhalevski.
to bring the knight to d4. After 2S tllfl White can also exchange queens
tlle7 Black need not fear 26 ii.xcs bxcs with 13 '1i'xd8 l:txd8, although this
followed by ...tllc6-d4 and ... ..11i. h6 with a should not be dangerous:
pleasant position for him. a) 14 tlld2 cs!? (or 14 ... ..11i. e6) lS f3
tllc6 16 o-o-o tlld4 17 ..11i. d3 ..11i.e6 18 tllb3
C2) 12 'ii' d S+ �ac8 was equal in V.lvanchuk­
T.Radjabov, Sochi 2007.
b) 14 tlld s tlla6 1s o-o-o (White
cannot even grab the bishop-pair here
because, of course, lS tlle7 is not check)
1s ... ..11i. e 6 16 tlld2 (16 a3 c6 17 tllc3 tllcs
18 b4 tllb 3+ 19 'ltc2 tlld4+ 20 tllxd4
exd4 21 tlla4 was S.Volkov-A.Shomoev,
Krasnoyarsk 2007, and here 21...b6
looks best, while Black can vary earlier
as well, perhaps with 17 ... tlle7! ? or
18 ... tlld 7) 16 ... �d7 17 f3 c6 18 tllc3 ..11i.f8
This is White's most ambitious try. 19 tllb3 b6 20 �xd7 ii.xd7 21 ..tf2 ..11i. e6
Now Black does not want to exchange 22 �dl tllf7 and the position was bal­
queens himself, because after anced in E.Perelshteyn-M.Beelby, Tulsa
12 ... 'ii' xds 13 cxds White will have 2008.
pressure on the c-file. c) 14 h4 g4 lS tlld2 and now

G lig o ric Va ria ti o n : 7 il e 3 tlJg4 8 il gs f6 9 il h 4 g5

15 ... cs!? is possible, but in practice tZJbs (22 'Itel 1'.. h 6 23 �dl �xe4 24 1'..fl
Black has preferred to develop his 1'..d7 looks fine for Black) 22 ...�d8 23
queen knight: 'Itel 1'..h 6 24 l:i.dl a6 25 tlJc3 �d4 with a
strange situation. The al-knight is
trapped, but White can hardly win it.
He may exchange it indirectly with 26
'ltbl, but then 26 ... .i.xd2 27 'ltxal?!
ilxc3 28 bxc3 l:i.xe4 29 �d8+ 'ltg 7 cer­
tainly favours Black
13 '1i'e7

cl) 1S ... tlJa6 16 o-o-o 1'.. e6 17 CZJb3

b6! (the inferior 17 ... c6?! 18 tZJas �xdl+
19 l:txdl �b8 20 a3 gave White an edge
in A.Onischuk-A.Shomoev, Sochi 2007,
but Mikhalevski's 17 ...CZJf7 18 tZJas
l.h6+ 19 'ltbl tiJd6 looks solid enough)
18 f4 gxf3 19 gxf3 c6 20 'ltbl CZJf7 21 f4
exf4 22 1'..xf4 fs and Black had enough 14 'ii' as
play in V.Bhat- B.Sambuev, Montreal Slow is 14 h3 and 14... tlJa6 15 'ii' as
2009. c6 was fine for Black in P.Wells-L.Kritz,
c2) 1S ... tlJc6!? (Black logically takes Biel 2004.
aim at the d4-square, but there are Instead 14 h4 g4 15 tiJel tlJc6 16
some tactical issues) 16 tZJds (16 CZJb3 is tZJc2 fs 17 exfs was K.Sundararajan­
possible as well, but the text move is R.Shetty, Visakhapatnam 2004, and
certainly critical) 16 ...tlJd4!? (going for here the obvious 17 ... CZJxfs gives Black
it, but 16 ... l:td7 17 CZJb3 �f7 is safer) 17 counterplay.
!Uxo tZJc2+ (not 17 .. Jlb8 18 �cl with a 14 ...c6
healthy extra pawn) 18 'ltdl tlJxal 19 After the natural 14... tlJc6 White has
tlJxa8 CZJfs! (a clever resource) 20 tZJo 15 tZJds ! tZJxas 16 tZJxe7 1'..e6 17 tZJds
(not 20 exfs 1'..h 6!) was played in l:tac8 18 tiJd2 with an edge,
5.Feller-M.Al Modiahki, Paris 2010. Now E.lturri zaga-5.Feller, Moscow 2010.
Black could justify his play with 15 tiJd2 b6 16 'ii' a4 1'.. d 7 17 f3 cs!?
20 ... tZJxg3 21 fxg3 �d4! (but not Black gives up the ds-square, but he
21...1'..h6 22 tZJds, blocking the d-file) 22 will get d4 for his knight in return.

A ttacking C he s s : Th e King 's In dian, Volu m e 1

18 '1i'a3 tllc6 19 tlld s •t1 20 ii.d3 ii.e6 Black may break immediately or he can
develop first in order to better fight for
the e4-square.

D1: 11...f5

02: 11....�7

D1) 11 ...fs 12 exfs

The position was unclear in

A.Motylev-T.Radjabov, Wijk aan Zee
2007. Black's kingside set-up may look
funny, but he has possibilities to initi­
ate play with ...f5 or ...b5.

D) 11 d 5
This invention of Kasimdzhanov
commits Black to a pawn sacrifice. The
alternative is 12 ... tllxf5 (but not
12 ... ..txf5?! 13 h4 g4 14 tllg 5 ! ) 13 tlld2
tlld4 and here:
a) After 14 tlld e4 g4! (intending ...h 5
with counterplay) White could still play
15 ii.xg4 ii.xg4 16 '1i'xg4 tllc2+ 17 'ltd2
tllxal 18 �xal with some compensa­
tion, although Black is probably not
Closing the position is White's most worse.
popular option. After Black breaks with b) 14 ii.d3 was suggested by Bolo­
.. .f5 White will gain the e4-square for gan. After 14 ... ii.f5 (?!) 15 ii.xf5 l:txf5 16
his pieces. Black will have to either tllde4 White is better, but Black should
fight for control of this square or gen­ prefer 14... tlla6, as in E.Karavade­
erate enough counterplay to offset T.Abrahamyan, Yerevan 2006.
White's grip on the light squares. c) 14 o-o tllx e2+ (14...ii.f5 15 tllde4
Here a decision must be made. tlld7 16 ilg4 was a little better for

G l ig o ric Varia tio n : 7 ile3 tb g 4 8 il g5 f6 9 i.. h4 g5

White in S.Gross-V.Golod, Ceske Bude­ ilxe4 tbfs 23 '1i'd3 hS 24 f3 (or 24 cs as!

jovice 1993) lS '1i'xe2 '1i'e8 16 tl:Jde4 with counterplay) 24 ... tl:Jxg3 2S hxg3
1t'g6 17 f3 ilfs (instead 17 ... tl:Ja6 18 a3 'ii'g s 26 'ltf2 �f6 (simpler is 26...1'..h 6)
g4 19 b4 keeps the knight sidelined) 27 '1i'e3 '1i'xe3+ 28 'ltxe3 1'..h 6+ 29 'ltf2
was Ki.Georgiev-V.Bologan, Gibraltar 'ltg7 30 �h l �h8 31 'lte2 1'..g s?! (Black
2006. Here White could play 18 cs im­ can equalize with 31 ... gxf3+ and then
mediately with the initiative. 32 gxf3 iJ..g S OT 32 iJ..xf3 aS!) 32 fxg4
13 tl:Jd2 hxg4 33 �xh8 'ltxh8 34 a4 'ltg7 3S as
After 13 tl:Jh4? Black wins a piece and White kept an edge in R.Hungaski­
with 13 ... 1'..f6. White does not obtain D.Lemos, Riobamba 2007.
enough compensation after 14 h3
ilxh4 lS hxg4 1'..x g3 16 fxg3 tl:Jf7 with Another idea is 14 ... 'ii'g s lS tl:Je3
the idea of ... 0,g s. i..d7 16 tl:Je4 '1i'g6 17 1'.. d3 tl:Jfs with
13 ... 1'..xfs counterplay. Here Huzman gives 18
tl:Jxg4?! tl:Jxg3 (or 18...hs!?) 19 tl:Jef6+
ilxf6 20 1'..x g6 tl:Jxhl 21 1'..fs ! i..g 7! 22
1'..e6+ (Black is also doing well after 22
ilxd7 tbxd7 23 '1i'e2 hS 24 o-o-o �f4!)
22 ...1'.. xe6 23 dxe6 tl:Jc6 24 'ii'd s hS 2S
e7+ �f7 and because White will not be
able to round up the hl-knight, Black is
much better.
15 tl:Je3

14 tl:Jfl
This move wins a pawn and is
therefore critical. 14 tl:Jde4 is less dan­
gerous, leading to 14 ... tl:Jd7 and then:
a) lS 1'..d 3 'lth8 (1s ...'1i' e8? !) 16 o-o
as 17 '1i'e1 1'..g 6 18 '1i'e2 '1i'e7 was about
equal in L.Gofshtein-R.Kasimdzhanov,
Hoogeveen 1999.
b) lS 0-0, after which the following
play is very thematic: 1s ...'1i'e8 16 1'..d3 1s ...e4
'1i'g 6 17 '1i'c2 'lth8 18 �adl a6 19 b4 tl:Jf6 Th is looks like the best move order.
20 tl:Jxf6 '1i'xf6 2 1 tl:Je4 1'..x e4! (a typical Instead 1s ... tl:Jcs can be played when 16
idea; Black will target the g3-bishop) 22 o-o e4 transposes to the main line, but

A ttacking C he s s : The Kin g 's In dian, Vo l u m e 1

White can also try:

a) 16 b4 e4 17 �cl tlld 3+ 18 ii.xd3
exd3 19 tllxfs tllxfs gives Black some
counterplay: for example, 20 o-o (much
worse is 20 '1i'xg4? '1i'f6 with the idea of
21 o-o d2 22 l:i.c2 tlle 3) 20 ... tt:Jd4 (other
possibilities are 20 ... h s and 20 ...'1i'f6) 21
'1i'xd3 hS, targeting the g3-bishop.
b) 16 tllxfs tllxfs 17 ii.xg4 e4 and
bl) 18 ii.xfs ii.xc3+ (White is much
better after 18 ... tlld 3+ 19 'ltfl l:i.xfs 20 17 l:tcl
'1i'g4 '1i'f6 21 tllxe4 '1i'xb2 22 �dl '1i'c2 23 White lends the c3-knight some ex­
'1i'e2 '1i'xe2+ 24 'ltxe2) 19 bxc3 l:txfs tra protection and creates the possibil­
(Black's compensation looks insuffi­ ity of playing b2-b4. Others:
cient after 19 ... tlld 3+ 20 'ltfl l:txfs 21 a) 17 ii.xg4 is untried, but Black
'1i'g4+ 'ii'g s 22 'ii'x gs+ �xgs 23 'lte2) 20 should have enough for the pawn after
o-o 'ii'g s with some compensation. 17 ... ii.xc3 18 bxc3 tllxg4 (or perhaps
b2) 18 o-o tllx g3 (Huzman prefers 18 ... ii.xg4 19 tllxg4 'ii'g s) 19 tllxfs (19
18 ... ii.xc3 19 bxc3 'ii'g s with some tllxg4 'ii'g s 20 tlle 3 tlld3 gives Black
compensation) 19 hxg3 ii.xc3 20 bxc3 counterplay) 19 ... l:txfs 20 '1i'xg4+ 'ii'g s.
'ii'g s 21 ii.e6+ tllx e6 22 dxe6 �fs 23 b) 17 tllxg4 ii.xg4 18 ii.xg4 ii.xc3 19
'1i'd4 was agreed drawn in A.Huzman­ bxc3 '1i'f6 gives Black some compensa­
B.Avrukh, Pula 2000. Instead of 23 '1i'd4, tion. After 20 ii.e6+ (or 20 .!:!.cl tllxg4 21
Huzman suggested 23 'ii' bl. This has '1i'xg4+ '1i'g6) 20 ... tllxe6 21 dxe6 '1i'xe6
since been seen in practice and al­ 22 cs tllfs 23 '1i'g4+ 'lth8 24 cxd6 cxd6
though White eventually won after 2S l:i.fel the game was agreed drawn in
23 ... �es 24 �ael 'ii'fs 2s e7 �xe7 26 f3 R.Markus-N.Djukic, Herceg Novi 200S;
l:tae8 27 fxe4 '1i'g6?! 28 'ii' ds+ l:te6 29 2s ... l:tae8 would be unbalanced, but
'1i'xb7 in H.Grooten-N.Vanderhallen, level.
Vlissingen 2006, Black could have c) 17 '1i'd2 'ii' gs (or 17 ...ii.d7 !? 18 b4
maintained the balance with the move tlla4 forcing White to make a reason­
27 ...'ii'g s ! . able exchange sacrifice) 18 l:i.adl a6 19
16 o-o tllc s b4 tlld7 20 cs tlle s (not 20...dxcs 21 bxcs
Also interesting is 16 ...ii.xc3 17 bxc3 tllxcs 22 ..txa) 21 ii.xes ii.xes 22 tllxfs
'ii'g s with an unclear position in 'ii'xfs (and not 22 ...'1i'xd2 23 l:txd2 tllxfs
F.Atakisi-N.Gavrilakis, correspondence 24 tllxe4 when White's extra pawn
2006. means more than it does in the game)

G lig oric Va riati o n : 7 il e 3 tlJ g 4 8 ilg5 f6 9 il h 4 g5

23 '1i'xh6 1'..xc3 24 cxd6 �f6! 2S '1i'h4 24...tlJxe4 25 '1i'xe4 1'.. es

(White really had to try 2S '1i'e3 or 2S Black threatens ... 1'..xh2+. Instead
'1i'c1) 2s...cxd6 26 1'..xg4 'ii'e s. Now Black 2s ... 1'..xb2 26 1'.. e3 l:txfl+ 27 l:txfl �xfl+
held without much difficulty against his 28 'ltxfl is also possible and after
higher rated opponent: 27 f4 exf3 28 28 ... 'ii'hs Black should hold.
J:1xf3 �xf3 29 ilxf3 �f8 30 'lthl '1i'f4 31 26 1'.. e3 l:i.xfl+?
'1i'xf4 l1xf4 32 a3 1/2-1/2, A.Huzman- This is a mistake. Instead 26 ...'ii'h S!
1.Sidorenko, Ramat Aviv 2004. holds after 27 l:txfs (no better are 27 g 3
11 l:txfl+ 2 8 l:i.xfl �xfl+ 29 'ltxfl '1i'xh2
Black secures the cs-knight, but also and 27 h3 �xfl+ 28 �xfl 1'..h2+ 29
possible is Mikhalevski's 17 ...'1i'd7!? to 'ltxh2 l:txf1) 27 ... 1'.. x h2+.
cover fS and g4. After 18 b4 (18 �el 27 l:txf1 �xfl+ 28 'ltxf1 'ii' h s 29 h3
gives Black time to play 18 ... as and 18 1'.. xb2 30 '1i'e6+ 'ltg7 31 '1i'd7+ '1i'f7+ 32
ilf4 CZJd3! 19 CZJxfs! 'ii'xfs 20 1'..xh6 'ii'xf7+ 'ltxf7 3 3 a 4
1'.. xh6 21 1'.. xg4 '1i'g6 22 l:tc2 e3! gives This was M.Roiz-T.Radjabov, Saint Vin­
him counterplay according to Mik­ cent 200S. White has a clear advantage
halevski) 18 ... CZJd3 19 .i.xd3 exd3 20 in the bishop ending because Black's
CZJxfs 'ii'xfs looks okay for Black. pawns are fixed on dark squares and
18 tlJxg4 tlJxg4 19 1'.. xg4 'ii'g s 20 1'.. xfs White can create an outside passed
l:i.xfs 21 '1i'c2 e3 22 �eel exf2+ 23 .i.xf2 pawn on the kingside .
D2) 11 ...tiJd7

24 tlJe4!?
White return s the pawn to neutral­ Black makes this move in prepara­
ize Black's activity. 24 .lii.xcs is also pos­ tion for .. .fs, so that the knight can
sible, but after 24 ... dxcs 2s �xfs �xfs it quickly come to f6 to fight for the e4-
will not be easy for White to make pro­ square.
gress. 12 tiJd2 fs 13 exfs

A ttacking C h es s : The Kin g 's I ndian, Vo l u m e 1

The alternative is 13 f3 tllf6 (intend­ poses and 1 6 fxe4 can be met with
ing ... g4) 14 h3. 16 ... g4 17 hxg4 tllx g4! 18 ii.xg4 'ii'g s) 16
ii.f2 tllxe4! (this weakens White's con­
trol of dS) 17 tllxe4 (17 fxe4 tlld4)
17 ... cs 18 '1i'd2 (18 dxc6 bxc6 19 '1i'd2 dS
20 cxds cxds 21 tllcs tlld4 22 ii.xd4
exd4 is unclear according to Bologan)
18 ... h6 19 o-o-o (or 19 o-o tll d4 with a
good position) 19 ... a6 20 g4 tllh4!
( 21 h4!) 21 ii.xh4 gxh4 22 ii.d3
(better is 22 gs with complications)
22 ...�f4! 23 'ltb1 l:i.b8 (or 23 ... bs!? im­
mediately) 24 '1i'e2 '1i'e7 2S a3 ii.d7 26
Here Black has: 'lta2 bS and Black had the initiative in
a) 14...tllh S lS ii.f2 tllf4 16 g3 tllxe2 A.Huzman-V.Bologan, Moscow 2006.
17 '1i'xe2 cs (17 .. .fxe4!? looks better, as Moreover, Black can even improve on
after 18 tlldxe4 Black can activate his this with Bologan's 19 ... bS!? 20 cxbs a6
knight with 18...tllfs) 18 g4! f4 was 21 bxa6 (21 b6 �b8 22 'ltbl '1i'xb6 23 h4
B.Gelfand-1.Nataf, Cap d'Agde 2002. gxh4) 21 .....txa6 22 ii.xa6 �xa6 23 'ltbl
Here 19 h4 gives White an obvious ad­ '1i'a8 24 tllc3 �b8 with a strong initia­
vantage. tive.
b) 14 ... cs lS ..tf2 (White's play has 13 tllf6

been too slow to consider lS dxc6?!

bxc6 with ...ds coming) 1S .. .f4 16 �bl
(White tried the slower 16 a3 in
E.Bacrot-T.Radjabov, Morelia/Linares
2006) 16 ...b6 17 b4 cxb4 18 �xb4 tlld7
19 a4 tllcs 20 as ii.d7 was solid for
Black in B.Gelfand-T.Radjabov, Turin
Olympiad 2006.
c) With all of White's pawn moves,
it is logical to open the position with
14 .. .fxe4 and Black has scored well with
this move: lS tlldxe4 (after lS fxe4 This is the point of Black's play.
Black can play 1S ... g4 or 1 S ... tlld7, in- 14 tllde4
tending ...tll cs and perhaps ...l:tf4!?) Instead 14 o-o ii.xfs is not danger­
1s ...tllfs (Black can also play 1s ...tllxe4 ous, but White can play 14 ii.d3 tllxfs
immediately when 16 tllxe4 tllfs trans- lS tllde4 and here:

Gligoric Va ria tio n : 7 il e 3 tlJ g 4 8 il g 5 f6 9 il h 4 g 5

a) 1S ...1'.. h6 1 6 o-o 'lth8 17 cs g4 18 White can also prepare to castle

tiJxf6 '1i'xf6 19 tZJbs '1i'e7 20 '1i'e2 and long, but Black has resources: 17 '1i'e2
then: 'lth8 18 o-o-o '1i'e8 19 'ltbl '1i'g6 20 h3
al) 20... tZJxg3 21 fxg3 �xfl+ 22 .!:i.xfl gxh3 21 gxh3 tlJg8! (a typical regroup­
dxcs 23 '1i'e4! threatens 24 d6! and ing) 22 1'.. h4 1'.. xe4! 23 1'.. xe4 �6 24
gives White a winning attack ilg3 CZJf6 2S 1'..f 3 e4 26 1'.. g4 tlJxg4 27
a2) 20 ... 1'.. g7 21 cxd6 cxd6 22 '1i'xg4 hxg4 '1i'g6 28 'ital as and Black had the
tlJxg3 23 '1i'xg3 1'.. d7 was V.Kramnik­ initiative in J.Dorfman-1.Nataf, Mon­
T.Radjabov, Wijk aan Zee 2007. After 24 dariz 2000.
'1i'e3 ! Black would not have enough for
the pawn.
a3) 20 ... a6 21 1'..xfs �xfs (not
21 ... .i.xfs 22 cxd6 cxd6 23 tlJxd6) 22
tiJc3 �f8 23 tlJe4 'ltg8! 24 '1i'c4 1'..f s is
okay for Black according to Bologan.
b) 1S ...tZJxe4 16 tZJxe4 tiJh4! 17 o-o
(or 17 ilxh4 gxh4 with the idea of
...1'..fs) 17 ... g4 18 0.d2 (if 18 cs then
18 ... tlJg6 intending ... tiJf4 or ...h s)
18 ...'ii' g s 19 tlJe4 '1i'e7 20 tiJd2 h S ! ? 21 f3
CZJfs 22 1'..f2 1'..h6 23 '1i'e2 '1i'g7 24 1'..xfs White has control of the e4-square,
1'..xfs 2S 1'..e3 l:tf7 (Bologan suggests but Black has gained some space on
2s ... 1'..f4!?) 26 �ael l:i.af8 27 1'.. xh6 the kingside. White will be able to initi­
'1i'xh6 28 '1i'e3 '1i'xe3+ 29 �xe3 gxf3 30 ate queenside play with c4-cS, but
l:tfxf3 was Wang Yue-V.Bologan, Ji­ Black's main concern is the activation
angsu Wuxi 2008. Here 30 ... 'ltg7 would of his h6-knight.
leave Black with a slight advantage in 17 'lthB

the endgame. The king is a little safer here and

14 tZJxe4 15 tZJxe4 1'.. xfs 16 1'.. d3
••• Black can consider regrouping with
A good example of Black's possibili­ ... tlJg8 at some point. Nataf has used
ties is 16 f3?! g4 17 1'..f 2? (better is 17 this move frequently, but Black has
o-o) 17 ...gxf3 18 gxf3 tlJg4 19 1'.. d3 CZJxf2 some choice here and it is not clear
20 CZJxf2 'iih4 21 o-o e4 22 tZJxe4 1'..h 3 which move is best. There are two
23 �el 1'..e s 24 '1i'e2 'lth8 2 s 'lth1 �xf3 other logical tries:
26 '1i'xf3 ilg4 27 '1i'f2 ilf3+ 28 'ltgl a) 17 ... 1'..xe4 really tries to force
1'..x h2+ 0-1, J.Donner-S.Gligoric, Eersel things when 18 1'..xe4 'ii'g s protects the
1968. g4-pawn and intends ... tZJfs and ...hs.
16 ...g4 17 0-0 White has:

A t tacking C h e s s : Th e Kin g 's Indian, Vo l u m e 1

al) 19 'itb3 !Iab8 20 cs lLifs 21 'it'a3 White has:

was suggested by Mikhalevski, but bl) 18 !Iel 'if g6 19 'if d2 (19 cs!
19 ...b6 looks like a better try. transposes to variation 'b2') 19...!If7 20
a2) 19 'it'd3 lLifs 20 !Iae1 !If7 21 cs b4 !Iaf8 21 ..ih4 ..ixe4! 22 ..ixe4 lLifs 23
!Iaf8 22 cxd6 cxd6 23 !Xcel a6 24 b3 hS cs ..ih6 24 'it'd3 ..igs 2s ..ixg s 'it'xg s 26
gives Black counterplay. After 2S f4 ..ixfs !Ixfs gives Black good play in a
(forced) 2s ... lLixg3 26 fxgs !Ixf1+ 27 position with all the major pieces. The
!Ixf1 !Ixfl+ 28 'it'xf1 lLixfl 29 @xfl Black following game serves as a good illus­
had to work a little bit to draw in tration of Black's chances in this type of
R.Markus-V.Sikula, Hungarian League position: 27 !Ie2 h S 28 c6 b6 29 'it'a6?
2004. Instead Black could have tried h4 30 'it'xa7 !Ixf2 ! 31 !Ixf2 'it'e3 and 0-1
2s ...'i'h6 26 ..ixfs !Ixfs 27 fxes !Ixf1+ 28 in P.Smirnov-D. Kryakvin, Belorechen sk
!Ixfl !Ixfl+ 29 @xf1 h4! with equality. 2009.
a3) 19 'it'c1 'it'hs (19 ...'it'xcl followed b2) 18 cs looks more to the point:
by 20 ...lLifs is also possible) 20 cs !If7 21 18 ... 'it'g6 19 !Iel lLif7 20 ..ih4 (White
cxd6 cxd6 22 'it'c2 lLifs (after 22 ... ..if6 23 was also better after 20 !Icl lLig s 21
!Iad1 !Iaf8 24 'it'a4 lLifs 2s 'it'xa7 lLid4 26 lLixgs ..ixd3 22 lLie6 in Yang Shen­
'itb6?? ..id8 White was facing ...lLie2xg3 Xiaowen Zhang, Jiangsu Wuxi 2008)
and resigned in J.Vanheste-J.Piket, Am­ 20 ...!Iae8 (20...'it'h6 21 g3 is not such a
sterdam 1988, but of course he can im­ help to Black) 21 !Ic1 dxcs 22 !Ixcs lLid6
prove with 26 'it'a3) 23 ..ixfs 'it'xfs 24 23 'ifa4 ..ixe4 24 ..ixe4 'it'h6 (perhaps
'it'xfs !Ixfs 2s !Iac1 e4! gave Black better is 24 ... lLixe4 2S 'it'xe4 - but not
enough counterplay in D.Kuljasevic­ 2S !Ixe4? !If4! - 2S ...'it'xe4 26 !Ixe4 !If4
D.Popovic, Zupanja 2008. 27 !Iec4 !Id4 28 @f1 bS 29 !Ic3 a6 as
b) With 17 ... 'it'e8 Black brin gs his suggested by Mikhalevski, but Black is
queen to g6 before addressing his side­ still worse) 2S ..ig3 'it'd2 was L.Aronian­
lined knight. T.Radjabov, Morelia/Lin ares 2006. Now
26 !Id1 'it'e2 27 ..id3 'it'xb2 28 !Ixe7
would leave White much better.
18 .l:i:el
Instead 18 'it'e2 ..id7 19 f3 'it'e7 20
fxg4 and Vi-Vi B.Gelfand-T.Radjabov,
Russian Team Championship 2006, was
not very revealing, while 18 f3 'ife7 19
'it'c2 ..ig6 20 fxg4 lLixg4 21 'it'e2 lLih6 22
..if2 lLifs 23 !Iael ..ih6 gave Black
enough play in A.Kovalyov-1.N ataf,
Montreal 2008.

2 00
G l ig o ric Varia t i o n : 7 j_ e 3 0i g 4 8 j,_95 f6 9 j_ h 4 gs

18 'ife7
.•• After 2S cxd6 cxd6 26 0ic3 j_xd3 27
After 18 ... j_d7 19 cs 'if e7 20 l:i:c1 'it'xd3 the g4-pawn is not under attack
0ifs 21 cxd6 cxd6 22 .l:f.c7 White won and Black can play 27 ... 0ifs with coun­
material in Y.Shulman-1.Nataf, Mont­ terplay.
real 2008. The alternative, 18 ... 0ig8, 2s l:i:a4 26 0ic3 j_xd3 27 'it'xd3 l:i:b4

was played by Radjabov, but he did not Not 27 ... l:i:d4?? 28 'it'xd4.
repeat it when given the chance. After 28 cxd6 cxd6 29 !Ie4 l:i:xe4 30 0ixe4
19 f3 (perhaps 19 J:!.c1 OT 19 CS!? later £Ufs
worried the Azeri) 19 ... gxf3 20 'it'xf3
0ih6 21 'it'e2 j_g6 22 j_f2 0ifs 23 �f1
the game was drawn in A.Onischuk­
T.Radjabov, Mallorca Olympiad 2004.
19 !k1 b6 20 b4 as!
Black creates some play for himself
on the queenside.
21 a3 axb4 22 axb4

At last Black activates the knight.

Now ... hs is coming and it becomes
clear that the g3-bishop is more of a
problem for White than Black's 'bad'
bishop on g7, which may become ac­
tive with ... j_h6. The game V.Akobian-
1.Nataf, Montreal 2008, concluded as
22 j_g6
••• 31 'it'd1 hs 32 h3 j_h6 33 !Ic6 h4 34
Instead 22 .. Jfa3 23 cs bxcs 24 bxcs hxg4?
j_g6 (24.. JHa8! ? is also possible) 2S White had to give up a piece with
j_b1 .l:f.a1 26 'it'e2 �fa8 27 lUC3 j_xb1 28 34 j_xh4 'it'xh4 3 S 'it'xg4.
�xb1 !Ixb1 29 !Ixb1 'if gs was Wang 34 0ixg3 3S fxg3 j_e3+ 36 @h2 'ifh7

Yue-T. Radjabov, Elista 2008. White still Black can also win with 36 ...hxg3+
seems a bit better here, but Black 37 @xg3 j_f4+ 38 @f3 'i'h4.
gradually e qualized and drew. 37 'it'c2
23 cs bxcs 24 bxcs !Ias Instead 37 �c4 loses to 37 ... hxg3+
Black eyes the dS-pawn. 38 @xg3 j_f4+ 39 @f3 'it'h2.
2S 'it'e2 37 hxg3+ 38 @xg3 j_g1! 0-1

Chapter 11
Petrosian Variation
7�d-S aS withOut 8

1 d4 lLif6 2 c4 g6 3 lLic3 ..ig7 4 e4 d6 5 sian Variation was once a big favourite

lLif3 o-o 6 ..ie2 es 7 ds of Kramnik and he used this system to
defeat Kasparov on more than one oc­
casion, so it certainly has to be taken
seriously. Personally I have been happy
to play the Petrosian with both colours.
1 as!

This prophylactic move is the main

line and it has been played more than
twice as often as all other moves com­
bined. A couple of Black's obvious
moves run into difficulties: for exam­
ple, 7 ... lLie8 8 h4! or 7... lLih s 8 g 3 ! fs 9
This variation was pioneered by the exfs ..ixfs (9 ... gxfs 10 lLixes!) 10 lLig s
ninth World Champion. Although it lLif6 11 g4! and White has an unshake­
appears to be a simple line, the Petro­ able grip on the e4-square.
sian Variation leads to a real heavy­ The old main line was 7...lLibd7 8
weight strategic struggle. Both sides ..ig s (8 ..ie3 may be even better) 8 ... h6
often have the chance to play on either g ..ih4 gs 1 0 ..ig3 lLih s 11 h4!. In this
side of the board. Although it is not line Black will have to depend com­
considered critical nowadays, it is still a pletely on piece play, because a subse­
popular choice. The positions are gen­ quent .. .fs will hand the e4-square to
erally not as sharp as they are in the White. A respectable line is 7 ... lLia6
Mar Del Plata Variation, but the theory (which was Gallagher's recommenda­
does run deep in some lines. The Petro- tion in his 2004 book) 8 lLid2! and now

2 02
Petrosian Varia tio n : 7 dS a s with o u t 8 j,_9 5

8 ... c6 9 a 3 will sideline the a6-knight, Lines A and B are seen occasionally
while Gallagher's 8 ... cs looks too stodgy but are not very dangerous, while Line
to me after 9 h4 h S 1o lLif3!. C is a kind of transpositional device.
With Black intends to play Line D was fashionable for a while, but
... lLia6, from where it can hop to cs at 8 j_g5 is the absolute main line and
an appropriate moment to pressure will be considered in Chapter 12.
the e4-pawn. From a6 the knight will There are a couple of other options:
also help to hold up both White's b2-b4 a) 8 h4 lLia6 9 lLid2 is Line A.
and c4-c5 advances without interfering b) 8 a3 j_d7! (8 ... lLia6 allows 9 b4
with Black's development (compared to because of the pin on the a-file; this is a
7...lLibd7). After White makes the nec­ common theme in the Petrosian) 9 j_e3
essary preparations for the b4-advance (or 9 !Ib1 a4! intending ...'it'e8 and
(commonly !Ib1, b3 and a3), Black will ... lLia6) 9 ... lLig4 10 j_g5 f6 11 j_h4 lLia6
often play ... lLics, in order to meet b4 12 o-o 'it'e8 13 lLid2 lLih6 14 f3 lLif7 1 5
with ... lLia4!. Of course the knight will @h1 f S gave Black equal chances in
need the proper support to make this l.lvanov-J.Fedorowicz, USA 1989.
leap, but we will see that the moves c) 8 g3 was another specialty of the
...'it'e8 and ...j_d7 are often an integral late GM Igor Ivanov that I have even
part of Black's plans. tried myself. My friend GM Jesse Kraai
After 7 ... as, White has a diverse called this 'a move from outer space'
choice. when he saw it and thought chess
must be a great game if such moves
were playable. 8 g3 is not very danger­
ous theoretically, but Black should be
aware of White's ideas after 8 ...lLia6 9
lLih4!. This looks strange, but it is part
of White's plan, as we shall see.

A: 8 lbd2
B: 8 .i.e3
C: 8 0-0
D: 8 h3

A ttacking C h e s s : Th e King 's In dian, Vo l u m e 1

Now we have: With the centre closed White hopes

c1) 9 ... lLie8 10 o-o fs 11 exfs gxfs 12 to stir up trouble on the kingside.
f4 e4 13 ..ie3 intending ..id4 and lLig2- 8 lLia6 9 h4

e3 was D.Vig orito-A.Matikozian, Las Instead 9 o-o transposes to Line D,

Vegas 200S, when we can see White's while 9 g4 lLics 10 h4 is considered in
idea take shape, although even here the notes to White's 10th move, below.
Black is not without resources. 9...lLic s
c2) 9 ... lLics 10 f3 lLih s 11 o-o (it is T h e most dynamic. Black h a s no
probably better to play 11 lLig2 ..ih3 12 reason to fear the advance of the White
0-o fs 13 exfs gxfs with an unclear po­ h-pawn. 9 ... hs is also possible, but it is
sition) 11 ...lLif4! (taking advantage of a bit stodgy:
the knight's position on h4) 12 lLig2 a) 10 lLif1 lLics 11 ..ig s 'it'e8 12 lLid2
lLixe2+ (after 12 ... lLih3+ 13 @h1 fS 14 ..id7 13 f3 lLih7 14 ..ie3 fS gave Black
exfs gxfs 1s f4 the h3-knight's position counterplay in Y.Teplitsky-A.lstratescu,
looks a bit precarious) 13 'it'xe2 ..ih3 (or Yerevan Olympiad 1996; White does
simply 13 .. .fs) 14 ..ie3 b6 1s @h1 fs 16 not have an ideal home for his king.
exfs gxfs 17 !Ig1 'it'd7 18 lLid1 !Iae8 19 b) 10 f3 c6 (10 .....ih6 is given as
lLif2 ..ixg2+ 20 !Ixg2 @h8 21 !Iag1 'it'f7 equal by Nunn and Burgess) 11 lLif1
22 b3 lLid7 23 g4 and now, instead of lLics 12 ..ie3 and now rather than
23 ... !Ig8?! 24 gxfs 'it'xfs 2 s lLie4 when 12 ... cxds 13 cxds, which hands the c4-
the g-file and e4-square were under square to White, Black should play
White's control in l.lvanov-B.Gelfand, 12 ... ..id7, intending ... a4, ...'it'as, ... cxds
New York 1989, Black could play 23 ...f4! and ... bs.
24 ..id2 e4! 2s lLixe4 lLics with good c) 10 lLif3 ! ? heads for the weakened
counterplay. gs-square: 10...lLics 11 lLig s c6 12 ..ie3
..id7 and a complicated strategic
A) 8 lLid2 struggle lies ahead.
10 h s
Instead 10 g4 is somewhat unnec­
essary, but it has been seen often
enough in practice. Following 10 ... a4
(10... c6 is also possible) 11 hS gxh S ! ? 12
gs (after 12 gxhs ..ih6! intending
... @h8 the opening of the kingside is
welcomed by Black) 12 ...lLig4 13 lLif1 fS
we have:
a) 14 exfs ..ixfs 1s f3 a3! gives Black
good play. One possibility is 16 fxg4

2 04
Petro sian Va riatio n : 7 dS as with o u t 8 j,_95

..i.xg4 17 j_xg4 hxg4 18 'it'xg4 axb2 19 queen on the c-file) 11 ... a4 12 lLif1 cxds
..i.xb2 lLid3+ 20 @e2 lLixb2 21 'it'hs l:i:f4 13 cxds j_d7 14 hxg6 fxg6 (14...hxg6 is
22 'it'xh7+ @f8 with a messy position also possible - Black is not about to get
that looks very playable for Black. mated on the h-file) 1s f3 lLihs 16 g3
b) 14 f3 lLif2 (Black could also play bS! gave Black good play in
14.. .fxe4 1 S fxg4 j_xg4 16 j_xg4 lLid3+ A.Dunnington-Z.Lanka, Paris 1990. If
17 @d2 'it'xg s+ 18 @c2 lLib4+ 19 @b1 White takes on bS then f3 will be hang­
"ihg4 with an unclear position) 1 S ing.
ltixf2 fxe4 1 6 @g2 a3 ! 17 !Ixhs exf3+ 18 11 cxds

..i.xf3 e4 19 j_e2 j_fS gave Black excel­ Often is it better to delay this ex­
lent compensation for the piece in the change because White gains access to
famous game B.Kouatly-G.Kasparov, the c4- and bS-squares, but here there
Evry (simul) 1989. is a neat tactical point. Black has also
avoided the exchange in practice with:
a) 11 ... 'it'b6 is a common idea. The
queen eyes the b2- and f2-squares,
thereby creating some tactical possi­
bilities: 12 f3 (not 12 gs? lLixhs 13
j_xhs lLid3+ 14 @e2 lLif4+) 12 ...j_d7 13
lLif1 cxds 14 cxds !Ifc8 1s lLie3 a4 16
lLic4 'ife7 17 lLia3 ! was A.Aleksandrov­
R.Leitao, Yerevan Olympiad 1996.
White's knight manoeuvring has been
very time-consuming, but he has man­
10 c6
... aged to grip the bS-square. Instead of
Black looks for queenside counter­ the game's speculative 17 ... bs, I would
play. This move was relatively untested suggest 17 ... 'it'b6 when Black looks
when Nunn and Burgess wrote in the okay.
1990s, but now there are many exam­ b) 11 ...a4 is logical and then:
ples and Black has scored well. bl) 12 g s should backfire: 12 ...lLixhs
11 g4 13 j_xhs lLid3+ 14 @f1 'it'b6 1 s !Ih2 and
Others: now, instead of 1s ...gxhs?! 16 'it'xhs
a) 11 h6 limits White's possibilities. h6? 17 gxh6 j_f6 18 lLif3 with a win­
After 11...j_h8 12 f3 lLihs 13 lLif1 lLif4 ning attack in A.Aleksandrov­
14 g3 lLixe2 1 S 'it'xe2 fS Black had the V.Kotronias, Pula 1997. Black should
initiative in A.Bykhovsky-1.Caspi, play Huzman's 1s ... a3 ! 16 b3 'it'd4! 17
Herzliya 2006. lLidb1 lLixc1 18 'it'xc1 gxhs when he has
b) 11 'it'c2 (it looks odd to place the a huge advantage.

A ttacking C h e s s : Th e King 's In dian, Vo l u m e 1

b2) 12 f3 ..id7 13 lLidb1! 'it'b6 14 'it'b 3 ! being good for White. 1 2 ... a4 and
'it'd2!? (White makes some unusual 12 ... ..id7 are possible, but if Black
moves, but they are all designed to wants to play more slowly it is better to
shut down Black's counterpl ay) delay the exchange on dS.
14 ... !Ifc8 1s lLia3 'it'as 16 @f1 l:!.ab8 17 13 gs
'it'e1 'it'd8 18 @g2 cxds 19 cxds lLie8 20 Altematively:
lLiabS and White was better in a) 13 f3 b4 14 lLia4 lLifd7 is comfort­
C.Gabriel-E.Grivas, Corfu 1999. able for Black. Even ... ..if6-gS is a possi­
12 cxds bility.
Instead 12 exds e4! gives Black ob­ b) 13 hxg6 fxg6 14 f3 b4 1S lLia4
vious counterplay: lLifd7 is similarly pleasant for Black.
a) 13 g s lLixh s 14 ..ixh s gxh s 1s c) 13 a3 and here:
lLicxe4 l:i:e8 16 'it'xh s .ifs was clearly cl) 13 ...'it'b6 14 f3 ..id7 1S lLib3 b4
better for Black in D.Shchukin­ 16 lLixcs 'it'xcs (instead 16...bxc3 17
E.lnarkiev, St Petersburg 2001. lLixd7 lLixd7 18 bxc3 fS looked too
b) 13 h6 ..ih8 14 gs lLih s 1s lLidxe4 speculative in l.Tsesarsky-M.Tseitlin,
lLixe4 16 lLixe4 l:i:e8 17 l:i:h4 .ifs 18 ..if3 Givataim 2000, although Black won) 17
..ixe4 19 !Ixe4 l:i:xe4+ 20 ..ixe4 'if e7 21 lLia4 'ifa7 looks fairly level.
f3 !Ie8 22 'it'd3 lLig3 won back the pawn c2) 13 .....ixg4! now works.
and left Black with some initiative in
A.Voll-D.Petrovic, correspondence 2004.

After 14 ..ixg4 b4! both 1S lLie2?

lLixg4 and 1S lLia4? lLixa4 hang pieces,
12 bs!
... which leaves:
Black takes advantage of the loose C21) 1S l2Jcb1 l2Jd3+ 16 @fl jVb6 17
pawns on e4 and g4. Instead 12 .....ixg4 !Ih2 lLixg4 18 jVxg4 lLixc1 wins.
does not work, but the idea is worth c22) 1S axb4 axb4 16 !Ixa8 'it'xa8 17
keeping in mind, despite here 13 ..ixg4 lLibs lLid3+ 18 @f1 lLixg4 19 jVxg4 'it'b8
lLid3+ 14 @fl 'it'b6 1S l:i:h2 lLixb2 16 and both the knight on bS and bishop

Petrosian Va ria tion : 7 ds a s with o u t 8 j,_95

on cl are hanging. 17 ... 'it'xg 5 18 @e1 'it'g2 19 .l:f.f1 j_h3 20

c23) 15 lLia2 lLid3+ 16 ..tf1 bxa3 17 j_f3 'it'g 5 Black has a strong initiative
bxa3 jVb6 18 !Ih2 lLixg4 19 'it'xg4 'it'd4 for the sacrificed material.
20 lLib3 'ifc4 21 lLid2 'it'c2 22 'it'e2 lLif4
intending ....l:f.ab8 looks like fun for
There is no reason to play the pas­
sive 13 ...lLie8. Black must exploit his
tactical possibilities.
14 j_xhs

As in the Gligoric Variation, White

allows this bishop to get chased around
a bit.
8 ...lLig4
Worse is 8 ... lLia6 9 lLid2 when White
has the ideal piece set-up and total
14... b4! 1s lLia4 flexibility with his king position.
Avoiding the challenge is worse. Af­ 9 j_g5 f6 10 j_h4
ter 15 lLib3 bxc3 16 lLixc5 cxb2 17 j_xb2 Instead 10 j_d2 is not very danger­
dxc5 18 j_e2 'it'xg5 White's position ous after 10...f5 (however, 10...lLia6 11
was a mess in B.Kohlweyer-D.Stets, h3 lLih6 12 h4 is a bit annoying) and
German League 2005. now:
1s...lLid3+ a) 11 o-o lLia6 (or 11 ... lLif6) 12 lLie1
15 ...lLixa4 16 'it'xa4 gxh 5 is probably lLif6 13 exf5 j_xf5!? was V.Gallego
even better. After 17 !Ixh5 j_g4 18 !Ih4 Jimenez-J.Gallagher, Benidorm 1991.
j_d7 Black will play ...'it'xg5 with a clear Taking with the bishop is often a de­
advantage. cent choice when Black has good con­
16 @fl lLif4! trol of the e4-square, but 13 ...gxf5 is
Now 17 j_f3? j_a6+ 18 @g1 'it'xg5+ also fine.
19 @h2 was R.Kozlov-P.Gnusarev, As­ b) 11 exf5 gxf5 (again 11...j_xf5 is
tana 2007, when 19 ...j_e2 wins in­ fine, as in A.Huss-J.Gallagher, Samnaun
stantly. Better is 17 j_e2, although after 2008) 12 lLig5 lLif6 is variation 'c' below.

A ttacking C h e s s : Th e King 's I n di a n , Vo l u m e 1

Instead 12 ... 'it'e8 13 cs!? dxcs 14 'ifb3 a) 1 3 ... ..id7 1 4 o-o fS (Black should
@h8 1S 0-0 gave White compensation consider 14 ... g4 or 14... 'it'e8) 1S exfs
in B.Zlotnik-A.Morozevich, Moscow lLixfs 16 lLide4 looks somewhat better
1991, although he soon went wrong for White.
and lost. b) 13 ... 'it'e8 14 f3 (14 b4 fS 1S f3
c) 11 lLigs lLif6 12 exfs gxfs (this axb4 16 axb4 ..id7 17 l:i:b1 g4 gives
time 12 ... ..ixfs 13 g4! ..id7 14 lLige4 is Black counterplay according to Avrukh)
to White's advantage) 13 f4 e4 is fairly 14... ..id7 1S ..if2 and now, instead of
level. 1S ... a4? 16 lLibS! which was very good
for White in E.Bacrot-E.Relange, French
Team Championship 200S, Avrukh
suggests 1s .. .fs 16 exfs lLixfs 17 o-o
lLid4 with sufficient play.
11 lLid2
No one has tried 11 h3 lLih6. Pre­
sumably Black gets a favourable ver­
sion of the note to his next move. If
Black really does though prefer to play
with a pawn on hS, he could avoid this
possibility altogether by playing 10 ... hS.
After 10 ..ih4, White wants to play
lLid2, f3 and ..if2 with a flexible and
harmonious position. In most cases
Bl ack should maintain his knight on g4
until forced to retreat. White generally
cannot play f3 without allowing ... lLig4-
e3, so Black usually waits for White to
play h3, which will cost a tempo and
somewhat weaken the white kingside.
10 lLia6

10 ... h s 11 lLid2 lLia6 transposes.

Another idea is 10 ...lLih6. The point 11 hs

is to play and .. .fs before White Instead 11...lLih6!? allows White to
gets in f3 and ..if2. After 11 lLid2 gs 12 reach his desired set-up with 12 f3, but
..ig3 lLia6 (12 .. .fs !?) 13 a3 ! ? (or 13 f3 fS Black still looks fine following 12 ... lLif7
when Black gets counterplay following (or 12 ... ..id7 13 o-o 'it'e8) 13 a3 ..id7 14
14 ..if2 g4 or 14 exfs lLixfs 1s ..if2 lLid4) l:i:b1 lLics 15 b4 axb4 16 axb4 lLia4.
we have: Black is also not bothered by 12 a3 ..id7

Petrosian Varia tion: 7 ds as witho u t 8 ii. g s

1 3 !Ib1 lLics (13 ... 'it'e8 14 b4 sidelines in the air. Better is 18 lLixa4 ii..xa4 with
the a6-knight) 14 b4 axb4 1S axb4 l2Ja4 equality in J.Speelman-A.Strikovic,
with the idea of ... gs and ... fs. Oviedo (rapid) 1992.
12 a3 a2) 13 ...'it'e7 14 .l:f.b1 lLics 1S b4 axb4
The immediate 12 h3 lLih6 has 16 axb4 l2Ja4 17 lLixa4 !Ixa4 18 h3 lLih6
scored well for Black: 13 g4 (instead 13 19 f3 was J.Speelman-J.Piket, Tilburg
a3 ii..d7 transposes to the main line, 1992, when Black should just play the
while after 13 f3 lLif7 Black will play obvious 19 ....rf.fa8 with equality, as in­
... ii..h 6) 13 ... hxg4 14 hxg4 lLif7 with the dicated by Speelman.
idea ...ii..h 6, ... @g7 and ...!Ih8, which b) 13 .l:f.b1 lLics and here:
gives Black good play. bl) 14 b3 g S ! 1S ii..g 3 fS 16 h3 lLif6
12 ...ii..d 7 gives Black the initiative after either 17
ii..xh s lLid3+ 18 @f1 lLif4 (or 18...lLixhs
19 'it'xh s lLif4), or 17 f3 'it'e8 18 'it'c2
fxe4 19 fxe4 'it'g6, V.Loginov-A.Fedorov,
St Petersburg 1997.
b2) 14 b4 axb4 1S axb4 lLia4 16
lLixa4 (after 16 'it'c2 Black again seizes
th e initiative with 16... gs 17 ii.. g3 fS)
16...ii.. xa4 17 'it'c1 ii..d7 and Black had a
good position in G.Danner-V.Spasov,
Leon 2001.
13 ...lLih6 14 .l:f.b1 lLics 15 b4 axb4 16
13 h3 axb4 lLia4
White finally tires of having Black's
knight hanging around. This move i s
generally criticized, but the alterna­
tives are also unimpressive:
a) 13 0-0 and now 13 ...lLih6 is com­
pliant, wh ile 13 ...lLics allows 14 b4, so
we have:
al) 13 ...'it'e8 14 b3 fs 1S l:i:b1 lLics 16
b4 axb4 17 axb4 lLia4 and here White
should avoid 18 'it'c2 lLixc3 19 'it'xc3
fxe4 20 lLixe4, since Black has
20 ... lLixh2! when 21 @xh2 fails to 17 'it'c2
21 ...J:i.f4 and 21 lLixd6 cxd6 22 @xh2 Black also had fine play after 17 lLixa4
.l:f.f4 gives Black the initiative, as ...e4 is ii..xa4 18 'it'c1 gs 19 ii..g3 h4 20 ii..h 2 ii..d7

A ttacking C he s s : Th e Kin g 's In dian, Vo l u m e 1

21 f3 f5 22 ..ig1 g4 in l.Morovic Feman­ a) 9 lLie1 lLic5 and here:

dez-G.Kamsky, Las Palmas 1994. al) 10 'it'c2 transposes to note 'a' to
17 ...lLixc3 18 'it'xc3 gs 19 ..ig3 h4! 20 White's 10th move in our main line,
..ih2 fs 21 cs g4 22 c6 ..ic8 although this way White has managed
Black keeps the tension. Kasparov to avoid 9 'if c2 lLih5.
gives 22 ... bxc6 23 dxc6 ..ie6 24 ..ic4 as a2) 10 f3 has scored terribly for
unclear. White after 10 ...lLih5 ! . Indeed, Black has
23 hxg4 fxg4 a pleasant choice after both 11 ..ie3
23 ...lLixg4!? looks good too. (between 11 ...b6, 11 ...lLif4 and 11 ... f5),
24 cxb7 ..ixb7 2s o-o 'it'gs and 11 lLid3 (between 11 ...b6 and
Black had the initiative in E.Bareev­ 11 ...lLixd3 12 ..ixd3 lLif4), so White usu­
G.Kasparov, Tilburg 1991. ally prevents the knight hop with 11 g3.
Then 11.. ...ih3 12 !If2 f5 (or 12. . ...id7)
C) 8 0-0 13 lLid3 lLixd3 14 'it'xd3 (14 ..ixd3 f4 15
g4 lLif6 is similar) 14.. .f4 15 g4 lLif6
(15 ... lLig3 is possible but more risky af­
ter the cool 16 ..id1) 16 ..if1 ..ixf1 17
'it'xf1 lLid7, with the idea .....if6, gives
Black good play.
b) 9 lLid2 lLic5

This move is played rarely nowa­

days, but it is still important, especially
for transpositional reasons.
8...lLia6 9 'it'c2
We will take this as the main line,
but there are several other moves. 9
..ig 5 is actually the most common and here:
move, and after 9 ... h6 10 ..ih4 we are in bl) 10 'it'c2 transposes to note 'b' to
Chapter 12. Instead after 9 ..ie3 lLig4 10 White's 10th move in our main line,
..ig5 f6 White has a less flexible version again with White avoiding 9 'it'c2 lLih 5 .
of Line B. White's two remaining alter­ b2) 10 f3 lLih5 11 lLib3 b6 has done
natives to 9 'it'c2 are the typical knight well for Black.
retreats: b3) 10 !Ib1 will transpose to varia-

Petrosian Va ria t i o n : 7 ds as with o u t 8 j,_95

tion 'b4' after 10...lLie8 11 b3, while lLia6 (or 7... lLibd7) 8 dS lLics 9 'it'c2 as.
10 ... j_h 6 is also sensible. With the move order we are going by,
b4) 10 b3 lLie8 (10 ... j_h6 has also however, Black has another promising
scored well for Black after 11 'it'c2 lLie8 option in 9...lLih s.
or 11 ...j_d7) 11 a3 fS 12 l:i:b1 lLif6 and

White has not been able to show

anything here. On the contrary, Black
b41) 13 b4 is a strange pawn sacri­ has done very well. Some examples:
fice. After 13 ... axb4 14 axb4 lLicxe4 1S a) 10 g3 fs 11 exfs j_xfs 12 'it'd1 (or
�dxe4 lLixe4 16 lLixe4 fxe4 17 j_e3 12 lLie4 lLif6 13 lLifd2 lLib4 14 'it'b1 c6)
Black is doing well after all of 17 ....rf.a2, 12 ... lLif6 13 lLih4 j_h3 14 lLig2 'it'd7 gave
17 ...j_d7 and 17 ... 'it'M. The extra pawn Black some initiative in S.Hamann­
may not mean much itself, but Black R.Hiibner, Skopje Olympiad 1972.
certainly has no problems. b) 10 !Ie1 lLif4 11 j_f1 is a bit trick­
b42) 13 f3 j_d7 (preparing for ier. After 11 .. .fs 12 j_xf4 exf4 13 es
White's queenside advance) 14 b4 axb4 White has good chances, so Black
1S axb4 lLia4 (we will see this idea should play 11 ...j_g4!.
countless times) 16 'it'c2 lLixc3 (Black c) 10 lLie1 lLif4 11 j_f3 fS 12 j_e3 b6
could also play 16 ...j_h6 or 16 ... lLih s) 17 13 exfs j_xfs 14 j_e4 lLics 1s j_xfs !Ixfs
"if'xc3 and now Black has several prom­ 16 g3 lLih 3+ 17 @g2 'it'd7 18 lLid3?!
ising ideas, such as 17 ... lLihs , 17 ... j_h6 (better is 1 8 j_xcs bxcs 19 lLie4, al­
and 17 .. .fxe4 18 lLixe4 lLixe4 19 fxe4 though Black is fine after 19 ... l:i:af8 or
�xf1+ 20 j_xf1 'ifh4. 19 ... a4) 18 ... lLixd3 19 'it'xd3 lLif4+! 20
Now we return to 9 'ifc2: gxf4 exf4 gave Black a strong attack in
9 lLics
... J.Nikolac-G.Ligterink, Wijk aan Zee
After this move we reach a position 1976.
that has occurred hundreds of times, Even though 9 ...lLih S looks promis­
although it usually arises after 7 o-o ing, we will also examine 9 ...lLics in

A ttacking C he s s : The Kin g 's I n d i a n , Vo l u m e 1

detail because the position can arise in choice between 10...lLie8 11 lLib3 fS (or
various ways and doing so provides 11 ...b6) 12 lLixcs dxcs! 13 ..ie3 b6, in­
Black with a couple of good options. tending ...lLid6, and 10.....ih6 11 lLib3
..ixc1 12 l:i:axc1 lLifd7 (12 ... lLixb3 and
12 ... b6 are satisfactory as well) when
Black has no problems.
10. h6 11 ..ie3 b6

This so lid move has gradually

evolved to become the main line. Black
has also tried 11 ... lLifd7, 11 ...lLihs and
11 ... 'ife7, while after the forcing
11 ... lLig4 12 ..ixcs dxcs 13 h3 lLif6 14
lLixes lLixds 1s cxds ..ixes 16 f4 ..id4+
17 @h2 White has a big centre, so I pre­
10 ..tgs fer the text, which keeps more tension
After 10 b3 lLihs Black is even better in the position.
off than after 9 ...lLihs above. 10 ..ie3 12 lLid2
gives Black a choice between the solid
10...lLifd7 or 10...b6 and the more en­
terprising 10...lLig4 11 ..ixcs dxcs 12 h3
when he has the extra option of
12 ... lLih6! ? compared to 11 ... lLig4 in the
note to his next move.
Again there are also the knight re­
a) 10 lLie1 lLifd7 (this retreat keeps
control of both cs and es, but Black has
decent options too in 10...b6, 10.....id7
and 10 ... lLie8) 11 ..ie3 fS 12 f3 (after 12 12.....ig4!?
exfs gxfs 13 f4 exf4! 14 ..ixf4 lLies This unusual-looking move has be­
Black has good piece play) 12 .. .f4 13 come popular. It looks rather anti­
..if2 gs and Black is doing rather well. positional to offer a trade of light­
White still has to prepare b4 and Black squared bishops, but Black hopes that
should get in ... g4 rather easily - White must make some concession to
White's queen being on c2 (as opposed make this exchange. The more natural
to d1, from where it controls g4) helps 12 ... lLig4 is also viable. After 13 ..ix94
Black. ..ixg4 14 a3 (14 f3 ..id7 planning . ts ..

b) 10 lLid2 gives Black a pleasant gives Black counterplay, so White tries

Petros ian Va riation: 7 ds as with o u t 8 il.. g s

to ignore the bishop) 14 ..lLia6 15 !Iab1 the untried 13 ... i.d7 ! ? when it is not
fS ! (1s ... i.d7 16 b4 axb4 17 axb4 fS has clear what White's rook is doing on el)
also been played, but the text is 14 l:i:xe2 lLig4 15 i.xcs bxcs and we
sharper) 16 f3 (16 exfs gxfs 17 f4 e4 is have:
unclear) 16 ... i.h s 17 b4 axb4 18 axb4
f4 19 i.f2 g s Black has counterplay. The
break ... g4 is coming and if White plays
h3, Black will retreat the hS-bishop and
play ...h s and ... g4 with even greater
13 f3
This natural move gains a tempo,
but Black has a reason for provoking
this advance. The natural 13 i.xg4
lLixg4 14 i.xcs bxcs is fine for Black
despite the exchange of light-squared bl) 16 lLibs hs 17 a4 i.h6 18 'ifc3 fs
bishops. It is not easy to come up with 19 f3 lLif6 20 lLib3 ?! (winning the as­
an active plan for White and Black will pawn at great cost; instead 20 exfs
play .. .fs or ...h s , activating the bishop gxfs 21 lLib3 h4 22 lLixas lLih s with
on h6. Instead after 13 'it'd1 Black counterplay is given by Panczyk and
should just retreat with 13 ... i.d7, con­ llczuk) 20 .. .fxe4 21 fxe4 lLig4 with a
tent th at White has undeveloped his strong initiative for Black in A.Korobov­
queen. Others: N.Kuren kov, Moscow 2007. White's
a) 13 h3 i.d7 (happy to have in­ knights are too far away from the ac­
duced a kingside weakness, Black re­ tion.
treats, but 13 ... i.xe2 14 lLixe2 lLih s has b2) 16 g3 looks silly, but the follow­
al so been tried) 14 b3 lLih7 15 a3 fS 16 ing play shows the strategic complexity
exfs (or 16 b4 lLia6 and 16 f3 f4 looks inherent in the Petrosian Variation.
good too) 16 ... gxfs 17 f4 exf4 18 i.xf4 With his rook on e2, White wants to
'ifh4! 19 !Iae1 lLigs 20 @h 2 l:i:ae8 21 play on the kingside him self: 16 ...h s 17
'it'c1 22 i.xes l:i:xes 23 b4 axb4 24 f3 lLih6 (or 17 ...lLif6 18 f4) 18 f4 fS 19
axb4 f4! gave Black a winning attack in lLif3 and White had the initiative in
Y.Zilberman-1.Smirin, European Club A.Korobov-P.Czarnota, Cappelle la
Cup, Panormo 2001. White cannot Grande 2004.
avoid detonation on h3 - a far reaching 13 i.d1 14 b3 lLihs!

consequence of his 13th move! This is precisely why Black provoked

b) 13 !Iae1 is more challenging. f2-f3.
Then 13 ...i.xe2 (Black should consider 15 .l:f.fel

A ttacking C h e s s : Th e Kin g 's I n dian, Vo l u m e 1

For the pawn Black has accelerated

his kingside play.

1s ... ..tf6!
Both 1s ...t2lf4 16 ..if1 and 1s .. .fs 16
a3 look too slow, so Black offers a pawn 19 exfs gxfs 20 a3 a4 21 b4 lLib3
to activate his dormant bishop. Black had excellent play in
16 ..ixh6 l.Naumkin-1.Smirin, lschia 199S.
After 16 a3 ..igs White cannot play
17 ..if2? because of 17 .....ixd2 18 'it'xd2 D) 8 h3
l2lxb3. With his bishop now activated,
Black also has good play after 16 !Iab1
..tgs 17 ..tf2 fs.
16 ..tgs! 11 ..ixgs

White cannot take the rook. He gets

mated after 17 ..ixf8 ..ie3+ 18 @h1 (or
18 ..tf1 'ifh4 19 t2ld1 l2lg3+) 18 ... l2lg3+
19 hxg3 'it'xf8 20 l2lf1 'it'h6+ 21 l2lh2
11 'it'xgs 18 tlJfl

A quick finish was seen after 18 ..if1

fS 19 a3 !If7 20 !Iab1 t2lf4 21 b4? (bet­ This is a tricky move that has been
ter is 21 g3 !Ih7, although Black has favoured by Zvjaginsev. 8 h3 in the Pet­
reasonable compensation for the rosian Variation is very similar to the
pawn) 21 ...l2lcd3 and 0-1 in R.Wukits­ Makogonov Variation (S l2lf3 0-0 6 h3)
G.Timoscenko, Oberwart 2003. White is and there are several transpositional
losing the exchange because 22 !Ied1 possibilities: for example, 1 d4 l2lf6 2 c4
walks into 22 ...l2lh3+ 23 @h1 l2ldf2 g6 3 t2lc3 ..ig7 4 e4 d6 s l2lf3 0-0 6 h3 es
mate. 7 dS as 8 ..igS l2la6 9 lLid2 (or 9 ..ie2)
1s fs
... 9 ...'it'e8 10 ..ie2 transposes to the Petro-

Petros ian Varia ti o n : 7 dS as with o u t 8 j,_95

sian. Here we will look at lines where White will seize control of c4 and bS
White has played an early j_e2, while and it will be difficult for Black to cre­
in the Makogonov Variation (which will ate counterplay, so Black should gen er­
be covered in Volume 2), we will look at ally delay this exchange until he can
positions where White delays or omits successfully control the bS-square. lfhe
th is move. can do this, then the advance will
There are a couple of ideas behind not only control c4, but will give him
the subtle 8 h3. White is ready to play the possibility of playing ...b4.
the sequence 9 j_g5 h6 10 j_e3 because Black's other set-up is seen in the
now Black cannot play ...l2lg4. White main lines of both Lines D1 and D2,
then hopes that ... h6 will weaken the below. Black will play ... 'it'e8, ...t2ld7 and
black kingside - the h6-pawn itself may ... tlJdcs. This will usually allow him to
be weak, but more importantly the di­ get in ...fs because the diagonal for the
agonal b1-h7 will be weakened after a c8-bishop remains open. It is in these
subsequent .. .fs. This may seem minor, lines that we see the importance of
but in practice this detail can create big leaving out the weakening move ... h6.
problems for Black. Another idea for Because the gs-bishop is hanging in
White is to play g2-g4. Then if Black the air, Black may drive it away first
plays .. .fs, White can pl ay gxfs, opening with .. .f6 before playing .. .fs, or play .. .fs
the g-file, and exchanging twice on fS straightaway. Having both knights on
could grant White the e4-square for his the queenside may also give Black
knights. If Black does not play ... fs then some tactical possibilities. The moves
White may tum his attention to the ... j_d7 and ... a4 are also thematic (espe­
other wing. cially if Black does not play .. .fs), but in
The downside to 8 h3 is that it general I think th at Black should avoid
weakens White's kingside, so castling playing ... c6 in this set-up.
on that side of the board becomes less 8 l2la6

appealing for the first player. Also the Instead 8 ...tlJh S is not a bad idea,
g4-advance creates weaknesses and but White can play either 9 g3 (as he
Black's .. .fs break can be very danger­ would after 7 ...tlJh s), or the exotic 9
ous for White if he is unable to keep l2lg1 ! ? intending 10 j_f3.
the position under control. An interesting idea is 8 ...l2lfd7!?. In­
Generally Black has two different deed, Panczyk and llczuk give this as
ways to deploy his pieces. He can play their main line even though it has only
... tLJcs, ... a4, ... c6, ... j_d7 and ... 'it'as to been played a handful of times. Now 9
create play on the queenside. However, g4 tLJcs 10 j_e3 l2lba6 intending .. .fs
Black should be careful concerning the looks pretty comfortable for Black. In­
move ... cxds. If it is played too early, stead g h4! l2lf6 (9 ... fs 10 h S ! ) 10 t2ld2 ! ?

A ttacking C he s s : Th e King 's In dian, Vo l u m e 1

leaves White a tempo up on 8 lLid2 lLia6 16...lLicxe4 17 lLixe4 'it'xc2+ and

9 h4, although this is not exactly fatal 18 ... lLixe4 with an extra pawn)
in this kind of position. 1S ... lLifxe4! 16 lLidxe4 (or 16 lLicxe4
lLixe4 17 lLic4 'i't'b4 18 'it'xe4 !Ifc8 19
..id2 'it'cs 20 ..ie3 and here, instead of
repeating moves with 20 ... 'i't'b4, Black
can play 20 ... 'it'bs! with the idea of
... .ifs) 16 ...lLixe4 17 'it'xe4 (17 lLixe4 loses
to 17 ...!Ifc8 18 lLic3 a3!) 17 .....tfs 18 'i'h4
!Ifc8 19 @d2 !IxC3! 20 bxc3 'it'xds+ and
Black had a winning attack in L.Yurtaev­
Y. Shulman, Vladivostok 199S.
9 'it'e8

Instead 9 ... h6 10 ..ie3 shows one

9 ..tgs idea behind White's 8th move (there is
This pin is a typical weapon in both no ...lLig4), and Black has weakened his
the Petrosian and Makogonov varia­ kingside in order to break the pin. I pre­
tions. White tries to entice Black to fer the text move.
weaken his kingside. Other moves are
less dangerous:
a) 9 ..ie3 lLics (9 ...lLih s is also possi­
ble) 10 lLid2 lLie8 11 g4 fs 12 gxfs gxfs
13 exfs (13 'it'c2 lLia6! threatens ... f4
and then 14 exfs .�xfs 1s .�d3? is not
possible because of 1s ...lLib4) 13 ... ..ixfs
gives Black excellent play. Note that
Black's h-pawn would be much more of
a target on h6 here, which is one point
of White's 9 ..igs.
b) 9 g4 is not well timed: 9...lLics 10 The move ... 'it'e8 is very typical in
'it'c2 c6 11 ..ie3 a4 (better than 11 ...cxds the Petrosian Variation. Black breaks
12 cxds when bS is weak) 12 o-o-o the pin without weakening his kingside
(White cannot even take a pawn be­ and observes the a4-e8 diagonal. Here
cause 12 ..ixcs dxcs 13 lLixa4 runs into he hopes to show that h3 cost a tempo
13 ...lLixe4!) 12 ... cxds 13 cxds 'ifas 14 and weakened the white position,
lLid2 ..id7 1S gs (this runs into a great while the gs-bishop may also be mis­
shot, but after 1S lLic4 'ife7 Black threat­ placed. White has two important con­
ens ... bs and 16 lLia3 can be met by tinuations:

Petros i a n Variati o n : 7 dS as with o u t 8 j,_ 9 5

01: 10 g4 a) 11 lLid2 lLidcS 12 lLif1 (in stead 12
02: 10 �d2 h4 could be met by 12 .. .fs or 12 ...f6, 12
a3 should be met by 12 ...j_d7, and 12
Dl) 10 g4 !Ig1 transposes to the notes to White's
12th move in our main line, below) is
the same as variation 'a1' in the notes
to White's 12th move below, but with­
out l:i:g1 and ... @h8. Here the position
looks even better for Black after 12 .. .f6
OT 12 .. .fS.

b) 11 'it'd2 lLidcS 12 l:i:g 1 lLib4! cre­

ates tactical problems for White. Black
intends ... @h8 and .. .fs, and White can­
not play a3 because of the fork on b3.
c) 11 a3 and now 11 ...lLidcS? is met
This lunge can be the prelude to an by 13 b4 because of the pin along the
attempted kingside attack, but often a-file, but Black has 11 ...lLib6 ! .
White is just hoping to limit Black's
10 lLid7!

Black can also play 10...lLics, but af­

ter 11 lLid2 but the f6-knight is in the
way. By playing the king's knight to cs,
Black keeps the possibility of playing
...fs and ... j_xfs. Black has also tried to
move the knight to g8, but not surpris­
ingly, this is less dynamic. Indeed,
10 ...@h8 11 lLid2 (or 11 l:!.g1 lLig8)
11 ... lLig8 12 h4 f6 13 j_e3 fs 14 f3 (after This i s an unusual square for the
14 gxfs gxfs 15 exfs j_xfs 16 lLide4 lLif6 black knight, but White has made a lot
Black had decent play in T.Polak­ of pawn moves and so Black adapts to
S.Novikov, Pardubice 2007) 14 ...f4 15 the situation. He plans .. .fs or ...j_d7
j_f2 is better for White, because all the and ... lLia4. After 12 lLid2 j_d7 13 h4
play will shift to the queenside, where practice has seen:
White has a significant space advan­ c1) 13 ... lLics 14 hs fs (14 ... lLiba4!?)
tage. 15 hxg6?! (better is 15 gxfs gxfs 16
11 !Igl j_e3) 1s ...'it'xg6 16 j_e7 fxe4 17 j_xfg

A ttacking C h e s s : Th e Kin g 's I n dian, Vo l u m e 1

!Ixf8 gave Black compensation in to Black in both cases) 1 8...lLixb2 19

B.Thorfinnsson-E.Berg, Reykjavik 2009, 'it'e2 'it'd7 is good for Black. The knight
but 16 .. .fxg4! 17 ..ixf8 !Ixf8 would have can escape with ...lLia4 and 20 'it'xb2 is
been even stronger. met by 20.....ixe4 21 lLixe4 'it'xh3+, win­
c2) 13 ... lLia4! 14 lLibs lLi4cS 1s hs fs ning.
16 gxfs (not 16 hxg6 'it'xg6, hitting the a2) 16 lLihs with a further divide:
bishop) was agreed drawn here in
L.Nisipeanu-V.Nevednichy, Sovata
1999. After 16 ... gxfs 17 l:i:g1 @h8 Black
has good counterplay.
11 lLidcs

11 ... @h8 12 'it'd2 lLidcS (J.Alvarez

gives 13 ..ih6 ..ixh6 14 'it'xh6 f6 1S h4
as better for White, but I prefer Black
after 1S ... ..id7) 13 o-o-o ..id7 (or 13 ... a4)
is a good alternative.

a21) 16 ... 'iff7 is tricky: 17 ..ih4 ..ih6!

(Black should avoid the complications
of 17 ... l:i:g8 18 !Ixg7 !Ixg7 19 ..if6 'it'xf6
20 lLlxf6 !Ig1+ 21 ..if1 ..ixh3 22 'it'hs) 18
..if6+ 'it'xf6 19 lLixf6 l:i:xf6 gave Black
compensation for the queen in
A.Aleksandrov-1.Zakharevich, St Peters­
burg 1994.
a22) 16 ...'it'g6! 17 lLixg7 (one point
of Black's play is that after 17 .l:f.g3 'iff7
12 h4 the move 18 ..ih4 does not protect f2,
Instead 12 'it'd2 lLib4! is note 'b' to but Black has even better in the shock­
White's 11th, above. White's main al­ ing 17 ... ..ig4!! 18 ..ie7 'it'xhs 19 ..ixg4
ternative is 12 lLid2 @h8 and here: 'iff7 20 ..ixf8 l:i:xf8 with a strong attack)
a) 13 lLif1?! is slow and a bit clumsy 17 ... 'it'xg7 18 'it'd2 lLib4 (Black does not
after 13 .. .fs 14 gxfs gxfs 1s lLig3 fxe4! miss his dark-squared bishop because
and now: his knights are tremendous) 19 o-o-o
al) 16 lLigxe4 .ifs 17 ..ihs lLid3+! lLixa2+! 20 @b1 lLixc3+ 21 bxc3 .ifs
(even better than 17...'it'd7) 18 @f1 (al­ was winning for Black in O.Jovanic­
ternatives are worse: 18 'it'xd3 'it'xhs or Z.Kozul, Ljubljana 2004.
18 @d2 lLixb2, with a clear advantage b) 13 a3 and now:

Petro s i a n Va ria tion: 7 ds as with o u t 8 j,_95

bl) 13 ...j_d7 14 h4 f5 15 gxf5 (worse I like this little nudging move, al­
is 15 h5 fxg4) 15 ... gxf5 16 h5 j_f6 17 though 13 .. .f5 14 gxf5 gxf5 15 j_h6
�xf6+ !Ixf6 18 'it' c2 'it'f8 19 o-o-o fxe4 comes to the same thing.
20 lLicxe4 l:i:f4 21 f3 j_f5 22 !Ig2 'it'h6 14 j_e3 f5 15 gxf5 gxf5 16 j_h6
with counterplay, M. lvanov-0.Cvitan,
Cappelle la Grande 1995.
b2) 13 ...a4 is simple enough: 14 'it'c2
f5 15 gxf5 gxf5 16 o-o-o fxe4 17 lLidxe4
�f5 gives Black good play.

This may look dangerous, but Black

has enough resources to hold the bal­
ance. Note that strategically Black does
not mind the exchange of dark­
squared bishops as Jong as his king re­
12 ...@hB! mains safe.
A typical reaction to White's rook 16 ... j_xh6
move. Instead 12 ... c6 is inconsistent Or 16 ... l:i:g8 17 j_xg7+ l:i:xg7 18 !Ixg7
and looks panicky to me, but Black may @xg7 19 'it'g 5+ 'it' g6 20 'it'e7+ 'it'f7 with
be able to get away with it: 13 h5 cxd5 equality according to Zvjaginsev.
14 h6 j_h8 15 cxd5 (after 15 lLixd5 f6 16 17 'it'xh6 'it'e7 18 h5 !If6
..i.e3 Black should avoid 16 ... lLixe4 17 Instead 18 ...!If7 19 !Ig6 j_d7 20 lLig 5
�b6 j_e6 18 lLixa8 'it'xa8 19 lLid2, when !Ig7 is also okay, J.lvanov-J.lbarra Jerez,
White is much better according to Mondariz 2000.
Piket, and instead play 16 ... j_e6!) 19 !Ig6 .l:f.xg6 20 hxg6 'it'g7
15 ...j_d7 16 a4 lLic7! ( 17 axb5 Black quickly takes over the initia­
�c7 18 b6 lLib5 19 lLixb5 j_xb5 20 lLid2 tive in the endgame.
gave White a clear advantage in J.Piket­ 21 '1Wxg7+ @xg7 22 exf5 j_xf5 23 gxh7
J.Polgar, 1st matchgame, Aruba 1995) lLib4 24 @d2 j_xh7 25 !Igl+ @f6 26
17 lLid2 b5 18 axb5 lLixb5 gave Black lLig5 j_g6 21 f3 !Ihs
counterplay in S.Giddins-R.Leitao, Ant­ Black had some advantage in the
werp 1998. ending in J.Alvarez-E.Pupo, Holguin
13 'it'd2 f6 1997.

A ttacking C h e s s : The King 's I n dian, V o l u m e 1

02) 1o ll:id2 17 b4 axb4 1 8 axb4 �xal 19 '1!r'xa1 ll:ib3

20 '1!r'b2 ll:id4 21 it.d3 f5 22 ll:ie2 fxe4 23
it.xe4 ll:ixe2 24 '1!r'xe2 '1lr'a4 25 '1!r'b2 it.d7
with the initiative in J.Tihonov­
A.Fedorov, Minsk 2010.
13 ... il.h6 14 l':tbl
Or 14 o-o ll:ic5 15 �a2!? (15 l:tb1)
15 ...f5 16 f3 il.e3+ 17 il.f2 il.xf2+ 18
.l:!.xf2 ll:ibd7 (other possibilities are
18 ... '1lr'e7 and 18 19 b4 axb4 20
axb4 �xa2 21 ll:ixa2 fxe4! 22 fxe4 (22
bxc5 e3) 22 ...l:txf2 23 '1¥o>xf2 '1lr'f8+ 24
A flexible move. '1¥o>g1 ll:ia6 with equality in E.Mirosh­
10...ll:ifd7 11 a3 nichen ko-A.Tukhaev, Kharkov 2007.
Instead 11 g4 ll:idc5 takes play back 14...ll:ics
into note 'a' to White's 11th move in
Line Dl, while 11 o-o ll:idc5 looks fine
for Black
11...f6! 12 il.h4
After 12 it.e3 Black plays 12 ...f5 im­
mediately, with good play.
12 ...ll:ib6
Again we see this move. Instead
12 ... f5? 13 exf5 gxf5 loses an exchange
after 14 iLh5, but 12 ... il.h6!? is possible.
Then 13 b4 ll:ib6 14 �bl axb4 (or when 15 bxa5?! ll:ia4! and 15 b4
... ll:ia6-c5 is good for Black) 15 axb4 f5 After this move Black takes over the
16 o-o it.d7 17 '1!r'c2 ll:ia4 18 ll:ixa4 Jt.xa4 initiative. 15 o-o is more solid and after
19 '1lr'd3 il.d7 gave Black a good position 1 5 ... il.d7 (instead 1 5 .. .f5 16 b4 axb4 17
in D.Brunsteins-Y.Yarmolyuk, corre­ axb4 ll:ica4 18 ll:ib5 was awkward for
spondence 2006. Black in M.Krasenkow-A.Vovk, Vlissin­
13 b3- gen 2009) White has tried:
13 b4 is not possible, because a) 16 '1!r'c2 f5 17 b4 ll:ica4 (17 ... axb4
Black's last move protected the a8-rook. 18 axb4 ll:ica4) 18 ll:ib5 Jt.xb5 19 cxb5
Instead 13 f3 just invites 13 ... il.h6. A '1lr'f7 was unclear in A.Jorgensen­
good example of Black's play went 14 P .Hertel, correspondence 2003.
Ji.f2 ll:ic5 1 5 ll:ib3 ll:iba4! 16 ll:ixc5 ll:ixc5 b) 16 b4 axb4 17 axb4 ll:iba4 18

Petrosian Varia t io n : 7 d5 a5 with o u t 8

ll:ixa4 il..xa4! 19 '1!r'e1 il..xd2 20 '1!r'xd2 white queen. This is solid enough, but
ll:ixe4 21 '1!r'e3 fS snared an important he could consider 18 ... cs!? or even the
centre pawn for Black in S.Nenciulescu­ forcing 18 ... il..xd2+!? 19 '1¥o>xd2 c6. After
D.Matic, correspondence 20os, al­ 20 dxc6 bxc6 21 cs+ dS 22 cxb6 cxbs 23
though White can claim some compen­ '1lr'xds+ '1lr'xds+ 24 exds ll:ixb6 2s il.. xbs
sation with his bishop-pair. �a2+ 26 '1¥o>e1 ii.ts Black has a strong
15 ...axb4 16 axb4 ll:ica4 17 ll:ibs initiative in the ending.
After 17 '1!r'c2 ll:ixc3 18 '1!r'xc3 Bl ack 19 �d1 c6 20 ll:ia3 cs 21 bs ll:id7 22
has 18 ...�a2, intending .. .fs with a clear ll:iab1 '1!r'as
advantage. Black intends ...'1lr'b4 with the initia­
17 ...'1!r'd8! 18 '1lr'b3 '1¥o>g7 tive, N.Legky-A.Shirov, French League
Black gets off the diagonal of the 199S.

Chapter 12
Petrosia n Variation

1 ds a s a ..tgs

1 d4 lLif6 2 c4 g6 3 lLic3 1lg7 4 e4 d6 S when it is difficult for Black to create

tLif3 o-o 6 1le2 es 7 ds as 8 counterpl ay. Black has sometimes tried
9...�e8 here, as in Line D of the previ­
ous chapter, but White has not wasted
time on the weakening move h2-h3 in
this line.
9 1lh4 lLia6 10 lLid2
This is the most common move, but
White can also play 10 o-o �e8 11 lLJd2
tran sposing to the main lines. Instead
11 lLiel is sometimes played. After
11 ... gs (worse is 11 ... tLics 12 f3 lLih s 13
lLibs, but Black can keep the game
This is White's main move in the tense with 11 ... 1i.d7!? 12 lLJd3 lLJh 7 13
Petrosian Variation. By pinning the f6- f3 fS) 12 1lg3 Black has the well-timed
knight he makes it difficult for Black to 12 ... tLixe4! (a common motif that is al­
get in .. .fs. This variation has been fa­ ways worth checking ) 13 tLixe4 fS 14
voured with White by GM Igor Naum­ ..\;lhs �e7 15 f3 fxe4 16 fxe4 �xfl+ 17
kin for over 25 years. '1¥o>xf1 tLics 18 �e2 g4! (without this
8 h6
... move Black would just be worse) 19
In this case breaking the pin is ..\;lxg4 1lxg4 20 �xg4 �g s! when the
Black's best because retreating to e3 e4-pawn falls after 21 �e2 �g6 or 21
can be met by ...lLig4. It is important to �xg s hxgs.
play this move immediately, else White 10 �eB

can play 8 ... lLia6 9 lLJd2 h6 10 1le3 Generally 10 ... 1i.d7 will just trans-

Pe tros ian Va ria t i o n : 7 dS a s 8

pose after 11 o-o (or 11 a3 �e8 12 b3) the pin on the a-file: for example,
11 ...�e8. 11 ... lLJh 7?! 12 b4! saves White some

White's basic plan is to play o-o, a3,

b3 (usually necessary at some point if 12 b3 lLih7
White wishes to avoid his pawns being This is a multi-purpose move. Black
fixed by ... a4), l':tbl and only then b4. prepares either .. .fs or ... hs and ....ii.h6.
However, the move order is flexible for Another idea is 12 ... tLixe4!? 13 lLJdxe4
both sides. Usually White castles here fS 14 f3 fxe4 15 tLixe4 il.fs. If Black can
or on the next couple of moves, but he fi ght for the e4-square with pieces, as
can also delay or omit castling in order he can here, this idea is usually good
to accelerate his queenside play. We enough for equality.
will examine: 13 l':tbl
Instead 13 o-o takes play into the
A: 11 a3 (without an early 0-0) realm of Line B.
B: 11 0-0

Instead 11 b3 lLih7 12 a3 returns

play to Line A, as does 11 l':tbl il.d7 12
b3 lLih7 13 a3, while the lunge 11 g4 is
well met by 11...lLih7!.

A) 11 a3
Here we only examine lines where
White delays castling.
This develops and protects the a8- 13 ... h s
rook. Black must always be aware of Black prepares t o activate his bishop

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

on h6. This idea is usually more effec­ il.xbs 2 0 cxbs e4! with dangerous ideas
tive than pushing the f-pawn, but here like ... ll:ic3 and ... gs.
13 .. .fS!? is quite viable for tactical rea­ 14 f 3 il.h6 1S il.f2
sons: 14 f3 (14 exfs il.xfs! is one point, Again White can transpose to Line B
because 1s ll:ide4? walks into 1S ... il.xe4 with lS o-o.
16 ll:ixe4 �f4) 14 ... ll:ics (Black does not
want the a6-knight to be sidelined; in­
deed, this looks better than 14...ll:if6 lS
b4 axb4 16 axb4 ll:ihs 17 1lf2 ll:if4 18
..\;lf1 when Black's knights are ineffec­
tive) and now:
a) lS b4 axb4 16 axb4 ll:ia4 17 '1!r'c2
(not 17 ll:ibs? ..\;lxbs 18 cxbs ll:ic3 and
wins) 17...ll:ixc3 18 '1!r'xc3 ll:if6 gives
Black comfortable play: 19 cs?! can be
met by 19 ... gs 20 1lf2 fxe4 or 19 ... !
when White may regret his decision to 1s '1!r'e1

postpone castling for so long. Worse is 1s ...fs 16 b4, but 1s ...ll:ics !?

b) lS '1!r'c2 ll:if6 16 b4 axb4 17 axb4 is certainly worth consideration. After
ll:ia4 18 ll:ibs (after 18 o-o ll:ixc3 19 16 '1!r'c2 '1lr'e7 17 b4 axb4 18 axb4 ll:ia4
'1!r'xc3 ll:ihs Black has good counterplay: 19 ll:ibs White must goes on an adven­
for example, 20 l':tfcl ll:if4 21 1lf1 gs 22 ture to fight for the initiative. Follow­
1lf2 '1!r'g6 23 cs g4). ing 19 ... c6 there is:
a) 20 ll:ia7 cxds 21 cxds (better is 21
exds, but Black's position is healthy
here too and even the pawn sacrifice
21 ...e4 looks promising) 21 ...l':txa7! 22
..\;lxa7 l:tc8 gives Black a very strong ini­
b) 20 ll:ie7 l':tac8 21 dxc6 (21 '1!r'xa4
l':txe7) 21 ... bxc6 22 ll:ia6 cs 23 l':tal!? (23
bS ll:ib6 is unclear) 23 ... cxb4! ? with a
further divide:
bl) 24 l':txa4 il.xa4 2S '1!r'xa4 gives
Here 18 ... il.xbs 19 cxbs '1lr'd7 20 o-o Black a pleasant choice between
ll:ihs 21 l':tfcl ll:if4 22 1lf1 has scored 2s ... il.xd2+ 26 '1¥o>xd2 '1lr'gs+ 27 .ii.e3
well for White in practice, but instead '1!r'xg2 and 2s ... '1!r'gs 26 '1!r'xb4 '1!r'xg2 27
Black has 18 ...ll:ixds!, intending 19 exds �gl '1!r'xh2.

Petro s i a n Varia tion : 7 d5 a5 8 1l g 5

b2) 24 ll:ixb4 �b8 2s ll:ids (after 2s a ) 1 9 bxcs!? ll:ixcs 2 0 ll:ic4 ll:ia4

l:txa4 il.xa4 26 �xa4 Black again has (20 ...h 3 ! ?} 21 d6 �gs is very messy.
the 26 ...�g s idea) 2s ...�gs 26 ll:ie3 ll:ics b} 19 bS ll:ib4 20 ll:ic4 and here
is pretty level. rather than 20 ... b6 21 d6 cxd6 22 ll:ixb6
16 b4 OT 20 ...ll:ia2 21 �C2 ll:ixc3 22 �XC3, in

Instead 16 h4 is very ambitious. both cases with compensation for

White prevents Black from gaining White, I suggest 20 ... h 3 or 20 ...�fd8!?.
space on the kingside with ...h4, but 18 ...l:tfbB!?
makes castling kingside less attractive: This prophylactic move looks
16 ...ll:ics 17 �c2 fS 18 b4 axb4 19 axb4 stronger than the 18 ... �g S 19 �gl c6
ll:ia4 20 ll:ibs has been played a few (19 ...b6 is safer) 20 cs! of P.Rossiter­
times by Zilberman. After 20 ...c6 21 J.Gallagher, British Championship,
ll:ic7 �ac8 22 ll:ie6 1lxe6 (22...cxdS !? 2 3 Swansea 1987.
tllxf8 ll:ixf8 24 exds e4 was a bit specu­ 19 1ld3 �gs 2o l:tg1 cs!
lative in Y.Zilberman-J. Becerra Rivero, A useful idea to know. Black seizes
Havana 1998) 23 dxe6 bS 24 0-0 �xe6 the initiative right across the board.
2S .rf.bdl ll:if6 matters were unclear in 21 dxc6 bxc6 22 bs ll:ics 23 g3 hxg3 24
Y.Zilberman-S.Porat, Tel Aviv 2002, al­ hxg3 lLif8 2S ll:ib3 ll:ixb3 26 .t:!.xb3 ll:ie6
though Black was up a pawn. With active pieces and a safer king,
16...axb4 17 axb4 h4 Black was much better in G.Danner­
G.Timoscenko, Budapest 1989.

B) 11 0-0

A thematic move. Black gains some

space and creates various possibilities
an the kingside, such as ... il.f4, ...�g s,

-�tf6-hS and ... h3. Castling at some point is White's

• Wc2 most logical way of playing.
White has also played the pawn sac­ 11...ll:ih 7
nfice 18 cs!? dxcs and now: Here too Black can play 11 ... il.d7

A ttacking C h e s s : Th e King 's I n dian, Vo l u m e 1

which will usually transpose to the this, then White will play b2-b4 in one
main line after 12 b3 (or 12 '1¥o>h1 ll:ih7, move. Conversely Black should usually
while 12 a3 can be met by 12 ... a4 13 wait for l':tbl before playing this ad­
ll:ibs il.xbs 14 cxbs ll:ics 15 f3 b6) vance, because after the sequence ... a4;
12 ...ll:ih7 13 a3. b4 axb3; ll:ixb3 White's rook is usually
12 a3 better posted on al, from where it de­
White threatens 13 b4. Instead 12 fends the a3-pawn and supports the
'1¥o>h1 h S 13 f3 il.h6 14 a3 il.d7 15 l':tbl advance a3-a4-aS.
(15 b3 transposes to the main line)
runs into 1S ...a4! (if Black does not play
this then White will save a tempo on
b2-b3), and then 16 ll:ibs b6 or 16 b4
axb3 17 ll:ixb3 b6, with a good game for
Black in both cases.
White has also been known to go 12
f3 when 12 ... hs, 12, and 12 .. .fs
are all reasonable and may well trans­
pose to the main line.
12 il.d7

Instead 12 ...hs 13 f3 il.d7 trans­ Now, after 14 ll:ibs hS (the most

poses, but here Black has to avoid flexible, but 15 cxbs ll:ics 16
13 ... il.h6?! 14 b4!. Premature too is f3 b6 is fine too) 15 f3 il.h6 (Black
12 .. .fs 13 exfs il.xfs (13 ... gxfs 14 il.hs is should probably avoid 15 ... il.xbs 16
annoying) 14 g4! (not 14 ll:ide4?? il.xe4 cxbs ll:ics 17 b6!?) 16 il.f2 (or 16 b4
15 ll:ixe4 �f4) 14 ... il.d7 15 ll:ide4 with axb3 17 ll:ixb3 il.e3+) Black has:
an advantage for White in A.Veingold­ bl) 16 ...b6 17 b4 axb3 18 ll:ixb3
G.Kasparov, Moscow 1979. il.xbs 19 cxbs ll:ics 20 ll:ixcs dxcs 21 a4
13 b3 '1lr'e7 22 '1!r'c2 ll:if6 23 l':tal ll:id7 24 il.d3
The alternatives are also important: l':ta7 25 z:ta2 �fa8 26 l':tfal h4 27 '1¥o>h1
a) 13 ll:ibs hs 14 f3 il.h6 15 '1¥o>h1 (15 '1¥o>g7 28 '1!r'e2 (White intends il.el and
b3 is the main line and 15 l':tbl a4 varia­ as, so Black sacrifices an exchange to
tion 'b', but 15 '1!r'c2!? is possible) 1S ... il.e3 keep the position closed) 28 ... z:tas 29
16 l':tbl and here 16 ...a4! can be played il.e1 il.f4 30 il.xas l':txas 31 il.c4 ll:if6 32
because the d2-knight is loose. d6! (White gives back a pawn to open
b) 13 l':tbl is well met by 13 ... a4. lines for his rooks) 32 ...'1!r'xd6 33 l':tdl
This advance is usually good after '1lr'e7 34 '1!r'd3 l':ta8 35 l':ta3 with a slight
White has spent a tempo moving his advantage for White in D.Vigorito­
rook. Moreover, if Black does not play A.Sherzer, Philadelphia 1997.

Petro s i a n Va ri a t i o n : 7 d5 a5 8 il.g5

b2) 16 ... '1!r'e7 is more flexible. Black to cover the a3-pawn and play a4-a5, in
does not need rush to exchange on b5. which case here White could claim to
After 17 b4 (17 '1lr'c2 can be met by be two tempi ahead. After 14 ll:ib5
17... h4 or 17 ...b6) 11... axb3 18 ll:ixb3 b6 Black has:
19 a4 f5 20 ll:id2 (White should proba­ cl) 15 cxb5 ll:ic5 16 f3 and
bly change gears and play 20 exf5) although it seems that Black is not so
20 ... ll:ic5! Black developed pressure badly off, White has saved time and
against the e4-pawn in D.Vigorito­ scored well from here.
A.Matikozian, Los Angeles 2003. Note c2) 14 ... h5 15 f3 il.h6 and now:
that 21 ll:ixc7? loses to 21 ... il.xa4. c21) 16 ll:ib1!? il.xb5 17 cxb5 ll:ic5 18
c) 13 '1¥o>h1 ! ? is a way of reaching the il.c4 (with the idea of il.f2 and ll:ic3 to
main lines of 'B2' and 'B3', below, with­ target the a4-pawn) 18 ... il.e3 19 ll:ic3 f5
out allowing Black the option of 13 b3 (19 ... b6) 20 exf5 gxf5 21 '1!r'c2 '1!r'g6 22
f5 (after 13 '1¥o>h1 f5 White isn't vulner­ l':tael '1!r'h6 (22 also runs into 23
able down the long diagonal, so can b6! cxb6 24ll:ib5 and 22 6 23 il.f2 is
play 14 exf5! with advantage). Black no help) 23 b6! favoured White in
can deviate as well, but it is probably J.Nogueiras-A.Zapata, Matanzas 1994.
better to transpose to the main line c22) 16 b4 is the most logical when
with 13 ... h5 14 f3 il.h6 15 b3 (15 �b1 16 ... axb3 17 ll:ixb3 has scored tremen­
a4! was briefly discussed in the notes dously for White, even though Black's
to White's 12th, above). position is not so terrible. That said,
11 .. .f5 (11 ... b6 18 a4!) 18 exf5 il.xf5
(18 ... gxf5 19 f4! also looks good for
White) 19 il.d3 was still pleasant for
White in l.Yanvarjov-0. Loskutov, Mos­
cow 1994.
Returning to 13 b3:

Instead 13...a4 has been tried too,

but White has made a certain gain in
that he has reached this structure
without spending a tempo on �bl as
explained above. In fact, sometimes
(after l':tb1) the white rook return s to al

22 7
A ttacking C h e s s : Th e Kin g 's I n dian, Vo l u m e 1

13 h s
... il.hs �c8 (Kasparov's exchange sacri-
Black has a sharp alternative in fice) 16 il.e7 l':te8 17 il.xe8 �xe8 18 il.h4
13 .. .fS!?. e4. White has tried both 19 l':tcl and 19
�c2 here with no clear conclusions.
b2) looks quite sound, in­
tending lS g4 e4 (1s .. .il.d7 16 ll:ide4 b6
is also feasible) 16 l':tcl e3!.

Then :
a) 14 f3 is played rather infre­
quently, but it may be best:
al) 14...ll:if6 lS il.f2 (after lS l':tbl
ll:ihs 16 �e1 ll:if4 17 il.fl gs 18 il.f2 g4 White has:
Black has counterplay) lS ...ll:ihs 16 g3 b21) 17 fxe3 �xe3+ 18 il.f2 �g s 19
and it is not s o easy for Black t o create h3 (no better are 19 '1¥o>h1 il.d7 20 ll:ide4
play on the kingside. �e7 or 19 h4 �f4 20 gxfs 21 lLif3
a2) 14...b6 lS �bl f4 16 b4 axb4 17 �g4+ 22 '1¥o>h1 �h3+ with a draw)
axb4 gs 18 il.f2 hS transposes to varia­ 19 ... il.d7 2o ll:if3 �f4 is level.
tion 'a3'. b22) 17 gxfs exd2 18 �xd2 ll:ics 19
a3) 14... f4 lS l':tbl (White could also �dl (worse is 19 il.dl �xfs 20 �e2?
play lS il.f2 or even lS b4!? axb4 16 ll:igs 21 �xe8+ �xe8 22 ll:ibs ll:ixb3
axb4 ll:ixb4 17 �3 ll:ia6 18 �xb7 �8 with a clear advantage, P.Enders­
19 l':tfb1 with the initiative) lS... gs 16 V.Babula, Berlin 1996) 19 ...�xfs 20 b4
il.f2 hS 17 b4 axb4 18 axb4 b6 (18...�g6 (after 20 il.g4 �f4 21 �el Black can
19 cs g4 20 c6 gives White the advan­ play 21...�f8 or sacrifice his queen with
tage) 19 cs! bxcs 20 bxcs ll:ixcs 21 21 ...�xel+ 22 �xel �xg4+ 23 il.g3
il.xcs dxcs 22 d6 (22 ll:ic4 and 22 ll:ibs ll:igs 24 �e2 hS) 20 ... axb4 21 axb4 ll:ie4
are good alternatives) 22 ...cxd6 23 ll:ic4 22 ll:ibs l':tf7 23 il.g4 lLihg S with ap­
and White had the initiative in J.Kraai­ proximate equality in l.Naumkin­
D.Gross, German League 1998. D.lsonzo, Lido Estensi 2003.
b) 14 exfs and here: While 13 ...fs is playable, I prefer
bl) 14. ..gxfs is very sharp after lS 13 ... h s to which we now return. Do also

Petrosian Varia t i o n : 7 d5 a 5 8 1lg5

note that White can sidestep 13 .. .fs This keeps Black's bishop out of e3,
with 13 '1¥o>h l !? as discussed above. but it allows Black to improve the posi­
14 f3 1lh6 tion of his queen.
15 '1lr'e7!

From here the queen helps to hold

up White's cs-advan ce, while it may
become active via the gs-square. Typi­
cal ideas for Black include
...h4,...1lf4, ...ll:if6-hS and ... '1¥o>g 7 when a
piece may hop into g3 and ... l':th8 will
be a possibility too.
16 '1!r'c2
By far the most common move, but
there are others:
Black has activated his bishop, a) 16 �bl '1!r'g s looks strong, but af­
which may come to e3. White can pre­ ter 17 l':tb2 .lii.h3 18 .lii. g3 h4 (Black can
vent the intrusion, dodge the check or force a draw with 18 ... '1lr'e3+ 19 1lf2
continue his queen side play. We have: '1lr'g s, but must avoid 19 ... '1lr'xC3? 20
l':tc2) 19 f4 (not 19 gxh 3 hxg3) 19 ... exf4
81: 1s i.f2 20 ll:if3 '1lr'f6 21 es dxes 22 il..xh4 g s 23
82: 15 �h1 gxh3 gxh4 24 .lii.d3 matters are not so
83: 1s .:b1 clear. Black could also just play the
move 16 ... h4.
Line Bl is independent, whereas b) 16 �a2 protects the d2-knig ht.
Line B2 can easily transpose into B3. Black should just play 16 ...h4.
c) 16 l':tel h4 17 il..fl 1lf4 18 �bl
81) 15 1lf2 '1!r'g s 19 �b2 ll:if6 intending ... ll:ih s gives
Black a strong attack.
d) 16 '1¥o>h 1 seems to be asking for
trouble because a knight may come to
g 3 with check. Following 16 ...h4 17
�gl?! (even after 17 �bl ll:ics 18 b4
axb4 19 axb4 ll:ia4 20 ll:ixa4 �xa4 21 cs
h3 22 g3 z:ta2 23 ll:ib3 fS Black had the
initiative in V.Antoshin-B.Kantsler, Ni­
kolaev 1981) 17 ....lii . gs 18 '1!r'e1 '1¥o>g7 19
�bl ll:ics 20 b4 axb4 21 axb4 ll:ia4 22
ll:ib3 ll:ixC3 23 '1lr'xC3 lLif6 24 �al ll:ih s 2s

A ttacking C h e s s : Th e K i n g 's I n d i a n , Vo l u m e 1

�a8 �xa8 26 l':tal l':th8 27 '1¥o>g1 il.f4 28

il.e3 '1!r'g s 29 �2 fS Black had enduring
pressure in P.Kiss-V.Kotronias, Kavala
16...h4 17 �fdl

Now 2 1...ll:ihs was given as clearly

better for Black by Strikovic in Infor­
mant 45, based on 22 cs dxcs 23 bxcs
ll:ixcs 24 il.xcs? il.e3+ 2s il.xe3 '1!r'xe3+
26 '1¥o>h1 ll:ig3+ 27 hxg3 hxg3 and 0-1 in
11 ... il.f4 L.Santa Torres-A.Strikovic, New York
Black can also lead with the queen. 1988. However, White later improved
One successful example of this idea with 24 ll:ic4! in L.Santa Torres-M.Al
was 11 ...'1!r'g s 18 il.f1 ll:if6 19 l':tab1 ll:ihs Modiahki, Moscow Olympiad 1994,
20 b4 axb4 21 axb4 fS 22 cs h3 23 g3 f4 when he had good compensation for
24 g4 il.xg4 2S fxg4 '1!r'xg4+ 26 '1¥o>h1 the pawn. Instead I think that Black
ll:ig3+! 27 hxg3 fxg3 28 il.e2 g2+ 29 should play:
'it>gl h2+ 0-1, G.Flear-R.Douven, Charl­!
ton 1983. Black has a good position.
18 �ab1
18 ll:ifl '1!r'g s 19 il.d3 ll:if6 20 �ab1 82) 15 'it>hl
was drawn here in B.Malich­
W.Schmidt, Decin 1976, but Black could
certainly continue with 20 ...ll:ics,
20...'i¥o>g7 OT 20 ...ltJh S.
18 ...'1!r'gs 19 il.f1 ll:if6
Black has a couple of alternatives
here that should be investigated.
19 ... il.e3 fights for the dark squares and
19 ...ll:ics avoids Black's knight being
potentially sidelined on a6.
20 b4 axb4 21 axb4

Petro s i a n Va ria tio n : 7 d5 a5 8 il.95

This is a tricky prophylactic move. z:ta2 '1¥o>g7 17 '1!r'b1 il.cs 18 ll:id1 c6 19

Often play will transpose to Line B3, ll:ib2 should be met by 19 ... il.d4 20 ll:id3
but here we will only look at independ­ ll:ics with equality, rather than
ent lines where White generally avoids 19...cxds?! 20 cxds ll:ic7 21 ll:id3 il.d4
or delays l':t bl. (M. Narciso Dublan-F.Nijboer, Barbera
1s il.e3
... del Valles 2007) when 22 ll:ic4 gives
Black has a couple of alternatives: White a large advantage after either
a) 1s ...'1!r'b8 intends to bring the 22 ... gs 23 il.e1 '1!r'e7 24 ll:ixas or
queen to a7. Then 16 '1!r'c2 il.e3 (or 22...ll:ibs 23 �cl ll:ic3 24 l':txc3 il.xc3 2S
16 ...'1!r'a7 17 l':tael!) tran sposes to note ll:ixd6 '1!r'b8 26 il.e7.
'f to his 16th move, below, but White
could also play 16 il.f2 when 16 ...'1!r'd8
17 '1lr'c2 gives him an extra tempo over
Line Bl, with Black's queen on d8 in­
stead of the more flexible e1-square.
b) 1s ... ll:ics is worth considering. Af­
ter 16 l:tb1 (but not 16 b4? il.xd2) 16 .. .fs
17 '1!r'c2 ll:if6 18 b4 axb4 19 axb4 ll:ia4
White has:
bl) 20 exfs ll:ixc3 21 '1!r'xc3 il.xfs 22
..i..d3 l':ta2 23 ll:ie4 .�xe4! 24 fxe4 ll:ig4
gave Black the initiative in l.Naumkin­ 16 fs

A.Khalifman, USSR 1984. This move is very natural, but Black

b2) 20 ll:ibs was played in must keep in mind that White can also
l.Naumkin-J.Trapl, Namestovo 1987. play on the kingside, especially if
Black's best is 21 cxbs il.xd2 Black's dark-squared bishop goes wan­
22 '1!r'xd2 fxe4 23 il.xf6 �xf6 24 fxe4 dering off. The position is very rich and
which Palliser considers slightly better there are several possibilities, of which
for White because of the open c-file, 'f' is the most interesting:
but after 24 ... '1lr'e7 Black is probably a) 16 ... c6 looks premature: 17 il.f2
close to equality. The e4-pawn i s also a (even 17 cs!? is dangerous: for exam­
potential weakn ess and White's bishop ple, 17 18 ll:ic4 intending ll:ia4
is no better than Black's knight. with the initiative) 17 ...Lf2 18 �xf2
16 '1!r'c2 '1!r'd8 19 l:tffl ll:if6 20 '1!r'b2 h4 21 f4!
This is critical. White does not (with the dark-squared bishops ex­
bother with l':tabl and instead prepares changed and Black's knights unable to
for war across the board. Instead 16 reach the es-square, this is a strong
:b1 transposes to Line B3, while 16 break) 21 ... h3 22 g3 ll:ig4 23 fS ll:ie3 24

A t tacking C h e s s : The King 's I n dia n , Vo l u m e 1

�f3 '1!r'b6 25 lLifl ll:ig4 26 '1!r'd2! '1¥o>g7 ll:ics in D.Cummings-P.Cramling, Pula

(26 ...ll:if2+ 27 l':txf2 '1lr'xf2 28 ll:ie3 wins) 1997) 19 ... exf4 20 l':tael '1!r'cs 21 '1!r'c1
27 '1!r'gs ll:if2+ 28 �xf2 '1lr'xf2 29 ll:ie3 '1lr'd4 22 �4 �ae8 23 l':tffl '1lr'g7 with a
il.e8 30 f6+ 1-0 was M.lllescas Cordoba­ good position for Black in A.Shirov­
B.Gelfand, Linares 1994; 30 ...'1lr'xf6 31 J.Polgar, Madrid 1994.
ll:ifs+ wins. f2) 17 l':tael! is Kramnik's plan.
b) 16 ... gs is inflexible, but solid: 17 White intends ll:id1, '1!r'c1 and il.d3-c2
il.f2 il.xf2 18 �xf2 h4 (worse is 18.. .fS?! when all of his pieces will be aimed at
19 exfs .lii.xfs 20 il.d3! with a big ad­ the centre and kingside.
vantage in G.Vescovi-A.Bachmann, Sao
Paulo 2006) 19 l':tg1!? '1lr'e7 20 g3 '1¥o>h8
21 ll:if1 fs 22 ll:ie3 f4 23 ll:ifs '1lr'f6 24
gxh4 gxh4 25 l':tfg2 with just an edge
for White in Liang Chong-Ye Jiang­
chuan, HeiBei 2001.
c) 16 ... ll:ics 17 l':tabl fS 18 b4 axb4 19
axb4 ll:ia4 20 ll:ibs (Palliser suggests 20
ll:id1) 21 cxbs f4 22 ll:ic4 il.d4
with a strategically unclear position in
S.Savchenko-V.Grigoriev, Alushta 1994.
d) 16 ... '1¥o>g7 17 �ael '1!r'b8 18 il.d3 Black must take some care:
il.cs 19 ll:idb1 gs 20 il.f2 h4 21 ll:id1 f21) 17 ... '1lr'a7 18 ll:id1 il.cs 19 '1!r'c1
bS?! (this just weakens Black even fur­ l':tae8 20 il.d3 c6 21 ll:ib1 '1!r'b6 (better is
ther on the light-squares) 22 ll:ie3 '1¥o>h8 21 ...'1¥o>g7 22 ll:ibc3 with just an edge for
23 cxbs il.xbs 24 il.xbs '1lr'xbs 25 ll:ic3 White according to Kramnik) 22 il.c2
'1lr'b7 26 a4 �g8 27 ll:ibs ll:if6 28 ll:ic4 '1¥o>g7 23 ll:ibc3 '1lr'a7 24 il.g3 ! was
was much better for White in R.Palliser­ V.Kramnik-J.Nunn, German League
G.Jones, British Championship, Torquay 1994. White intends 25 f4! with the
2009. initiative.
e) 16 ... il.cs 17 il.f2 il.xf2 18 �xf2 fS f22) 11 ... il.cs 18 ll:idb1 '1lr'a7 19 ll:id1
(18...'1lr'e7 is safer) 19 exfs gxfs 20 l':tel c6 20 il.d3 cxds 21 exds (21 cxds l':tac8
gave White some initiative in R.Palliser­ looks okay for Black) 21 .. .fs 22 '1!r'd2
L.Trent, British League 2002. '1!r'b6 23 il.c2 (after 23 '1!r'h6 il.e8 24 il.c2
f) 16 ... '1!r'b8!? and now: '1lr'C7 Black is ready to evict the white
fl) 17 il.f2 seems to justify Black's queen with ...'1lr'g7) 23 ... '1¥o>g7 24 ll:if2
play: 17 ... '1lr'a7 18 il.xe3 '1!r'xe3 19 f4 il.d4 25 ll:id3 ll:ics and Black had good
(Black also reached a good position af­ play in T.Polak-L.Gofshtein, Vienna
ter 19 �ael '1!r'h6 20 ll:id1 h4 21 '1!r'b2 1996.

Petrosian Va riatio n : 7 dS as 8

17 exfs gxfs 2 0 '1¥o>h8 21 '1!r'b2 tt:\cs


Not 17 ... il..xfs? 18 il..d3 when Black's
bngside is starting to creak.

18 ii.fl
Or 18 �ael '1!r'g6 19 il..d3 and here:
a) 19 s 20 tt:ldb1 '1¥o>h8 21 il..f2
:gs 22 �gl �g7 and Black had coun­
terplay in R.Tuominen-M.Hartikainen,
Finland 1997.
b) 19 20 g4! ? '1¥o>h8 21 gxfS 22 b4 axb4 23 axb4 tt:\a4 24 tt:\xa4 z:txa4
25 il..d1 �as 26 f4

according to Komarov.
That may be a little bit extreme, but
Black has good play in any case.
18 il..xf2

After 18 6!?, as played in Y.Pelle­

tier-G.Mohr, Buekfuerdo 1995, switch­
ing back to the queenside with 19 �abl
has been suggested by Palliser. 26 exf4

19 z:txf2 tt:\f6 20 �gl! Opting for piece play, but 26 ... e4 27

Here we see another plan from tt:\f1 '1!r'g6 28 tt:\e3 '1¥o>h7 also looks okay
Kramnik. White again plays on the for Black.
kingside. 27 �e2

A ttacking Chess: Th e King 's I n di a n , Vo l u m e 1

27 l':txf4 �es ends White's initiative.

Black's position looks a little loose, but
there are not many pieces on the board
and he has enough activity to hold the

1 S....lii.e3+ 16 '1¥o>h1
16 .lii.f2 is also possible, though it is
less dangerous.

21 ...�g6
Black could also keep e7 covered by
playing 27 ...�f7. After 28 �d4 bS! 29
.lii . b3 l':tae8 30 l':txe8 l':txe8 31 �xf4 �e7
Black has counterplay. Moreover, 32
�h6+ gets nowhere after 32...'1¥o>g8 33
�g6+?! �g7 34 �xg7 'it>xg7 and Black
was even better in the endgame in
S.Cade-B. Benko, correspondence 2004.
28 l:te7 �f7 29 �xf7 �xf7 30 �d4 After 16 ... .lii.xf2+ (16 ... .lii.cs is sound,
Not 30 .lii.xh s �g7 31 .lii.f3 ll:ig4 e.g. 17 .lii.xcs ll:ixcs and now both 18
when Black has the initiative. �c2 �e7 19 b4 axb4 20 axb4 ll:ia4 and
30...�g7 31 �xf4 18 b4 axb4 19 axb4 ll:ia4, as in V.Anto­
This was V.Kramnik-B.Gelfand, shin-M.Chiburdanidze, Baku 1980, look
Linares 1994. Instead of 31 ... '1¥o>h 7?! fine for Black) 17 �xf2 Black has:
Black should have played 31 ... bS! with a) 11. . .ll:ics 18 �c2 �e7 19 b4 axb4
counterplay. 20 axb4 ll:ia4 21 cs (after the immedi­
ate 21 ll:ibs, Black can play 21 ... .lii.xbs 22
83) 15 �bl cxbs ll:ib6) 21 .. .fs?! (a lapse; 21 ...ll:ixc3
White gets on with his queenside 22 �xc3 fS was better) 22 ll:ibs ! and
play. This does not exclude White from Black's a4-knight was in trouble,
operation s on the kingside, however. F.Vallejo Pons-F.Jenni, Aviles 2000.

Petro s i a n Va ria tio n : 7 d5 a 5 8 il.g5

b) 17 ... 'fke7 is solid when the posi­ Black uses the bishop t o prevent b3-
tion resembles a Bogo-lndian. With b4 and hits the a3-pawn at the same
some pieces exchanged, Black is less time. Others:
worried about his a6-knight being side­ a) 16 ... tLics looks inconsistent with
lined, because White's rook will be tied the bishop on e3. After 17 'fkc2 (17 lLJbs
to the defence of the b4-pawn. After 18 'fkb8!? 18 b4 axb4 19 axb4 lLJa4 20 l':tb3
b4 axb4 19 axb4 practice has seen: is also possible) 17 ...fs 18 b4 axb4 19
bl) 19 ... h4 20 'fkcl (the more enter­ axb4 lLia4 20 lLibs il.xbs 21 cxbs f4 22
prising 20 cs!? dxcs 21 bS lLib4 22 lLic4 lLib3 'fkd7 23 tLias! gs 24 il.f2 il.xf2 2s
with compensation has been played a l':txf2 lLib6 26 lLixb7 White had won a
couple of times with success by Porper) pawn in A.Yusupov-P.Scheeren, Eind­
20... l':tfb8 (more solid is 20 ... cs with hoven 1986.
equality) 21 cs! dxcs 22 bS lLib4 23 b) 16 ... fS 17 exfs (after 17 b4 axb4
�c4, intending d6 or f4, gave White 18 axb4 f4 Black has good play on the
compensation for the pawn in F.Vallejo kingside) 17...gxfs 18 f4! exf4 19 il.xhs
Pons-0.Cvitan, European Team Cham­ 'fies is very unclear. Black has played
pionship, Plovdiv 2003. this several times with mixed results,
b2) 19 ...cs prevents White's sacrifi­ but I do not like Black's ragged pawn
cial idea: 20 dxc6 (20 bxcs tLixcs 21 structure and open king position.
-':Jb3 b6 was equal in 1.Khenkin­ 17 'fkc1
D.Bokan, Moscow 1989) 20...bxc6 21
-'!Ja4 (21 bs lLics 22 lLib3 lLixb3 23 .l:txb3
-':Jgs 24 il.fl lLie6 heading to d4 or cs
can only favour Black) 21 ...l':tfb8 22 cs
occurred in B.Zlotnik-A.Kuzmin, Buda­
pest 1989. Now 22 ... il.e6 is level.

11 ...fs
This is the most direct and looks sat­
isfactory. The position is very rich and
both sides may initiate play on either
side of the board. While the text move
is natural, there are other moves worth
16 s considering:

23 5
A tt a c k i n g C h e s s : Th e King 's I n dian, Vo l u m e 1

a) With 17 .. .f6!? Black finds a way to That leaves:

initiate play on the kingside without cl) 18 ll:ia4 il.d4 19 il.f2 il.xf2 20
opening things up too much: 18 ll:ia2 l':txf2 '1!r'd8! (again the position looks like
gs 19 il.e1 (19 il.f2) 19 ... g4 20 b4 axb4 a Boga-Indian) 21 dxc6 il.xc6 22 ll:ic3
21 axb4 il.e3 was unclear in l.Naumkin­ ll:ics 23 b4 axb4 24 axb4 ll:ie6 2s bS
V.Tsarev, Moscow 1988. il.d7 26 ll:ib3 '1!r'b6 27 l':tfl ll:if6 gave
b) 17 ...'1¥o>h8 is a normal waiting Black an edge in A.Yusupov-B. Daml­
move, but here it is slow and well met janovic, Saint John 1988.
by 18 ll:ia2!. Now: c2) 18 ll:ia2 is the best try: 18 ... cxds
b1) 18 ... gs 19 il.e1 (19 il.f2!?) 19 ... fs 19 b4! (Black gets counterplay after
20 b4 axb4 21 axb4 il.e3 22 exfs il.xfs both 19 cxds bS! and 19 exds fS!)
23 z:tb3 il.d4 24 ll:ic3 ll:if6 2s ll:ibs il.b6 19...axb4 20 axb4 il.e3 21 exdS! (better
26 il.f2 was agreed drawn in a fairly than 21 cxds when Black gets good
level position, l.Naumkin-T.Paehtz, Bu­ play with either 21 ... �c8 or 21 ... '1!r'b8)
dapest 1991. when Black's position looks quite good.
b2) 17 ...fs 19 b4 axb4 20 axb4 il.e3
21 ll:ic3 cs 22 dxc6 bxc6 23 '1!r'd1!? il.d4
24 ll:ia2 ll:ic7 2 s ll:ic1 ll:ie6 26 exfs gxfs
27 ll:idb3 looked better for White in
l.Naumkin-J.Fedorowicz, London 1990.
Black's position is rather loose.
c) 17...c6 increases the tension. Now
both 18 dxc6 bxc6 19 ll:ia4 il.d4 and 18
'1!r'b2 il.d4 19 b4 axb4 20 axb4 cs look
fine for Black, while 18 il.f2 il.xf2 19
l::txf2 gives Black a pleasant choice be­
tween 19 ...'1!r'd8, 19 ...'1lr'e7 and 19 ...'1!r'b8!?. However, actually Black must be
careful: 21 ... il.h6?! weakens cs and af­
ter 22 ll:ic3 fS?! 23 '1!r'c2 (23 cs! looks
even stronger) 23 ...l':tc8 24 l':tfel g s 2S
il.f2 g4 26 cs! White had seized the ini­
tiative by targeting the es-pawn in
l.Naumkin-Ye Jiangchuan, Belgrade
1988. A better try would have been
21 .. .fs!?.
18 exfs gxfs 19 il.f2
After 19 f4 e4 Black looks fine, while
19 ll:ia2 '1!r'g6 20 b4 axb4 21 axb4 il.e3

Petro s i a n Va ria t i o n : 7 d5 a 5 8 il.g5

22 il.f2 il.xf2 23 l':txf2 has been seen a

few times. Then the untried 23 ... lLif6!
looks flexible and best.
19 lLif6

Also satisfactory is 19 ... �g6 20 il.xc S

tLixcs (Black has also tried 20 ... dxcs?!,
but I could not do that to the a6-
knight) 21 b4 axb4 22 axb4 lLia4 with
play similar to the main line in
Z. Franco-C.Checa, Seville 1994.
20 il.xcs tLixcs
Compared to Kramnik-Gelfand in 21 b4 axb4 22 axb4 tLJa4
Line B2, Black is even better off because Black certainly had no problems in
White's pieces are passively placed. R.Pogorelov-Z.Plenkovic, Budva 2009.

Chapter 1 3
Excha nge Variation
7 dxes dxes 8 1fxd8 !txd8

1 d4 ll:if6 2 c4 g6 3 ll:ic3 1lg7 4 e4 d6 5 1. Beggars

ll:if3 0-0 6 Ji.el es 7 dxes dxes 8 '1!r'xd8 These players just want to try to make
l':txd8 a draw. Usually the player of the white
pieces is lower rated or perhaps the
tournament situation dictates that
White should try to draw. Unless a
quick draw suits the black player as
well, I absolutely agree with Gallagher
that these players should not be
granted a draw until every resource
has been exhausted. It is my experience
that m any beggars will crack at the
first sign of trouble. At the very least
we should make the game as miserable
The 'dreaded' Exchange Variation. an experience for them as possible.
This notorious drawing line has gone
so far as to discourage some players 2. Psychologists
from even playing 6 ... es. While the Ex­ Some players will play the Exchange
change Variation does have obvious Variation anticipating that Black will
drawing tendencies, Black should cer­ feel uncomfortable in the endgame.
tainly not give up hope of playing for a There is some logic to this; most King's
win. As Gallagher has mentioned be­ Indian players are striving for a sh arp
fore, this line is primarily played by a game. Nevertheless, if Black is well pre­
few different types of players. I will pared this approach will lose its sting.
classify these opponents as follows: It is important in general, of course, to

Exch a ng e Va r i a t i o n : 7 dxe5 dxe5 8 'jkxd8 'gxd8

be a well-rounded player capable of The text reaches the basic position

continuing to play even in equal end­ for the Exchange Variation. The pin
games. gives White a certain pressure against
both f6 and c?. Indeed, both 10 ll:ids
3. S pecialists and 10 hf6 followed by 11 ll:ids are
Some players just love to pl ay dry, threatened. However, if Black can neu­
technical chess. These players believe in tralize White's slight initiative, he can
White's chances in the Exchange Varia­ feel confident, because White's 9th
tion and are happy to try to prove that move left him with a weak d4-square,
he has a little something. These players whereas Black can still guard the dS­
are the most dangerous, because square with ... c6 at some point.
unlike beggars and psych ologists, they From a theoretical perspective,
are likely to be well prepared. Special­ Black really only needs to study Line A,
ists will have studied the detail s of the but I will examine three systems in de­
Exchange Variation and will be ready tail. One reason for this is variety -
to take advantage of lackadaisical play some players may face an opponent
by Black. that frequently uses the Exchange
9 1i. gs Variation, so it is nice to have more
It is well known that g ll:ixes can be than one weapon. Another reason is
met by 9 ...ll:ixe4 when Black is at least circumstantial - Black may want to
equal. White can also force things with either play solidly or take some risks,
9 ll:ids. After 9 ... ll:ixds 10 cxds c6 11 depending on the opponent and the
l.c4 cxds 12 il.xds ll:id7 13 'ge8 we situation.
have tran sposed to Line A2, below. If We will explore:
Black wants to avoid this, he can play
11 ... bs!? 12 il.b3 il.b7 13 and then A: 9 .�e8

either 13 ...'gd7 or 13 ... 'gc8. B: 9. ll:ibd7


C: 9... c6

Th e first of these lines, 9 ... 'ge8, is a

solid equalizer that was favoured by
Fischer. This is the line recommended
by Gallagher as well. White has great
difficulties obtaining even a shred of
an advantage against 9 ... l:te8, but be­
cause a great deal of simplification of­
ten occurs, it can also be difficult for
Black to try to win. Line B is more com-

A ttacking C h e s s : Th e King 's In dian, Vo l u m e 1

bative, but White also has greater a2) 14....rf.e8 1 5 o-o-o (15 �f3 f6 16
chances of getting an edge. Neverthe­ tll d 3 �fs 17 tllf2 tll cs is level) 1s ...f6 16
less, White's advantage is not large and tlld 7! �xd7 17 .rtxd7 .rtac8 18 �f3 tllcs
this is a good choice as Black for those 19 l:tddl! with the idea of 19 ...tllx e4 20
that want to try to gradually outplay l:thel.
their opponents. Line C is a gambit con­ a3) 14...tllcs 15 o-o-o! f6 16 l:td8+
tinuation. This option is quite popular (16 tllg4 tllxe4 17 �f3 is a little better
and it is in decent theoretical shape. for White) 16 ... �g 7 and now White has
9 ... c6 is the most combative option, but several ways to get an advantage, in­
Bl ack does sacrifice a pawn, so he does cluding 17 tll d 3 (also good are 17 b4
risk more than in Lines A and B. tll x e4 18 �g4 and 17 tllf3 l:txe4 18
There are two other moves. tlld4) 17 ...l:txe4 18 tllxcs l:txe2 19 tlld3
a) 9 ... tlla6 is probably not very good. when he has the upper hand.
It is worth examining though, if only to b) 9 ... l:tf8! ? is an idea of Lanka that
compare it to a line we will see later in has been pioneered by Bologan and
the Makogonov Variation (in Volume Shirov.
2). Play continues 10 tllds (10 tllxes
l:te8) 10 ... l:td6 11 �xf6 �xf6 12 tllxf6+
l:txf6 13 tllx es l:te6 14 f4 when Black
will have trouble getting his pawn back
without making any concessions.

White has:
bl) 10 tll d s tllxds 11 cxds c6 12 �c4
bS 13 �b3 �b7 14 l:tcl as 15 a3 (after
15 a4 bxa4 16 �xa4 cxdS! the main
point of Black's play becomes apparent
The unsatisfactory choice is: - if we compare this to the note to
al) 14.. .f6 15 tll g 4 hs 16 tllf2 tllcs 17 Black's 12th move in Line A2, we can
o-o (or 17 �f3 tllxe4 18 �xe4 fS 19 �d2 appreciate that Black's rook is not
fxe4 20 �e3 with a nice advantage) pinned) 1S ... a4 16 �a2 l:tc8 17 �e3 b4
17 ...tllx e4 18 tllx e4 l:txe4 19 �d3 and 18 axb4 a3! gives Black good play as
White is clearly better. shown in several games.

Exc h a n g e Varia tio n : 7 dxes dxes 8 'ilixd8 1:l.xd8

b2) 10 �xf6 �xf6 11 liJds .�d8 12 This is the classical continuation.

liJxes leads nowhere after 12 ...1:l.e8 13 Black simply protects his es-pawn
l:.d1 c6 14 liJC3 �as 1s lLlf3 l:txe4. while breaking the pin on his king's
b3) 10 o-o-o liJc6 (10 ... liJbd7 is Line knight. Now 10 �xf6?! .�xf6 11 liJds
B) 11 h3 �e6 12 �e3 .rf.ad8 13 a3 h6 can be met with 11 ... �d8! when Black
(13 ... liJas ! ? 14 liJd2 liJc6 could also be will play ...c6 and have everything he
tried) 14 b4 as lS bS liJd4! 16 liJxd4 wants. 9 ... l:te8 has always been consid­
exd4 17 �xd4 l:txd4 18 l:txd4 liJg4 19 ered a solid equalizer, but correspond­
�xg4 �xd4 20 �xe6 fxe6 21 liJd1 �cs ingly it has a drawish reputation. Even
22 a4 l:td8 with some compensation for so, I think that with careful study, Bl ack
the pawn in Wang Yue-V. Bologan, will be well prepared to seize the initia­
Moscow 2006. tive if White gets careless.
b4) 10 liJxes is critical according to There are two main lines, with the
Bologan. After 10 ...liJxe4 11 liJxe4 �xes second being much more popular.
12 o-o-o liJc6 13 f4 �d4 (or 13 ... �g7 14
.!l:if6+ �g7 1s h4!) 14 liJf6+ �g7 1s lLlds A1: '10 0-0-0
f6 16 �h4 �fs, instead of the 17 �d3 A2: 10 ll'id5
of R.Ravisekhar-V.Bologan, Calcutta
1992, Bologan gives 17 l:thel when Al) 10 0-0-0
White's good centralization gives him
slightly the better prospects.
We can conclude that 9 ... liJa6 is just
dubious, but 9 ... l:tf8 is playable if Black
is unhappy with the three (!) recom­
mendations given here.

A) 9 l:te8

White just continues his develop­

ment, hoping th at his space advantage
will mean more than Black's long-term
control of d4. Generally some pieces
stay on the board in this variation and
there is scope for independent play.
10 liJa6

This is a natural way for Black to de-

A ttacking C hess: Th e King 's I n d i a n , Vo l u m e 1

velop. The knight covers the c7-square 12 tllc2

and may go to cs, from where it pres­ Instead 12 f3 tllcs 13 tllc2 leads to
sures the e4-pawn and can continue to the same thing, but White could con­
e6, with an eye towards the weakened sider controlling cs with 12 tlld3 ! ? here.
d4-square. Instead 10... h6 is also possi­ For this reason 11 ... tll cs might be pre­
ble, but it is not really necessary for ferred.
Black to chase away the white bishop, 12 ... tllcs 13 f3
so I prefer to keep this move in reserve. This is a typical position. White can
11 tlle1 double rooks on the d-file and try to
Instead 11 tllx es should be met by expand on the queenside. Black will
11 ...tllc s ! (11 .. .l:txes 12 .l:td8+ tlle8 13 f4 often play moves like ... tlle 6, ...�f8,
l:te6 14 �g4 l:tb8 lS es is probably ...tllfd7, and perhaps ... as and ... b6. Ide-
somewhat better for White, but here ally Black would like to get in ...�cs too
12 ... l:te8 13 �xf6 l:txd8 14 �xd8 �h3 is to activate his dark-squared bishop.
probably okay) 12 �xf6 �xf6 13 f4 c6
when Black wins back the pawn, while
11 tlld2 c6 12 tllb3 (12 f3 tllcs 13 �e3
tlle6 is also fine for Black) 12 ... tlle7 13 f3
tlle6 14 �e3 �f8 lS a3 b6 16 .l:!.d2 �a6
17 �bl was agreed drawn in
V.Akobian-A.Yermolinsky, Agoura Hills

13 ...tlle6
By avoiding ... h6 Black is able to play
this move with tempo. Black could also
maintain the knight on cs for the time
being and play 13 ... as 14 �e3 (or 14 b3
�e6 with the idea of ... tllfd7) 14 ...�f8,
which is level.
14 �e3 tllf4
11 ... c6 Black heads in a different direction
Black has also tried both 11 ...�e6 and probes the kingside. He has also
and 11 ... h6, while another move order tried 14.. .il.f8 with the idea of ... �cs. If
for Black is 11 ... tllcs 12 f3 tlle6 (or the lS b4 as 16 a3 axb4 17 axb4 he could
immediate 12 ...c6) 13 �e3 c6. try 17 ... cs or 17 ... tllf4. However, I would

Exch a n g e Varia tio n : 7 dxes dxes 8 'ilixd8 '!J.xd8

prefer the prophylactic! ? when exds e4 14 tlld2 Black can safely play
15 tlla4 tlld7 intending ...�f8 looks fine 14...�xb2, while 13 '!J.xds �e6 13 �bs
for Black. �d7 14 l:txds �xbs 15 '!J.xbs b6 16 �e2
15 �fl hs 16 h4 �f8 17 b4 tlle6 18 a3 tll a6 is fine for him too.
b6 19 �b2 �b7 d) The most interesting try is the
Chances were about equal here in gambit continuation 12 o-o!?.
Lesiege- Smirin, Biel lnterzonal 1993.

A2) 10 tlld s
This forcing move is more popular.
It is not necessarily better, but it forces

simplification and generally leads to a

symmetrical pawn structure.

Black has:
dl) 12 ...cxds 13 exds e4 (after
13 ...�g4 14 h3 �xf3 15 gxf3 the
bishop-pair was more important than
White's fractured pawns in J.Jansson­
G.Gnichtel, European Club Cup, Fuegen
2006, but 13 ...tlld7 14 l:tfdl, as in
10... tll x ds 11 cxd s c6 J.Jansson-H.Harestad, Oslo 2006, and
If Black refrains from this move, his then 14 ... e4! ? is an idea) 14 tlld2 l:tes
backward c-pawn will become weak. (after 14...�xb2 15 '!J.abl �g7 16 '!J.fc1
12 �C4 tlld7 17 d6 tllb6 White has obvious
White wants to bring his relatively compensation for the pawn) 15 �e3
passive bishop to the nice dS-square. '!J.xds 16 tllxe4 tllc6 17 �c4 '!J.d8 18
Other moves are also possible: '!J.adl �fS 19 tlld 6 gave White the ini­
a) 12 dxc6 tllxc6 just helps Black de­ tiative in J.Jansson-Thomassen, Norwe­
velop. gian League 2008.
b) 12 d6?! f6 13 �e3 �e6 leaves the d2) 12 ... h6 13 �e3 cxds 14 exds fs
d6-pawn weak. Black will play ... '!J.d8 15 '!J.fd1 tlld7 16 .l:!.acl f4 17 �cs tllxcs?!
and ...�f8 to round it up. 18 '!J.xcs �g4 19 '!J.a was better for
c) 12 '!J.dl is an inferior way to sup­ White in J.Jansson-K.Tryg stad, Euro­
port the dS-pawn. After 12 ... cxds 13 pean Club Cup, Fuegen 2006, and 19

A ttacking C h e s s : The King 's I n d i a n , Vo l u m e 1

d6 ! looks even stronger. Instead Bolo­ h i s main idea without any hindrance.
ga n gives 17 ... e4 1s tll d4 tllxcs 19 l:txcs White can also try 14 l:tcl, which is
<Ms with equality. not so good, although Black has to be
12 ... cxds precise: 14 ... h6! (14 ... tllf6? lS �xf6
We will stick with this classical con­ �xf6 16 l:to is much better for White
tinuation. Black can also play 12 13 and 14 ...tllb6 lS �b3 �e6 16 �xe6
�b3 �b7 14 lkl as lS a4 bxa4?! 16 l:txe6 17 l:to is not want Black wants
�xa4, but here, compared to the 9 .. J:US either) lS �e3 (lS �h4 g s 16 �g3 tllf6
line, Black has problems because his c6- is also good for Black) 1s ...tllf6 16 �b3
pawn is pinned to the eS-rook. After tllxe4 17 l:te7 �e6 lS �xe6 l:txe6 19
16 ...l:tcs 17 d6 f6 1s �b3+ <Ms 19 �e3 l:txb7 l:ta6 20 a3 tlld6! when Black's
tlld 7 20 h4 cs 21 �e6 l:tdS 22 tlld2 better development and active pieces
White was much better in F.Vallejo give him an edge.
Pons-A.Morozevich, Monaco (blindfold) 14...tllc s
13 �xds tlld 71
Black has also played 13 ...tlla6 and
13 ...tllc6, but White can still hope for an
advantage in those lines.

15 0-0-0
other moves do not trouble Black at
a) lS o-o �e6 16 �xe6 tll xe6
(16 ...l:txe6 is also good and has been
With the text, Black intends ...tllf6 or played with success by Hebden) 17 �e3
... h6 followed by ... tllf6 when the pres­ tlld4 and Black has the initiative. In­
sure on the e4-pawn will force White to deed, White has trouble stabilizing the
part with one of his bishops. position: for example, lS �xd4 exd4 19
14 tlld2 l:tacl l:tacS 20 f4 d3! 21 es f6 and
After 14 �e2 h6 lS �e3 tllf6 or 14 Black's bishop is very strong.
o-o h6 1 s �h4 g s 16 �g3 tllf6 14 o-o-o b) lS �e3 tll d 3+ 16 �e2 tllf4+ 17
h6 lS �e3 tllf6 Black is able to execute �xf4 exf4 lS l:tacl �xb2 19 l:tc7 �e6!

Exc h a n g e Varia tio n : 7 dx es dxes 8 'ilixd8 '!J.xd8

20 �xe6 '!J.xe6 21 '!J.xb7 (21 '!J.b1 �es 22 d3) 16 '!J.d1 has been analysed by
'!J.cxb7 '!J.a6 is good too for Black) Grivas: 16 ... �e6 17 f3 '!J.ac8 18 b3 (18
21 ...�c3 22 liJb1 '!J.xe4+ 23 'itf3 '!J.b4 24 liJe3 liJa4! gives Black counterplay as he
.rtxb4 �xb4 2S '! '!J.d8 26 �xf4 �d2+ points out) 18 ...�xds (or 18 19 lLle3
27 liJxd2 '!J.xd2 and Black won a pawn �g7, whereas Grivas only quotes a
in J.Vanheste-J.Gallagher, Metz 1991. game with the stran ge 19 ... �d7) 19
c) lS �e2 �e6 (it may be better to .l:!.xds bS 20 liJd6 (20 liJe3? liJe6 is good
just play 1S ...liJe6 16 �e3 liJf4+ which for Black) 20 ... �xd6 21 l:.xd6 liJe6 22
is variation 'b') 16 '!J.hcl '!J.ec8 17 �e3 �e3 '!J.c2 23 '!J.d2 and now, rather than
�xds 18 exds liJd7 19 liJe4 �f8 20 �gs Grivas's 23 ... '!J.ec8 24 �e2 when he
fS 21 liJf6+ liJxf6 was equal in claims an edge, it looks to me like Black
M.Rutkowski-B.Socko, Warsaw 2010. can play the simple 23 ... '! 24 '!J.d1
d) 1S lLlc4 �f8 and here: '!J.c2.
Returning to lS 0-0-0:

dl) 16 o-o-o �e6 17 �bl '!J.ac8 put s

pressure on c4 and e4. 1s liJe6

d2) 16 o-o �e6 17 �xe6 (17 '!J.fdl is Black must avoid the greedy
probably a better chance to maintain 1S ... liJd3+ 16 �bl liJxf2? because 17
equality, intending 17 ...liJxe4 18 �xe4 '!J.dfl wins for White. However, 1S ...�e6
�xc4 19 �xb7 '!J.ab8 20 �c6) is a good alternative. After 16 �xe6
17 ... '!J.xe6!. At first this looks equal, but liJxe6 (16 ... '!J.xe6 is similar to Acebal­
Black actually has some initiative. He Gallagher above, but in this case
can kick White's knight with and White's king is more active, which al­
his own pieces will all become quite lows him to more easily defend and to
active. Indeed, after 18 f3 bS 19 lLle3 h6 attack on the queenside: for example,
20 �h4 liJd3 21 liJds '!J.c8 Black was 17 �bl �f8 18 �e3 '!J.c8 19 '!J.c1 '!J.ec6
doing well in A.Acebal-J.Gallagher, Can­ 20 f3 and Black has to be careful about
das 1992. any simplification because White's

A ttacking C he s s : The King 's I n d i a n , Vol u m e 1

king can easily become active on the l:tgfl �h6 2 3 �bl �xe3 24 fxe3 l:txfl
queenside) 17 �e3 there is: 2S l:txfl l:tc8 was a little better for
a) 17 ...�f8 is solid enough when Black, albeit still drawish in
White has: Ki.Georgiev-E.Sutovsky, Gibraltar 2006.
al) 18 tll b 3 as 19 l:tds (Black is cer­ 16 �e3 tllf4 17 �xf4 exf4
tainly okay after 19 �bl a4 20 tllcl �cs
21 �xcs tll x cs) 19 ...a4 20 tll d 2 .l:tac8+
(both 20 ... tllf4 and 20... f6 are sufficient
as well) 21 �bl �cs 22 l:txes �xe3 23
fxe3 tll c s is equal.
a2) 18 �bl �cs 19 �xcs (or 19 tllc4
�xe3 20 tllxe3 �8 21 tllg4 tllcs 22 f3
h s 23 tllf6 .l:!.e6 24 tlld s .l:!.d6 2s l:tc1
V2-V2 B.lvkov-Z.Polgar, Wijk aan Zee
1986) 19 ... tllxcs 20 f3 tlld3 21 .l:thfl
l:tac8 22 tllc4 .l:txc4 23 .l:txd3 V2-V2
A.Onischuk-V.Bologan, Poikovsky 2009. Black has not had any problems
b) 17 ...tllf4 is more enterprising: 18 from this position.
g3 (18 �xf4 exf4 19 f3 .l:tac8+ 20 �bl fS 18 f3
gives Black enough counterplay) Instead 18 tllc4 �e6 19 tlld6 .l:te7
18...tllg 2! does not get White anywhere. other
tries are:
a) 18 tllb3 �e6 19 �xb7 (better is 19
f3, transposing to the main line)
19 ....l:tab8 20 �ds �xds (20... as! gives
Black good play) 21 l:txds (or 21 exds
f3 ! 22 gxf3 l:te2) 21 ... l:txe4 is level.
b) 18 �bl �e6 19 �xe6 .l:txe6 20 f3
fS ! 21 exfs (after 21 l:tcl .l:td8 22 l:tc2
Black can play 22 ... �d4 23 exfs gxfs or
22 ... .l:ted6 23 tllc4 l:tdl+ 24 l:txdl l:txdl+
2S .!:tel l:txcl+ 26 �xcl fxe4 27 fxe4 bS)
19 tllf1 (19 �gs h6) 19....l:te6 20 .:tg1 21 ....:te2 (21 ... gxfs looks okay too - the
(otherwise, 20 �d2? l:tf6! 20 l:td7 l:tf6 doubled pawns control some impor­
gives Black good play, while 20 �bl tant squares) 22 fxg6 hxg6 23 g3 (23
l:tf6 21 .l:td2 l:tf3 22 l:te2 tllxe3 23 tllxe3 .l:thgl l:td8) 23 ... bS! with good compen­
l:td8 24 .!:tel �h6 with equality is given sation.
by Bologan) 20 ... tllx e3 21 tllxe3 l:tf6 22 18 ...�e6 19 tllb 3

Exc h a n g e Va ria t i o n : 7 dxes dxes 8 'ilixd8 '!J.xd8

19 �xe6 '!J.xe6 is level as well. �7 28 tllc4 .rf.e6 and 25 ...�7 26 .:tel

19... �xds 20 '!J.xd s '!J.e8 27 '!J.C7+ .l:te7 28 '!J.xe7+ �xe7
should draw.
26 '!J.e2
The idea is that after 26 .l:te7 �e3 27
tllb3 .l:tc6! 28 .rtxb7 '!J.g6 Black has coun­
26 ... �e3 2 7 tllb3 b6 28 '!J.c2 '!J.d8
Black has good counterplay in this

B) 9...tll b d7

20 ...fsl
Thematic. Black must open lines for
his rooks.
21 exfs l:tac8+
21 ....rf.e2 22 .rtd2 '!J.xd2 23 tllxd2 gxf5
is also good enough.
22 �bl '!J.e2 23 '!J.d2
Pia Cramling has preferred to make
draws after 23 f6 Lf6 24 '!J.d2 '!J.ce8
{1/2-1/2 P.Cramling-J.Gallagher, Biel
1991) or here 24 ... '!J.e7 M-V2 This combative move aims to keep
P.Cramling-E. Grivas, European Team pieces on the board without giving up
Championship, Debrecen 1992). material. It has been played by Rad­
23...l:txd2 24 tllxd2 gxfs 25 '!J.e1 jabov, Nakamura and a young Kas­
Or 25 .!:tel '!J.e8 26 .rf.C7 .rf.e2, while 25 parov. If White does not act quickly,
.!Llb3 .rf.e8 26 tllcl l:td8 27 �c2 '!J.c8+ 28 Black will consolidate his position and
it>bl '!J.d8 29 �c2 '!J.c8+ 30 �bl and aim to exploit his dark-square control,
1h-V2 was J.Nogueiras-D.Anagnosto­ so White must play purposefully.
poulos, Linares 1996. 10 0-0-0
I know all of this seems rather bor­ The most obvious and best. Others:
ing, but Black should know how to a) 10 tlld5 is premature: 10 ... c6 11
generate enough activity to hold the tlle 7+ �8 12 tll x c8 '!J.dxc8 is very com­
balance. fortable for Black. White has spent sev­
2s ...�d41? eral tempi exchanging off the c8-bishop
Instead 25 ...'!J.c6 26 .l:te7 '!J.g6 27 '!J.e2 and Black's pieces coordinate well:

24 7
A ttacking C he s s : Th e Kin g 's In dian, Vo l u m e 1

al) After 13 tlld2 tllcs 14 f3 tlle6 nected, which should favour Black
(14 ... as) l S �e3 tlld7 16 o-o-o �e8 17 somewhat) 13 ... tllcs 14 tllxc8 l:tfxc8 l S
g3 �f8 18 tllb3 as Black was quite com­ f3 as 16 tllb3 tlle6 17 �e3 �f8 Black
fortable in N.Brunner-M.Kazhgaleyev, was fine in V.Perekhodkin-E.Pakhomov,
Vandoeuvre 200S. Dagomys 2010.
a2) 13 o-o-o tll c s 14 �xf6 �xf6 lS 10 l:.f8

�d3 as 16 l:.hel l:te8 17 �fl �d8! was Again 10 ...l:te8? fails to 11 tllb s!.
S.Danailov-G.Kasparov, Dortmund
1980. Despite the opposite-coloured
bishops, Black is much better because
of his grip on the dark squares.
b) 10 .l:tdl l:tf8 (an important tactical
necessity because 10 ... l:te8? loses to 11
tllb s!) and here:
bl) 11 0-0 c6 is fine for Black, who
will play ... l:te8, ... as and ... �f8. The play
is similar to that in Line Al, except that
here White's king is not able to support
his queenside play. Now White has two main tries:
b2) 11 tllb s is a shot in the dark:
11...c6 (or 11...tllxe4!? 12 tllxe7 tllx gs 13 .. 8i1;J.1i)ct2
tllxa8 tlle 6! with compensation in 1b: 11..�lcfs
C.Gorobinskiy-M.Golubev, Ukraine
1999) 12 tlld6 tllcs 13 tllxc8 (13 �xf6 White has also tried 11 tll e l, but
�xf6 was quite comfortable for Black this does little to challenge Black's set­
in another C.Gorobinskiy-M.Golubev, up. After 11 ...tllcs (or 11 ...c6) 12 f3 tlle6
Ukraine 1999, clash) 13 ...l:taxc8 14 13 �e3 cs it's equal.
�xf6 �xf6 was better for Black in Another idea is 11 b4 as 12 a3 axb4
R.Asylguzhin-A.Kuzmin, Dubai 2003. 13 axb4 h6! (Golubev points out that
The play is similar to Danailov­ 13 ... l:ta3 14 �b2 l:txc3 fails to l S tlld2!)
Kasparov above. 14 �h4 (both 14 �e3 tll g 4 and 14 �d2
b3) 11 tlld s should be compared to tllb6! l S cs tlla4 are pleasant for Black)
Line B2, below. After 11 ... c6 (not 14...l:ta3 ! ? (also tempting are 14 ... cs
11 ... tllxe4 12 �e3 !) 12 tlle 7+ �h8 13 and 14...tllh s) lS �b2 l:txc3 16 �xc3
tlld2 (or 13 �e3 l:te8 - Black could also (not 16 tlld2 gs ! 17 �g3 l:txg3)
try 13 ...tllb6 or 13 ... b6 - 14 tllxc8 l:taxc8 16 ...tllx e4+, as given by Golubev, which
when the position is similar to Line B2 gives Black decent compensation for
except that White's rooks are not con- the exchange.

Exc h a ng e Va riati o n : 7 dxes dxes 8 'ilixd8 '!J.xd8

91.J 11 tll d2 the idea of ... tll e6, while after 13 h4 I

would suggest 13 ... tllf8!? heading for
the e6-square.
12 ...'!J.e8
It is worse to play 12 ... as 13 a3 l:te8.
Indeed, 14 �b2 tllf8 (14 ... b6 lS bS �b7
may be a better try) lS cs tlle6 16 �xf6
.�xf6 17 tllc4 axb4 18 axb4 tlld4 19
tll d 6 '!J.f8 20 �c4 '!J.b8 21 tlla 4 �e7?
(better is 21 ...�d8, but Black is
cramped in any case) 22 tllb6 was good
for White in T.Engqvist-T.Ernst, Han­
This has been played rather often, inge 1997. Here Golubev has pointed
but it does not seem dangerous to me out the resource 22 ...�xd6 23 cxd6 '!J.d8
if Black takes care. 24 f4! when Black is in big trouble.
11. •.c6 13 �b2
Black intends ... tllcs. The immediate Instead 13 cs looks premature after
11...tllc s? was dealt a serious blow by 13 ... as 14 a3 axb4 l S axb4 b6!, M.Haag­
12 tll d s! (12 �e3 tll e 6 13 tll d s c6 leads A.Kochyev, Balatonbereny 1988, while
nowhere) 12 ... tllcxe4 13 tllxe4 tllxe4 14 Black has counterplay after 13 tllb3
ie3 ! when Black has great difficulties: �f8 14 a3 as (14 ...b6 and 14...h6 also
for example, 14... c6 (or 14...b6 lS �f3 look good) lS cs (not lS tllx as cs!)
fS 16 tllxe7 '!J.b8 17 �xe4 fxe4, as in 1s...axb4 16 axb4 and now either
V.Kartachov-V.Evelev, Moscow 1998, 16 ... h6 or 16 ... �g 7, in both cases plan-
and now 18 tllbs would give White the ning ...b6.
upper hand) lS tlle 7+ �h8 16 tllxc8 u...tllf 81? 14 cs tll e 6 15 �xf6 �xf6 16
:fxc8 17 '!J.d7 and White won back the tllc4 tlld4 17 tlld6 l:.f8
pawn with a big advantage in L.S chan­
dorff-M.Golubev, German League 2002.
12 b4
White seizes space while he can. After
12 �e3 '!J.e8 (12 ... l:td8 !?) 13 cs �f8 14
b4 (14 tllb3 b6 14 tlla4 bS lS cxb6 axb6
16 tllxb6 tllxb6 17 �xb6 '!J.xa2 is also
pleasant for Black) 14... as lS a3 axb4
16 axb4 b6 Black seizes the initiative.
Black is fine too here after 13 f3 �f8 14
tllb3 tllh s ! 1s g3 tllg 7 (or 1s ... b6) with

A ttacking C h e s s : The King 's I n dian, Vo l u m e 1

Black intends ... b6 (Golubev) with tllxe4 14 tllxc8 allows the nice trick
counterplay - this is the reason for 14...tlld cS! when White cannot hold on
leaving out ... as. Already I prefer Black. to everything.
e2) 11 tlld s Another idea is 13 ... b6!?. Then 14
This is critical. tllxc8 .rtaxc8 lS g4 tllcs gives Black
11 ...c6 counterplay, while 14 l:td6 ..lri.b7 lS tllg s
Panczyk and llczuk suggest h6 1 6 l:thd1 i s unclear. Instead 14 tllxc6
11 ...tllx ds 12 cxds fs, but after 13 exfs! tllxe4 is untried and may be critical,
I think White is better prepared for the but the thematic move is 14 g4. Now
opening of the position. 14 ... �b7!? (14 ... tll cs 1s tllx es lllfxe4 16
Capturing the pawn with 11 ...tllxe4 tll sxc6 looks insufficient) lS gs l:tae8
runs into a familiar theme: 12 �e3! 16 gxf6 (16 lllfs gxfs 17 exfs ..Iii.cs 18
(instead 12 tlle 7+ �h8 13 tllxc8 tlld cS! gxf6 tllxf6 is murky) 16 ...tllxf6 17 tlld s
14 tlle 7 tllxg s lS tllxg s �f6 is certainly cxds 18 exds tllg4 gives Black counter­
fine for Black and 12 tllxe7 tllxg s 13 play.
tllxa8 tll e 6 is similar to one of the 14 tllxc8 l:tax c8
Gorobin skiy-Golubev games in the Black can also play 14 ... l:texc8 as lS
notes to White's 10th, above) 12 ... tllb6 tllg s �g8 16 ..lri.g4 .rf.d8 1 7 �xd7 tllxd7 is
(not 12 ... tlldcS?! 13 tlld2 ! when Black fi n e for him. Instead lS g4 looks more
has problems - Golubev) 13 tllxe7 l:tb8, dangerous, as below.
which leaves White with several good
continuations, such as 14 tlld 2.
12 tlle7+ �h8

After the text, White has the bishop­

pair, but Black has good squares for all
his pieces. White must try to seize the
13 ..lri.e 3 initiative or Black will have good
White prevents ...tll cs. After 13 chances to take over.
tllxc8 l:taxc8 Black is fine, while 13 l:td6 15 g41

Exch a n g e Va ria t i o n : 7 dxes dxe5 8 'ilixd8 '!J.xd8

The most dangerous try. White ex­ ther compromise his pawn structure or
ploits the shaky knight on d7 to disrupt put him self into a pin along the d-file.
Black's coordination. Advancing the g­ 19 �g4 �f8
pawn also gains space and gives the Now Black is ready to free himself
e2-bishop some scope. Other moves are with ... tlldcs.
less impressive: 2o �xe6 'fJ.. x e6 21 'fJ..d 3
a) lS tll g s �g8 (Black has also Black has experienced some diffi­
played 1s ... l:te7 16 f3 and then either culties from this position.
the flexible 16 ...�f8 or the rigid 16 ... cs) 21 ...tllb6
16 �g4 'fJ..e 7 17 �xd7 tllxd7 is fine, as in Other moves are possible as well:
the note to Black's 14th move, above. a) 21 ... cs 22 tlld2?! fS! 23 f3 f4 24
b) 1s cs �f8 16 b4 (16 tllgs �g7 11 ..lif2 h6 2s gxh6 gs 26 tll b 1 'fJ..x h6 gave
�c4 �xcs! is fine too) 16...b6 17 tll g s Black counterplay in T.Roussel Rooz­
�g8 18 �c4 'fJ.. e 7 19 cxb6 axb6 20 f3 bS mon-H.Nakamura, Philadelphia 2006,
21 ..lib3 h6 22 tllh 3 'fJ.. e e8 (22 ... cs!) 23 but as Mikhalevski points out, 22 l:thdl!
tllf2 �xb4 24 �xh6 cs 2s .lids c4 (or tllb6 23 tll a s is very good for White.
2s ... �c3 with the idea of ... �d4) was b) 21 ...l:.d6 is probably the safest.
very comfortable for Black in
S.Cvetkovic-G.Kuzmin, Linz 1990.
c) 1s tll d 2 �f8 16 tll b 3 b6 (16 ... as is
interesting, but then Black probably
should have recaptured on c8 with the
other rook) 17 g4! tllb8 (17 ... 'fJ.. ed8 looks
better) 18 �f3 cs 19 gs tllg8 20 �g4
'fJ..e7 21 h4 hS 22 �h 3 and White had
the initiative in N.Ostojic-M.Markovic,
Cetinje 1992.
Black cannot, of course, capture ei­ White has:
ther pawn. bl) 22 'fJ..h dl 'fJ.. x d3 23 'fJ..x d3 �e7 24
16 tlld 2 f4 b6 (24 ... �g7 2S �c2 f6 is a better try
White protects both e4 and g4 with to hold according to Mikhalevski) 2S fS
this move. f6 26 h4 and White maintained an
16 ...tlle 6 11 gs tlld 7 18 tllb 3 edge because of his space advantage in
18 �g4 lk7 19 tllb3 transposes. l.Piven-1. Drozdov, Budapest 1994.
18 ...'fJ..c 7 b2) 23 'fJ..d l �e7 24 h4 (Mikhalevski
After 18 ...'fJ..cd8 19 �g4 White will suggests 24 f4) 24 ... tllb6 2S tlld2 tll d7
take on e6, when Black will have to ei- (2 S ... �g7 with the idea of ...f6 looks

A ttacking C h e s s : The King 's I n dian, Vo l u m e 1

better) 26 tllf3 and White kept an edge pawn and then defend. Nowadays, this
in A.Mirzoev-M.Erdogdu, Istanbul 2010. line does not surprise anyone, but it
22 tllc sl l:td6 remains a viable choice for those will­
This walks into some clever tactics. ing to takes some risks.
After 22 ...�xcs 23 �xcs White's bishop 10 tll xes
is strong and Black cannot play White really has to take the pawn,
23 ...tllxc4? because 24 l:td8+ �g7 2S otherwise there is no chance to fight
�f8+ mates, but 23 ...h6!?, with the idea for any advantage at all. After 10 o-o,
of 24 gxh6 tllxc4, is probably the best 10 ...l:te8 is safer than 10 ... tlla6 11 tllxes
try. l:te8 12 f4, but Black could also try
23 l:txd6 �xd6 24 l:td1l 10...h6, 10 ...t2Jbd7 OT 10 ... �g4.
24 b3 was enough for an edge, but 10... l:te8
this is even stronger. Not 10 ... tll a 6? 11 tllxf7 (11 tllxc6
24 ...tll xc4 25 b3 tll x e3 26 l:.xd6 tllg4 may be even better) 11 ...�xf7 12 es
This was Z.Gyimesi-1. S mirin, Maalot­ when Black will remain down a pawn
Tarshiha 2008. Now the cleanest line is for nothing.
27 l:td2! tll xh2 (27 ...b6 28 tll a 6! followed
by 29 h3 traps the knight) 28 .l:td8+ �g7
29 tlld7!, winning the exchange.

C) 9 ... c6

11 0-0-0
White tactically defends the eS­
knight and accelerates his develop­
ment. This certainly seems to be the
most logical move, but he has tried
This enterprising gambit became other things too:
very popular in the late 1980s. Black a) 11 f4 tllh s ! 12 �xhs gxhs (Black
makes the move he wants to make and threatens 13 .. .f6, winning a piece) 13
leaves the es-pawn to its fate. At first �h4 tlld 7! 14 �g3 tll xes 1s fxes �xes
9 ... c6 scored tremendously - White 16 �xes l:txes was level in E.Maljutin-
players did not seem ready to grab a 1.Glek, Moscow 1989.

Exc ha n g e Va ria tion: 7 dxes dxes 8 'ilixd8 '!J.xd8

b) 11 �f4 tll a 6 (11...tllxe4 12 tllxe4 Grabbing the pawn is also poor:

�xes 13 �xes .l:txes 14 l:td1! is better 11 ...tll x e4 12 tllx e4 �xes 13 '!J.hel (not
for White) 12 o-o-o tllcs and Black will 13 tllf6+? �xf6 14 �xf6 tlld 7!) and
win back the pawn: for example, 13 f3 White has the initiative. However,
(or 13 .l:thel tllfxe4, while 13 �f3 runs 11 ...h6!? 12 �f4 tlla6 is an alternative.
into 13 ... g S !) 13 ... tllh s 14 �e3 tllxe4 lS
tllxe4 (after lS fxe4 �xes 16 �xhs
gxhs 17 �d4 �g4 18 �xes, as in
T.Johansen-S.Gabrielsen, Stavanger
1997, and then 18 ... �xdl 19 �f6 �g4
20 es Panczyk and llczuk claim that
White has compensation, but after
20...�fs 21 '!J.fl �g6 Black is much bet­
ter) lS ... �xes 16 �d4 �fs was
S.Bouaziz-J.Nunn, Szirak lnterzonal
1987. Black has recovered the sacrificed
pawn and has the initiative to boot. After 11 ... tlla6, the es-knight really
c) 11 '!J.dl is similar to 11 0-0-0, but is hanging. White has three options:
White's development is worse: 11...tlla 6
12 lllf3 (after 12 �6, 12 ...tllds ! ? is pos­ C1; U ti}fJ
sible, while 12 f4 h6 13 �h4 gs 14 �g3 C2:U f4
tllcs leaves White's king uncomfortable CJ: U lld6
on el, as shown by lS �f3 tllfxe4! 16
tllxe4 tllxe4 17 �xe4 gxf4 18 �xf4 Cl) 12 tllf3
�xes) 12 ... tll cs 13 tlld2 ( 13 es tllg4
does not help) 13 ... h6 14 �e3 (after 14
�xf6 �xf6 lS f3 as White's king can­
not help support the queenside)
14 ... tllfxe4 lS tll dxe4 tll x e4 16 tllxe4
'!J.xe4 17 '!J.d8+ �h 7 with the idea of
...�f6 compares favourably for Black to
Line Cl, below, because White's hl­
rook is not in play. There White's king is
on cl, although here it is still equal.
11 tlla 6

Taking the knight is bad: 11 ...'!J.xe s? This simple move has not scored
12 '!J.d8+ '!J.e8 13 �xf6 '!J.xd8 14 �xd8 well, but that is because White strug­
leaves White a pawn up for nothing. gled to consolidate in the early days of

A t tacking C h e s s : Th e King 's Indian, Vo l u m e 1

the gambit. In fact, the move is not l:tc2 when the passed a-pawn gives
harmless and Golubev considers that it White an enduring advantage) lS
causes Black some problems. Often �xg4 �xd4 16 f3 tllb4 17 �bl ! �xc3
Black will secure the bishop-pair and (better is 17 ...�f8, but White still has
good squares for his pieces, but an extra pawn) 18 l:txd8! (18 bxc3 tll d3
whether this is sufficient compensa­ 19 �c2 tlle s gives Black compensation)
tion for a pawn is open to question. 18 ...l:txd8 19 bxc3 tlla6 (not 19 ...tlld 3?
Perhaps it is a bit of a matter of taste. 20 l:tdl with a deadly pin) 20 �c2 and
12 tllc s!
... White has good chances of realizing his
I believe this is best, as an examina­ extra pawn.
tion of the alternatives reveals: 13 tlld 2
a) 12 ... tllg4 13 �h4 (13 l:thfl tllc s After 13 �xf6 �xf6 the pressure on
wins back the pawn pretty cleanly) c3 and e4 should allow Black to win
13 ... tllcs when both 14 h3 tllf6 lS l:thel back his pawn without any real conces­
and Golubev's suggestion 14 tlld 2 �xc3 sions.
lS bxc3 tllxe4 16 tllxe4 .rtxe4 17 l:td8+ 13 h6

�g7 look insufficient for Black. Another idea is 13 ...tllg4 14 �xg4

b) 12 ... h6 13 �xf6 �xf6 14 tlld4 (14 �xg4 lS f3 �e6 16 �c2 and now
es �xes lS tllx es .rtxes 16 .rf.d8+ �g7 16 ... tll a6 !? (Golubev gave 16 ... bs 17
with the idea of ... tlle7 is probably okay �e3 tll a6 18 cxbs cxbs 19 �bl) 17 a3
for Black) 14...tllcs lS f3 when Black has (17 �bl tllb4) 17 ... bs 18 cxbs cxbs 19
some compensation for the pawn, but tllx bs .l:tec8+ 20 tllc3 tllcs with the idea
White is well centralized and I imagine of ...tll a4 could be investigated.
only the first player can really try to
win this position.
c) 12 ... �g4 13 �xf6! (after 13 h3
tllxe4! 1 4 tllxe4 l:txe4 13 es tlld 7 1 4 h3
�xf3 lS �xf3 tllx es 16 �e2 tllc s
Black's easy development and dark­
square control compensated for
White's bishop-pair in A.Torrecillas
Martinez-G. Hernandez, Benasque
1997) 13 ...�xf6 14 tlld4! l:tad8 (after
14 ... �xd4 Golubev gives the line lS
l:txd4 �xe2 16 tllxe2 cs 17 .rf.d7 l:txe4 18 14 �e3
tllc3 .rtxc4 19 l:txb7 tllb4 20 �bl! l:td8 Perhaps better is 14 �xf6 �xf6 and
21 a3! tlld s 22 tllxds l:txds 23 .:tel! l:th4 then:
24 h3 l:tgs 2s g3 l:txh3 26 l:txa7 l:th2 27 a) lS f3 as (Black could try 1S ... �xc3

Exch a n g e Varia ti o n : 7 dxe 5 dxe5 8 'ilixd8 '!J.xd8

16 bxc3 fS) 16 tllb3 tll e 6 17 tlla4 and ity which has evolved into his main
Black struggled to find enough for the approach .
pawn in L.Voloshin-J.Helbich, Litomysl
b) 15 �c2 as (Bl ack should avoid
15 ... �xc3 16 �xc3 tllxe4+ 17 tllxe4
'!J.xe4 18 '!J.d8+ �g7 when both 19 �d3
and 19 �f3 !? '!J.xc4+ 20 �d2 are better
for White) 16 '!J.hel tlle6 (maybe not the
best; 16 ... a4! ? could be tried) 17 g3
tlld4+ 18 �bl was M. Sorokin-D.Pereyra
Arcija, Buenos Aires 1993. Black has
some compensation here, but again it
is not clear that it is enough. 12 h6

14... tllfxe4 15 tlldxe4 tllxe4 16 tllxe4 After 12 ... tllcs 13 �f3 several moves
'!J.xe4 17 '!J.d8+ �h7 have been tried, but without much sue-
a) 13 ... as looks too slow after either
14 '!J.d6 OT 14 �h4.
b) 13 ...�e6 14 �h4 tllfd7 15 tllxd7
tllxd7 16 �c2 just leaves White a pawn
up for nothing.
c) 13 ... tllfxe4 14 tllxe4 tllxe4 15
�xe4 f6 16 �xf6 �xf6 17 '!J.hel when
White is a pawn ahead and well cen­
d) 13 ... tlle6 14 Lf6 Lf6 15 tlld3
This looks uncomfortable at first, looks insufficient - White e4- and f4-
but Black intends .. if6, ...l:te8, and pawn duo is more of a strength than a
.. ie6 when the position will be com­ weakness.
pletely equal. e) 13 ...h6 14 �xf6 (14 �h4 trans­
poses to th e notes to Black's 13th in our
C2) 12 f4 main line) 14 ...�xf6 15 tlld3 and White
Perhaps this move is the most obvi­ keeps a useful extra pawn because
ous. White simply protects the ad­ 1s ...�xc3 can be met with 16 tllx cs!.
vanced knight. The downside to this f) 13...tllh s 14 �xh s (14 tlld3 ?!
move is that the e4-pawn may become tllxd3+ 15 '!J.xd3 h6 is already better for
weak. Also Black has a tactical possibil- Black) 14...gxhs (threatening .. .f6) 15

A ttacking C h e s s : The King 's I n dia n , Vo l u m e 1

�h4 �h6 16 �c2 �xf4 17 tllxf7! tllxe4 G.Camacho-L.Valdes, Cuba 1998, but
18 l:thfl tll xc3 19 �xc3 �g4 20 l:td3 is a 20... �xfs is fine for Black.
rather peculiar position, but White still a2) 19 �dl tllxe4 20 tllxe4 �xfs! 21
maintains the better chances. tll cs b6 22 .:te3 (or 22 l:tel bxcs 23
13 �h4 .l:.xe8+ l:.xe8 24 l:td6? �es 0-1 J.Bonin­
After 13 �xf6 �xf6 White's es- and K.Pohl, Philadelphia 2000) 22 ...bxcs 23
c3-knights allow Black to easily recover .:txe8+ l:txe8 24 �xcs hS with some ini­
the pawn. tiative for Black in J.Bonin-C.Riordan,
US League 2006.
b) 17 g3 ! is a problem, however, as
17 ...tll h3 18 tlld3 leaves Black strug­

u gsl

The alternative is 13 ...tllc s 14 �f3

gs lS �f2 tlle6 (after 1S ... gxf4 16 �xcs
l:txes 17 �d6 l:te8 18 �xf4 tll d7, as in
B.Finegold-G.Barbero, Wijk aan Zee 14 �g3
1991, Panczyk and llczuk claim that This is the most common move.
Black has compensation, but I doubt White lends support to the h 2-b8 di­
that it is enough) 16 fS! (after the 16 agonal. There are a couple of alterna­
tllxf7 �xf7 17 es tllxf4 18 exf6 �xf6 19 tives too:
tlle4 of J.Bellon Lopez-J.Becerra Rivero, a) 14 �f2 gxf4 1S tllf3 and then:
Santa Clara 1997, Black should play al) 1S ...tll x e4 16 tll x e4 �fs 17 tllfd2
19 ... �es 20 �d4 �fs, while after Bellon �xe4 18 tllxe4 l:txe4 19 �d3 l:tes is
Lopez's suggested 20 g3 Black has level. White has some compensation
20 ... g4!) 16 ...tllf4 (16 ...tllf8 17 �g3 is no for the pawn, but no more than that.
good for Black) and now: a2) 1S ... tll g4! ? 16 .lii.d4 tll e 3! ? (alter­
a) 17 tll d 3 tll x d3+ 18 l:txd3 g4 with a natives are 16 ... .lii.e6 and 16 ... �xd4 17
further divide: l:txd4 tllc s) 17 l:td2 �g4 18 h3 �xf3 19
al) 19 .lii.e2 tll xe4 20 tll xe4 l:txe4? 21 gxf3 �xd4 20 l:txd4 l:tad8 looked prom­
l:td8+ was winning for White in ising for Black in J.Bellon Lopez-

Exch a n g e Va riati o n : 7 dxe5 dxe5 8 'ilixdB '!J.xdB

S.Kindermann, Biel 1991, because of tllxg4 fS 21 tllf2 tll xg3 22 hxg3 �f7 23
White's bad bishop. �d3 '!J.e3 with compensation for the
a3) 15 ... �g4 16 h3 with a further pawn in R.Thomas-J.Hebert, correspon­
split: dence 1997.
a31) 16 ... tllxe4 17 tllxe4 '!J.xe4 b22) 16 ...tllcxe4 is more forcing.
(17 ...�fS! ? is also possible) and here, Black must be a little careful, but he
instead of 18 hxg4? .l:!.xe2 which was seems to be fine with exact play after
clearly better for Black in H.Hoeksema­ 17 tllxe4 tllxe4 18 �hs tllx g3 (18 ... �e6
J.Nunn, Groningen 1988, White should 19 �xf7+ �xf7 20 tllxf7 tllx g3 21 hxg3
settle for 18 �d3 �xf3 19 gxf3 '! '!J.e2 looks a bit risky) 19 �xf7+ �h7 20
which is level. �g6+ �h6 21 hxg3 �xes (but not
a32) 16 ...�hs 17 es �xf3 18 gxf3 21 ...'!J.xes 22 �c2! with a big advantage
.!l:ihs 19 '!J.hg1 �f8 20 tlle4 �xes 21 for White) 22 �xe8 �g4 23 �f7 �xdl
�d3 was P.Cramling-M.Wahls, Ham­ 24 '!J.xdl �xg3 with a drawn ending.
burg 1991, when Cramling gives 14...tllc s 15 � f3 g4 16 �f2
21 ... tllg 3 22 tllx g3 fxg3 23 �xg3 tll cs 24 This leads to complicated play, but
i.h7 �xg3 25 l:txg3 as equal. ultimately leads to approximate equal­
b) 14 fxg s hxgs 15 �g3 tllcs and ity. Instead 16 �e2 tllcxe4 17 tllxe4
here: tllxe4 18 �xg4 tllx g3 19 hxg3 �xes 20
�xc8 �xb2+ 21 �xb2 '!J.axc8 22 '!J.xh6
l:te2+ 23 �b3 '!J.xg2 is a drawish double­
rook ending.
16 ...�f8 has also been tried, but 17
.l:!.hgl! is a clever way for White to keep
the tension .
17 �xcs fxg2 18 l:thg1 tll h s

bl) 16 �d3 tllh s 17 lllf3 �e6 18 �c2

.!l:ixd3 19 .:txd3 �xc4 20 '!J.d7 g4 21 tlld2
�a6 was better for Black in D.Starke­
P.Bakalar, Prague 1990.
b2) 16 l:thfl when Black has:
b21) 16 ...�e6 17 �f3 (17 lllf3 tllfxe4
18 tllxe4 tllxe4 is pleasant for Black)
17 ...g4 18 �e2 tllcxe4 19 tllxe4 tllxe4 20

A ttacking C h e s s : The King 's I n dian, Vo l u m e 1

19 �e3 any initiative, the opposite-coloured

White has also tried 19 tlld 3, but bishops will give the game rather
Black gets sufficient counterplay here drawish tendencies.
as well: 2S b3
a) 19 ... �xc3 20 bxc3 l:txe4 21 .l:txg2+ This allows Black to protect the far­
�h7 22 �d4 �h3 23 l:tf2 �f5 24 tllc5 flung g2-pawn, but 25 .l:txg2 �xc4 is
l:te7 was level in B.lvkov-C.Werner, Wijk not too dangerous. After 26 tllg4
aan Zee 2000. Golubev gives 26 ...l:tg8 with the idea of
b) 19 ... b6 20 �d6 �xc3 21 bxc3 ... l:tg6 and .. .f6.
l:txe4 22 .rtxg2+ �h7 23 tll e5?! �h 3 24 2s ...J:tgs 26 �d2
l:tf2 f6 25 tllxc6 l:txc4 26 tll e7 .l:txc3+ 27 White wants to walk his king to f3
�b2 .l:tc4 and Black had a good extra to round up the g2-pawn.
pawn in F.Bindrich-G.Kamsky, Caleta 26... bsl
2010. Black is in time to create counter­
19... �h3 20 l:td3 play on the queenside.
White intends 21 �d2, uncovering 27 cxbs cxbs 28 �e2 .l:tc8 29 �d2 .l:tg8
an attack on the h3-bishop. 3o �e2 as
20....l:tad8 Black could repeat, of course, but
The most precise move order. In­ this does not change the assessment of
stead 20 ...�xe5 21 fxe5 l:tad8 22 l:txd8 the position.
l:txd8 23 tlldl �h7 would transpose, 31 �f3 a4 32 bxa4 bxa4 33 a3 �c4
but gives White the extra option of 22
21 .l::!. xd8 .l:txd8 22 tlld1 �xes 23 fxes
�h7 24 tllf2 �e6

34 .li es
Of course, 34 .l:txg2? allows
34... �e2+, while 34 tllg4 �fl (34 ... �e2+
35 �xe2 l:txg4 36 �f3 l:th4 37 l:txg2
The ending still looks complicated, l:th3+ 38 �e2 tll g7 is pretty level) 35
but if neither side is able to generate tllxh 6 l:tc8 looks very unclear.

Exch a n g e Va ria tion : 7 dxes dx e5 8 'ilixd8 '!J.xd8

34...�fl 35 tllg4 l:tb8 36 tlle3 tllcs lS '!J.d6! (even in a quiet variation

White could play 36 �b4, because Tal finds the tactics; instead lS f3 �e6
36...'!J.xb4 37 axb4 a3 38 tlle3 a2 39 tllc2 leaves the f6-rook in trouble) 1s ... tllxe4
stops the a-pawn. Still, after 39 ... tllg 7 16 '!J.d8+ �g 7 17 tllxe4 '!J.xe4 18 �f3
one gets the feeling that any chances '!J.xc4+ 19 �d2 was much better for
that there are lie with Black. White in M.Tal-1. S mirin, Podolsk 1990.
36...'!J. b3 Black always has to watch out for these
Now any neutral move such as 37 �d6 back-rank pins.
should draw, but after 37 �f2? tllf4 38 b) 12 ...tll x e4 hands White the
.!Lixfl gxfl'ili+ 39 �xfl '!J.b1+ 40 �2 bishop-pair: 13 tllx e4 �xes 14 tllf6+
.!Llh3+ Black won in H.Terrie-F. Felecan, �xf6 1s �xf6 tllcs 16 �f3 (16 �d3
Kona 1998. tllxd3+ 17 '!J.xd3 �fs is equal) 16 ...�fs
17 '!J.hdl and White is obviously better.
C3) 12 '!J.d6 c) 12 ...�f8 is rare: 13 '!J.xf6 '!J.xes 14
�f4 (Golubev's suggestion 14 �e3 may
be even stronger) 14...'!J.e8 lS �d6 �g 7
16 �xf8+!? (16 es �xd6 17 '!J.xd6 '!J.xes
18 '!J.d8 tlle7 looks okay for Black)
16 ... �xf6 17 �d6 with good compensa­
tion for the exchange.

C31) 12 ...�e6

This counterattacking move was

brought to prominence by Tal, and was
later employed in a well-known game
between Korchnoi and Kasparov. It is
quite challenging and I will cover the
two main lines.

C31: 12 ..i.e6

C32: 12 tll h s
••. This is Black's main move and it has
been more popular than all his other
Black has a few other options, of moves combined. Of course it is play­
which only 'c' is really important: able, but I do not see a clear way for
a) 12 ...'!J.xes 13 �xf6 �xf6 14 '!J.xf6 Black to equalize.

A ttacking C h ess: Th e King 's I n dian, Vo l u m e 1

13 f4 compensation according to Kasparov)

Instead 13 f3 leads nowhere after lS ... ..ixes 16 fxe S lfac8 (instead
13 ... h6 14 ..ih4 gs 1s ..ig3 ll'lh s 16 ll'ld7 16 ... 'iW8 17 cs! ll'lxcs 18 b4 is good for
l:tad8 17 l:thd1 ll'lxg3 18 hxg3 l:txd7 19 White, but 16 ...l:te7 17 l:td8+ l:te8 is
l:txd7 ..ixd7 20 l:txd7 ll'lcs 21 l:td2 (after fine) 17 'it>d2 l:te7 18 'iii> e3 l:tce8 19 l:td8
21 :d1 ..ixc3 22 bxc3 fS Black is slightly 'iii>f8 was V.Korchnoi-G.Kasparov, Til­
better) 2 1... ..ies when Black wins back burg 1991. White remains up a pawn,
the pawn, 1.Umanskaya-A.Shchekachev, although it will be very difficult to con­
St Petersburg 1994. vert.
Another less-critical continuation is b) 14 ..ih4 gs 1s fxgs hxgs 16 ..ixgs
13 ll'lf3 ll'lcs 14 ll'ld2 ll'lfd7 (trying to ll'lcs 17 ..ixf6 ..ixf6 18 ll'ld3 ..ie7 19
corral the d6-rook, but both alterna­ ll'lxcs (after 19 l:td4 ..if6 20 es ll'lxd3+
tives 14.....if8 and 14...ll'lg4 lS ..ih4 21 l:txd3 ..ixes Black has compensation,
..ic8!? look like better tries) lS ll'ldbl!? while after 19 es ..ixd6 20 exd6 it is
fS! when Black has some initiative for White who has compensation for the
the pawn. sacrificed material) 19 ... ..ixd6 20 ll'lxb7
is complicated, but it looks a little bet­
ter for White.
14 ..if3
Or 14 ..ixf6 ..ixf6 lS ll'ld3 (after lS
J:thdl ..ixeS 16 fxeS aS O T 16.. lte7 Black
is fine) 1s ... ll'lxd3+ 16 ..ixd3 (16 l:txd3 ?!
..ixc4) 16.....ie7 17 cs ..ixd6 18 cxd6 f6
19 es fxes 20 fxes l:tf8 when the posi­
tion is unclear, but I do not think Black
is any worse.
14 ...ll'lhs
13 ...ll'lcs Here 14...h6 lS ..ih4 gs does not
This is the most popular move, but work well because after 16 ..if2 g4? 17
13 ... h6 was Kasparov's choice. This al­ ..ixcs gxf3 18 gxf3 Black is just down a
ways means something, but it still does couple of pawns. Instead 14...ll'lfxe4 lS
not look like a lot of fun for Black. ll'lxe4 ll'lxe4 16 ..ixe4 f6 17 ..ixf6 ..ixf6
White has: 18 ..id3 also looks in sufficient
a) 14 ..ixf6 ..ixf6 lS l:thdl (after l S 15 ..ixhs gxhs 16 ll'ld3!
ll'lf3 ..ixc3 1 6 bxc3 ll'lcs White must This is better than 16 ..ih4 when
avoid 17 l:td4 ..ig4 18 h3 ll'lxe4 19 hxg4 Black can simply play 16 ... ..ixes 17 fxes
ll'lg3 when Black wins, and play 17 ll'ld2 ..ixc4 or try 16 .. .f6 17 ll'lf3 l:tad8 with
when 17.....ifS 18 es l:tad8 gives Black counterplay.

2 60
Exc h a n g e Va riat io n : 7 dxes dxes 8 -.xd8 :xd8

White's es-knight is now attacked

twice and .. .f6 is threatened, so the
knight has to move.
13 ll'lg4
This has been played several times
but may not be best. Alternatively:
a) 13 ll'lf3 ll'lcs 14 ..ie3 (14 ll'ld2 ..ies
lS :d8 .rtxd8 16 ..ixd8 ll'lf4 only gives
White trouble, while 14 es ll'le6 with
the idea of ...ll'lef4 is given by Golubev)
14.....ixc3 lS ..ixcs ..ig7 ! 16 ..ie3 (16 es
16 .....ixc3 ll'lf4 17 ..ifl ..if8) 16.....if8 17 l:tddl is
Instead 16 ... ll'lxd3+ 17 l:txd3 ..ixc4 given by Golubev, although after
18 l:tg3 'iW8 19 es is much better for 17 ...ll'lf6 18 ll'ld2 ll'lxe4 19 ll'lxe4 l:txe4
White according to Golubev. 20 .rf.d8 'iii>g 7 or simply 17 ...l:txe4 this is
17 ll'lxcs ..ib4 18 ll'lxb7 ..ic8 19 a3! probably not much for White.
White is better. Black should proba­ b) 13 ll'ld3 ! ? contains the a6-knight
bly play 19 ...l:txe4, but White has sev­ and may well be best: 13 .....if8
eral moves to keep some initiative: for (13 .....ixc3 14 bxc3 l:txe4 lS l:td8+ 'iii>g 7
example, 20 axb4 (other possibilities 16 l:tel must favour White) 14 l:td8
are 20 l:txc6 ..ixb7 21 l:te7, 20 l:td8+ 'iii>g 7 (similar is 14 l:td4 ll'lg7 1s :d8 l:txd8 16
21 ll'ld6 ..ixd6 22 :xd6 l:txc4+ 23 'it>d2 ..ixd8 ..ie6 17 ..if6, C.Lakdawala­
and 20 cs l:tc4+ 21 'iii>b 1 ..ixcs 22 ll'lxcs J.Peters, Irvine 1998) 14...l:txd8 lS ..ixd8
.:xcs 23 l:thdl) 20...l:txc4+ 21 'iii> b 1 ..ixb7 ..ie6 (1s .....ih6+?! 16 'iii>c2 ..ie6 17 ..ih4
22 l:thdl with a slight advantage. ..ixc4 18 ..ixhs gxhs 19 f3 was better
for White in J.Gustafsson-A.Matth aei,
c32) 12 ... ll'lhs Hamburg 1993) 16 ..ig s ..ixc4 17 l:td1
ll'lg7 18 f3 ll'le6 19 ..ie3 looks somewhat
better for White because of his strong
central position.
13 ... ll'lcs
After 13 .....ixc3 14 bxc3 (14 ll'lh6+!?)
14...ll'lcs (Black does better to play
14... ..ixg4, transposing to the next line)
lS ll'lh6+ 'iii>g 7 16 f3! ll'le6 17 ..id2!
White keeps an advantage according to
Golubev. A better move order for Black
is 13 ... ..ixg4 14 ..ixg4 ..ixc3 lS bxc3

A ttacking C h e s s : The King 's I n d i a n , Vo l u m e 1

l:txe4 with the idea of 16 l:td4 l:tes. a knight landing on the d3-square) 19
es ..ixes White starts to experience
some difficulties because of the knight
stuck on h6.

14 ll'lh6+
Instead 14 ..ie3 ..ixc3 lS ..ixcs ..ixg4
16 ..ixg4 l:txe4! is a clever trick given by
Golubev. After 17 ..ixh s Black can play 17 .....ic7!
17 .....ies! with equal chances. This move is an idea of the Ameri­
14...@fs 15 ..ixhs can IM Joe Fang. Black intends ... ll'le6,
Now lS f3 runs into 1S ...ll'le6! with hitting the gs-bishop and eyeing the
the idea of ...ll'lhf4. Probably best is lS f4-square. The bishop retreat main­
..ie3! ll'lxe4 16 ll'lxe4 l:txe4 17 ..ixh S tains control of f4 and uncovers an at­
gx hS 18 l:td8+ �e7 19 l:txc8! .l:i.xc8 tack on e4-pawn. All of White's at­
(worse is 19 ...l:txc4+ 20 @bl l:txc8 2 1 tempts to defend it make some kind of
ll'lfs+ @d7 2 2 ll'lxg7 l:tg8 2 3 ll'lxhs l:txg2 tactical concession, as his line-up of
24 ll'lg3) 20 ll'lfs+ @f8 21 ll'ld6 l:tce8 22 pieces on the cl-h 6 diagonal will prove
ll'lxe4 l:txe4 23 b3 as given by Golubev. to be vulnerable.
White has the better pawn structure, In stead, both 17 ... ..ig7 18 l:thdl ll'le6
but the position is drawish. 19 ..ie3 cs 20 l:tds b6 21 es! and
1s ... ..ixc3! 16 ..if3 17 ... ..ie6 18 b3 (or 18 ..ie3 b6 19 b3)
Not 16 bxc3 ll'lxe4. 18... as 19 ..ie3 b6 (A.Onischuk-
16 ... ..ies M.Golubev, Leningrad 1989) 20 @bl!
This is queried by Golubev, who in­ (Golubev) allow White to fight for an
stead suggests 16 ... ..ib4 17 l:tddl ll'lxe4 edge.
18 ..ixe4 l:txe4 19 l:td8+ l:te8 20 .l:i.xe8+ 18 .l:i.e2
@xe8 or 16 ... ..ias!?. In stead, 18 l:tel walks into 18 ... ..ias
11 .l:i.d2 and 18 ..ie3 ll'lxe4 19 l:tel @g7 20 ..id4+
After 17 l:tddl ..ic7! 18 l:thel fS ! @xh6 21 .l:i.xe4 l:txe4 22 ..ixe4 ..if4 23
(Black tries to exploit the possibility of ..ie3 ..ixe3 24 fxe3 ..ie6 is a little better

Exc h a n g e Va ria t i o n : 7 dxes dxes 8 -.xd8 'IJ.xd8

for Black. and maintains the initiative.

18 ...'iii>g 71 19 @bl 19 . ..ll'le6 20 h4
Black has all sorts of tricks: for ex­ If the bishop retreats, Black simply
ample, 19 h4 f6 20 ..id2 ..if4! 21 ll'lg4 plays 20 ...ll'lf4.
.ixg4 22 ..ixg4 l:txe4 and 19 g3 f6 20 20 ...ll'lxgs 21 hxgs ..if4
.i..d2 ll'ld3+ 21 'iii> c 2 ll'les. In both cases Black wins back the pawn and
Black wins back the sacrificed pawn stands well.

Part Ill
The Samisch: variation
1 d4 t'i)f6' z c4 g6 3 liJc3 i.g7 4 e4 d'6
5 f3 o-o 6 ..te3 tiJc6 7 ttJgez a6 8 1fd2 :bs

1 d4 ll'lf6 2 c4 g6 3 ll'lc3 .i.g7 4 e4 d6 5 ming-up of the kingside pieces, the d4-

f3 square is sometimes weak
s ...o-o 6 .i.e3
The alternatives 6 and 6 ll'lge2
will be discussed in Chapter 20. After 6
.i.e3 Black has a wide choice, as we
noted in the Introduction. The most
popular move today is definitely 6 .. cs,

but I had no hesitation in going for:


The Samisch Variation is White's

other big main line with 4 e4. White
creates a strong centre and maintains
a great deal of flexibility. There are a
couple of downsides to White's scheme
of development, however. The move 5
f3 takes away the most natural square
for White's king knight and developing
it to the e2-square will block in the fl- This is the Panno Variation. It is
bishop. Because of this slight gum- flexible and ambitious. Black takes aim

2 64
Th e S d m is ch Va riation

at the d4-square with his knight, aim­ Westerinen Variation of Chapter 20, in
ing for active piece play rather than the which case 7...l:te8 must be played, be­
protracted manoeuvring that is com­ cause 7 .. ltb8? 8 ds ll'les 9 i..x a7 wins a
mon with the more classical 6 ... es and pawn as the c4-pawn is still protected.
6... cs. Black holds ... es in reserve and This and other early deviations will be
after White overprotects the d4-square discussed in Chapter 20.
with ll'lge2, Black will look to the 1 a6 8 ..ifd2 ftb8

queenside with ...ftb8, ...a6 and ... bs.

This plan takes into account that White
may castle long and from a positional
standpoint it highlights the fact that
the c4-pawn is lacking defence.
The Panno is not very popular
nowadays, but I think this is mostly due
to the success of 6 ... cs at high levels
and consequently the decline of the
Samisch in general. 6 ...ll'lc6 has re­
mained popular in correspondence
play and has been played frequently by Black is ready to play ... bs and so
Nunn and Gufeld, while Kaspan;>v has White must decide on a plan. He may
been found on both sides of the board. play the aggressive 9 h4, which is con­
I believe it is due for a fresh look. sidered in the next three chapters, pre­
7 ll'lge 2 pare to develop his kingside with g ll'lcl
White sometimes plays 7 ..ifd2 first, (Chapter 17), initiate kingside play
although it makes little difference without 9 h4 (Chapter 18), or play on
unless Black intends to play the the queenside (Chapter 19).

Chapter 14
Panno Variation
g h4 without 9 hs ..•

1 d4 ll'lf6 2 c 4 g 6 3 ll'lc3 .i.g7 4 e 4 d 6 5 with 9 ...e 5 or initiate queenside play

f3 o-o 6 .i.e3 ll'lc6 7 ll'lge2 a6 8 'ilfd2 .l:i.b8 with 9 . .bs.

9 h4 Although 9 ...h s is considered the

main line nowadays and will be the
subject of our next chapter, 9 .. es and

9.. bs have not been refuted and they


lead to complex play. I imagine that

these two lines could prove to be un­
pleasant to face for some white play­
ers, especially those who intend to play
the positional 9 h4 h S 9 ll'lcl lines of
Chapter 16.

A: 9 es

This is the most aggressive move, al­ B: 9 bs


though White may still play position­

ally in certain lines. The move's basic A) 9 es

intention is obvious - White intends to Black follows the logic that a flank
attack the black king. The play very of­ attack should be met with play in the
ten is similar to the Sicilian Dragon. centre. However, because the d4-
Usually Black halts the advance of the square is well protected, it turns out
white h-pawn at once with 9 ... h s, al­ that Black actually cannot play in the
though there was a time when it was centre after all. Indeed, if 9 ... es is to
thought that this move was too weak­ prove viable, it will be by combining
ening. He may also strike in the centre defence and counterattack.

Pa n n o Va ria tio n : 9 h4 with o u t 9 . . . h 5

12 .....id7 13 ..ih6
It is hard to believe that 13 can
trouble Black after 13 ...b5, while 13
o-o-o b5 14 ..ih6 ..ixh6 15 �xh6 trans­
poses to the main line.
13 .....ixh6
This is a common idea that is also
frequently seen in the Dragon. Black
invites the white queen to h6 hoping
that White's queenside will prove to be
more vulnerable from its absence
10 ds ll'las 1 1 ll'lg3 there. It is also possible to play 13 ... b5!?
Geller's old book only considered 11 as after 14 ..ixg7 'iii>xg7 if White wants
b3, but this move just weakens the to invade with his queen, he must play
queenside for no reason and Black has 15 hxg6 fxg6 first, which could prove to
good play after 11...c5, intending ...b5. be premature. After 16 'ilfh6+ 'iii> g 8
The text move activates White's king­ Black can defend himself laterally
side and threatens 12 b4. along the second rank.
11... cs 12 hs 14 'ilfxh6 bs 15 o-o-o
After 12 I think Black should
play Gallagher's recommendation
12 ...h5!, preventing White from having
his way all over the board. Instead
12 ...b6 13 h5 ..id7 14 b4 cxb4 15 l:txb4
"fic7 16 h6 ..ih8 17 'ilfb2 b5 18 cxb5
.:.fc8 19 bxa6! was winning for White
in V.Kramnik-V.lsupov, Kuibyshev 1990.

White can also play 15 ..ie2 when

15 ...'ilfe7 (this is the most popular
move, but I do not like it; the queen
should hang around because White has
ll'lf5 ideas in the air, but d8 is a good,
flexible square! ) 16 o-o-o ll'lxc4 (Black
could also try 16 ...'iii>h 8!? or 16 ... l:tb7) 17
..ixc4 bxc4 18 ll'lfl allows him a typical

Attacking C h ess: Th e King 's Indian, Vo l u m e 1

build-up with ll'le3, :d2 and g4. More 2 1 ..We3 gxhs 2 2 ll'lg3 i s less clear) when
critical is the immediate 15 ... ll'lxc4 16 Black's attack is the stronger. For ex­
..ixc4 bxc4 17 0-0-0 l:tb4 when Black's ample, 20 'ilfxf6? .l:i.xb2 and White will
queen can still go to either side of the not last long.
1s ...ll'lxc4 B) 9...bS
Instead 1S ... 'ilfe7 16 l:td2 .l:i.b7 was
seen in Y.Kraidman-H.Westerinen,
Gausdal 1983, but this looks too defen­
sive to me.
16 ..ixc4 bxc4 17 ll'lfl ..Was!
Now that there is no ll'lfs to worry
about, the queen takes up an aggres­
sive post on the queenside.
18 .lld 2
After 18 g4? l:txb2! (also strong is
18 .....ia4! 19 l:td2 l:txb2! 20 ll'lxa4 l:tb7,
winning) White resigned in Black continues with his plan.
U. Lenhardt- .Crowdy, correspondence 10 hs es
1985. Following 19 'iii>xb2 l:tb8+ 20 @c2 This seems similar to the previous
..ia4+ 21 ll'lxa4 ..Wxa4+ 22 @d2 ll'lxg4! line, but here Black has played ... bS to
23 fxg4 c3+ White's position falls apart. attack c4 first, so the move ... cs will not
18 ...l:tb4 be necessary (although it is still possi­
ble). There are a couple of interesting
alternatives too:
a) 10 ... bxc4 and then:
al) 11 g4 ..ixg4!? (Black could inves­
tigate other moves) 12 fxg4 ll'lxg4 13
o-o-o ll'lxe3 14 �xe3 e6 15 hxg6 hxg6
(in stead 1S ...fxg6 16 'ifh3 ..Wgs+ 17 'iii>b1
l:txb2+ 18 'iii>xb2 .rf.b8+ 19 @al ll'lb4 20
..Wxe6+ ! 'iii>h 8 21 a4 ll'lc2+ 22 @a2 ll'lb4+
23 'iii>b l ll'ld3+ 24 'iii>al is winning for
White) was G. Kasparov-B. S passky, Nik­
sic 1983, and now 16 ll'lgl! looks good
Black has good counterplay. For ex­ for White.
ample, the natural 19 ..Wgs can be met a2) 11 hxg6 fxg6 12 ll'lf4 e6 (after
by 19 ...l:tfb8 (19 ... ll'lxhs 20 l:txhs .l:i.fb8 12 ... ll'las 13 o-o-o c6 14 g4 White has

Pa n n o Varia t i o n : 9 h 4 wit h o u t 9 . . . h 5

the initiative, as shown i n some of how he can untangle.

Murey's games) 13 ..ixc4 dS 14 ..ib3 b) 10...l'iJas 11 0ig3 (11 0,f4?! pre­
l:txb3! lS axb3 dxe4 16 o-o-o exf3 17 vents White from playing ..ih6 and
gxf3 is unclear after either 17 ...0ie7 or makes little sense) 11...0ixc4 12 ..ixc4
11 ... l'iJas. bxc4 and now:
a3) 11 ..ih6 has been considered a bit bl) 13 o-o-o c6! 14 ..ih6 ..Was 1s
of a refutation of Black's play. I am not ..ixg7 'iii>x g7 16 hxg6 fxg6 17 'it'h6+ 'iii> g 8
so sure about this, though: 11 ...0ib4! 12 18 es dxes 19 0ige4 was J.Peters­
0ig3 ..ixh6 13 'ilfxh6 0ic2+ 14 'iii>d l l'iJxal A.Matikozian, Los Angeles 2004, and
lS hxg6 (after 1s l'iJds, instead of 1 s ... e6 here 19 ...:f7! looks good for Black.
16 hxg6 which was winning for White in b2) 13 ..ih6 ..ixh6 14 'ilfxh6 and
M.Ceteras-D.Elliott, Mamaia 1991, be­ here:
cause of 16...exds 17 l'iJhs, Black can b21) 14...g s 1s ..Wxgs+ (1s es!? may
play 1s ...gS! abruptly ending the white also be good) 1 s ... @h8 16 o-o-o c6 17
attack) 1 s.. .fxg6 16 l'iJdS ..ie6 (but not :d2 l:tg8 18 'ilff4 ..Was 19 0if1 ..ie6 20
16....l:i.f7? 17 0ixf6+ exf6 18 ..ixc4, win­ 0ie3 gave White some advantage in
ning) and here: P.Van der Sterren-M.Bosboom, Brussels
b22) 14...l:txb2 !? is a sharp try: lS es
(or lS hxg6 fxg6 16 es dxes 17 0-0-0
l:txa2!?) 1s ... dxes 16 o-o-o (not 16 dxes
'ilfd4!) 16 ...exd4 is unclear. If 17 hxg6
(or 17 0ice4 .ifs) 17 .. .fxg6 18 0ice4 .l:i.f7
19 'iii>xb2, Black could try 19 ... cs,
19 .....ifds or 19 ...0ixe4 20 0ixe4 .ifs, in
all cases with a total mess.
We now return to

a31) 17 0,f4 ..if7 18 l'iJfhS was

V.Rajlich-K.Zalkind, Budapest 2000.
Here 18 ... gxhs 19 l'iJfs 0ie8 20 l:th3 ..ig6
21 ..ixc4+ e6! 22 ..ixe6+ �U7 23 l:tg3
'ilff6 would successfully defend.
a32) 17 0ixf6+ l:txf6 18 'ilfxh7+ 'iii>f8
19 'ilfh8+ ..ig8 20 ..ixc4 e6 2 1 l:th7 �U7
22 ..ixe6 lhh7 23 'ilfxg8+ 'iii>e 7 24 'ilfxg6
'it'h8 2S l'iJfS+ 'iii>f8 is unclear. Black is up
a lot of material, but it is hard to see

A ttacking C h e s s : Th e King 's In dian, Vo l u m e 1

11 d5 J.Tarjan-L.Nezhni, USA 1982, and Galla­

Invariably played, but 11 0-0-0 must gher's suggestion 16 ll'lfl, with the idea
also be possible. Indeed, we'll briefly of ll'ld2xc4, favour White.
discuss it in the notes to White's 11th 13 0-0-0
move in Line A of Chapter 18.
11 ll'las

13 J:tb4!?

This is a good, active move. More­

12 ll'lg3 bxc4 over, the alternatives may leave the as­
It can be difficult to decide whether knight looking vulnerable:
to take with the pawn or the knight. a) 13 ...ll'ld7 14 hxg6 fxg6 lS ll'lbl!
Taking with the pawn keeps the offside l:tbs (1s ...ll'lb7 16 ..ixc4 just looks good
knight, but this knight is more likely to for White) 16 b4 cxb3 17 ..ixbs cs 18
participate in an attack than White's dxc6 (18 ..ie2 is also strong) 18 ... axbs
light-squared bishop, which does not was J.Timman-G.Kasparov, Bugojno
have much of a role. Taking with the 1982. Here 19 cxd7 ll'lc4 20 ..igS ! would
knight can still make some sense, be decisive.
though, especially when White has al­ b) 13 ... ..id7 14 ..ih6 (White has an
ready spent a tempo moving his bishop extra option in 14 ll'lbl or 14 hxg6 fxg6
or when the black queen can go to the lS ll'lbl, although the latter is not so
as-square. Here neither of those factors clear after 1s ....rf.bS: for example, 16 b4?
come into consideration, so it is not �8! and White was in some trouble,
surprising that taking with the pawn is V.Rajlich-E.Kahn, Budapest 2000)
correct. 14... ..ixh6 lS �xh6 .l:i.b4 16 l:td2 'ilfe7
Instead after 12 ...ll'lxc4 13 ..ixc4 bxc4 (but not the 16 ... �b8? of A.Harakis­
14 o-o-o ll'ld7 lS ..We2 (lS hxg6 fxg6) E.Kahn, Budapest 2000, because of 17
1s ...ll'lb6 (some correspondence play has ll'lfs! ..ixfs 18 exfs with a winning at­
shown that 1s ...fs 16 hxg6 f4 17 ll'lhs! is tack) 17 ..ie2 leads back to the main
promising for White), both 16 .l:i.h2, from line.

2 70
Pa n n o Va r i a t io n : 9 h4 with o u t 9 . . . h 5

14 ..ih6 ..ixh6 1 5 �xh6 16 ..ie2 ..id7

Instead 16 ... ll'ld7 17 l:tdfl! is given
by Kasparov. White intends to play f3-
f4 with the initiative.
17 ll'lfl
This is a typical regrouping, espe­
cially when it is clear that the knight
will not be hopping into the fS-square.
If White starts instead with 17 :d2,
Black can try 17 ...c6!? (instead 17 ...l:tfb8
18 ll'lfl heads back to the main line) 18
ll'lfl when, instead of 18...'iii>h 8? 19 hxg6
15 .....We7 fxg6 20 �xg6 cxds 21 ll'le3 ! with a big
The alternative is 1s ...'iii>h 8. Black advantage for White in J.Lautier­
does not commit his queen and creates P.Svidler, Internet 2004, Black should
the possibility of chasing the white have played the logical 18 ... l:tfb8.
queen away with ... ll'lg8. After 16 hxg6 11 .. AfbB 18 .lld2
fxg6 17 ..ie2 (not 17 'ilfxg6? l:tg8)
17 .....id7 18 ll'lfl Black has sometimes
played the funny 18 ...'iii> g 8!?. This looks
like a silly waste of time, but Black is
happy to resolve the tension on the
kingside so that he may defend later­
ally along the second rank Thus he
moves the king back so that the g6-
pawn is no longer hanging. After 19
ll'le3 ..ifb8 20 :d2 ..ifb6 the position was
unclear in H.Rauch-J.Stephan, corre­
spondence 1996. Here Black could also White has solidified his castled po­
consider 18 ... l:tg8, although using the sition and is ready to build up on the
rook to protect the g6-pawn looks kingside. Black has:
rather inefficient. After 19 ll'le3 (19 g4 a) 18 ... cs?! (this does not do much
is given as good for White by Djurhuus, for the black cause) 19 ..id1! (after 19
but this is still not so clear) 19 ...'ilff8 20 g4 ..ia4, with the idea of ... �7. Black
�h2 ll'lhs ! 21 g3 �h6 22 'ilff2 �gs 23 can hope to create some counterplay)
:hgl, as in L.Psakhis-R.Djurhuus, 19 ... ll'le8 20 hxg6 fxg6 2 1 g4 'ilf g7 22 gs !
Gausdal 1994, both 23 ... :f8 and and White has the initiative even i n the
23 ...:gb8 give Black decent play. ending. Now 22 ...�xh6 23 :xh6 ll'lg7

2 71
A ttacking C he s s : The King 's I n dian, Vo l u m e 1

24 f4! prevented ...lllh s and after better for White. Matters are not so
24...exf4 2S l:tdh2 White was much bet­ clear, however, following 21 ... lllc s 22
ter in G.Kasparov-V.Loginov, Manila gs ..id7 (to stop lllg 4) when White can­
Olympiad 1992. Thus Black should not regroup so easily because both 23
probably leave the cs-square open for a ..id1? llld 3+ and 23 ..if1? llla b3+ are
knight. good for Black.
b) After 18 ... ..ie8 Kasparov gave 19 c) Another possibility is 18 ... ..ia4 19
g4 llld7 20 hxg6 fxg6 21 ll:le3 as much g4 llld7 20 hxg6 fxg6 21 ll:le3 lllc s.

2 72
Chapter 15
Panrio Vari:ation
9 h4 hS 10 a.o-o

1 d4 ll'lf6 2 c4 g6 3 ll'lc3 .i.g7 4 e4 d6 5 transposes, while 11 0-0-0 is Line A,

f3 o-o 6 .i.e3 ll'lc6 7 ll'lge2 a6 8 1lfd2 l:t b8 below) 11 ...axbs 12 l:tcl .i.d7 makes a
9 h4 hs 10 o-o-o funny impression . Instead of develop­
ing his pieces, White tries to play all
over the board.

This is the main line. White contin­

ues playing for an attack, although he
may prefer to completely change gears Play is likely to continue 13 ll'lxf6+
with 10 ll'lcl, which is covered in the .i.xf6 (or 13 ...exf6!? when Black has
next chapter. counterplay after 14 g4 hxg4 lS fxg4
Before we examine 10 0-0-0, we l:te8 or 14 ds ll'le7 1s ll'ld4 fS; in the lat­
should look at a couple of other moves ter variation, another option is 14... ll'les
which sometimes transpose to the 1s ll'ld4 fS !? 16 exfs .i.xfs 17 ll'lxfs gxfs
main lines: with an unclear position) 14 g4 hxg4
a) 10 ll'lds bS 11 cxbs (alternatively, lS fxg4 (after lS hS es 16 dS gxf3 17
11 ll'lxf6+ .i.xf6 12 cxbs axbs 13 l:tc1 dxc6 fxe2 18 'ilfxe2 .i.e6 19 'ilfg2 gs

2 73
A ttacking C he s s : The King 's I n d i a n, Vo l u m e 1

Black was better in S.Nikolic­ ..if3 ..ixcl 26 ..ixcl b4 with an unclear

M.Cvorovic, Kladovo 1991) 1s 16 dS ending.
ll'ld4 17 ll'lxd4 (Black has compensation b) 10 ..ih6 is fairly common and
after 17 ..ixd4 exd4 18 gs ..ies 19 ll'lxd4 looks more natural than variation 'a'.
l:ta8!?) 17 ... exd4 18 ..ig s (following 18
..ixd4 both 18 ... .l:i.e8 and 18 ... ..ixg4 look
good) 18 ... .l:i.e8 19 �f4.

Here Black can to transpose to some

other lines or explore independent
Perhaps slightly surprisingly con­ bl) 10... es 11 ..ixg7 (instead 11 dS
sidering White's 11th and 12th moves, ..ixh6 12 �xh6 ll'ld4 transposes to
we have come to a theoretical divide: variation 'b3' below, but that is dan­
al) 19 .....ixgs 20 hxgs �e7 21 ..ig2 gerous for Black, so he should consider
and here Black can play either 21 .....Wes 12 ... ll'las 13 ll'lcl cs, as pointed out by

22 �xes (22 o-o 'ilfxf4 23 nxf4 l:tb7 was Golubev) 11 ...@xg7 12 dS! ? (12 o-o-o bS
given by Gallagher) 22 ...l:txes 23 l:txc7 takes play into Line (32, below)
..ixg4 24 l:tfl l:tf8, with an unclear end­ 12 ... ll'le7 (or 12 ... ll'las! ? 13 ll'lg3 cs) 13
ing, or 21 ...l:tbc8!? with the idea of 22 ll'lg3 c6 14 dxc6 ll'lxc6 lS o-o-o ..ie6 16
'it'h2 ..Wes. @b1 ll'le8 (Black could also consider
a2) 19.....ies 20 �f3 and now 20.. .f6 16 ... ll'ld4!? 17 ll'lge2 ll'lc6 18 �xd6 'ilfxd6
2 1 ..id2 (R.Gunawan-Ye Jiangchuan, 19 .l:i.xd6 ..ixc4) 17 ll'lds bS was
Singapore 1987) 21 ... �c8!? 22 ..ie2 d3 B.Spassky-R.Fischer, 8th matchgame,
gives Black counterplay. Black can also Sveti Stefan 1992. White may try to
consider the immediate 20...�c8!? 21 claim some an edge here, but it is not
..ie2 d3! when White must find 22 hS much and in fact Black went on to win.
(worse are 22 ..ixd3 ..ixg4, 22 �xd3 b2) 10 ... bs is the simplest. Now 11
..ixg4 and 22 ..id1 ..ixb2) 22 ... ..ixg4! o-o-o is Line C and 11 ..ixg7 @xg7 12
(but not 22 ...dxe2? 23 hxg6 fxg6 24 o-o-o es is Line C32, which are both
'ifh3) 23 'ilfxg4 'ilfxg4 24 ..ixg4 ..ixb2 2S satisfactory for Black, as we shall see.

2 74
P a n n o Va r iatio n : 9 h4 h5 1 0 o - o - o

Instead 11 g 4 is not really sound after lllxd4 exd4 16 l:txd4 l:txb2.

11 ... hxg4 12 hS gxf3 and then:

If Black does not play this, he will

b21) 13 lllg 3 ..ixh6 14 'ilfxh6 f2+! lS have nothing to show for the pawn.
�dl lllg 4 16 ..ifd2 es is winning for Moreover, the temporary sacrifice looks
Black (Boleslavsky). very appealing, but it does not work
b22) 13 hxg6 fxe2 (or even 13 .. .fxg6) out well against exact play:
14 ..ixe2 (14 ..ixg7 exfl�+ lS @xfl b31) 17 @xb2 'ilfb6+ 18 lllb s axbs
fxg6 should also win for Black) 14.. .fxg6 gave Black good play in V.Khomyakov­
lS ..ixg7 @xg7 16 �h6+ 'i;f7 looks un­ M.Golubev, Ostrava 1992.
sound too. b32) 17 es .ifs 18 ..id3 :xg2 19
b23) 13 ..ixg7 @xg7 14 lllg 3 f2+! ? ..ixfs �as ! 20 ..ife3 l:te8 21 l:td2 'ilfa3+
( l S hxg6 fxg6 1 6 'it'h6+ @f7 is 22 @d1 :xes 23 ..ie4 l:txd2+ 24 'ilfxd2
also good) lS �xf2 es was much better lllxe4 2S fxe4 l:txe4!? 26 lllxe4 �f3+
for Black in E.Cooke-Y.Zimmerman, Bu­ was quickly drawn in R.Meessen­
dapest 2000. M.A hn, European Club Cup, Eupen
b3) 10 ... ..ixh6 11 �xh6 es (instead 2006.
11 ... bs 12 o-o-o es is Line Cl, which is a b33) 17 cs! (Golubev first pointed
tittle risky for Black) 12 dS!? (here out the strength of this move) 17 ...:b8
White can steer things towards the (no better are 17 ....l:i.b7 18 l:txd6! �as
aforementioned Line Cl with 12 o-o-o 19 @c2 or 17 ..ltf2 18 ..ife3 ..Was 19 l:ta4!
bS) 12 ...ll:ld4 (this does not work out ..Wxcs 20 ..Wxcs dxcs 21 llld 1) 18 l:txd6
well, so Black should consider 12 ...llle 7 �as 19 'iii>c2 ..ie6 (19 ...'ilfa3 20 'ilfc1
or 12 ...llla s) 13 o-o-o cs ( 13 ...lllx e2+ 14 should also win) 20 .l:i.xe6 fxe6 (after
.i..xe2 with the idea of g2-g4 looks good 20 .....ifa3 21 l:td6! :b2+ 22 'it>d3 ..Wxcs,
for White) 14 dxc6 bxc6 (no better are 23 'ilff4! is probably clearer than 23 :d4
14... lllxc6 l S cs or 14 ... lllx e2+ l S ..ixe2 'ilfa3) 21 ..ic4 'ilf7 22 es 'ilfb4 23 exf6
bxc6 16 'ilfd2!, as given by Golubev) lS ..Wb2+ 24 @d3 l:tbd8+ 2 s .ids! wins.

2 75
A ttacking C he s s : Th e King 's I n di a n , Vo l u m e 1

We can now return to the main sub­ a) 12 g4 ll'lxds (worse is 12 ...hxg4?!

ject of this chapter: 13 ll'lxf6+ ..ixf6, which is note 'b' to
10 0-0-0 Black's 13th move in our main line) 13
exds ll'lb4 14 ll'lc3 c6 lS ..ixc4 (instead
1s dxc6 ..ie6 16 gxh s ..Was 17 a3 ll'lds
18 h6 ..if6 looks good for Black, while
here 18 ll'lxds 'ilfxds 19 hxg6 c3! is
given by Nunn) 1s...cxds 16 ..ib3 (or 16
ll'lxds ..ie6 17 ll'lxb4 ..ixc4 18 ll'lc6 'ilfc7
19 ll'lxb8 ..ixa2+ 20 ..ifc3 'ilfxb8 when
Black is much better according to
Nunn) 16 .....ifb6 (also good is 16 ...hxg4
17 h S gxf3 with the idea of ....ifs as in
A.Ewald-M.Fischer, German League
Play becomes very sharp, but Black 2oos) 17 :hgl (Nunn gives 17 :dgl fS,
can hold his own in the complications. with the idea of 18 gxh s f4, and here 17 ... hxg4! ? is also possible) 17 ... ..ie6 18
Black initiates his counterplay. Note gxh s .ifs 19 l:tgs ..ih6 20 l:tdg1 ..ixgs
that there are certain similarities with 21 l:txgs e6 22 hxg6 fxg6 23 h s l:tb7 24
the Soltis Variation of the Sicilian ..Wg2 l:tg7 2s h6 l:tb7 26 l:txfs exfs 27
Dragon (1 e4 cs 2 ll'lf3 d6 3 d4 cxd4 4 'ilfxg6+ 'iii>h 8 28 ..Wg2 f4 29 ..ig1 l:tg8 0-1
ll'lxd4 ll'lf6 S ll'lc3 g6 6 ..ie3 ..ig7 7 f3 was A.Kuligowski-J.Nunn, Wijk aan Zee
ll'lc6 8 ..ifd2 o-o 9 ..ic4 ..id7 10 o-o-o lk8 1983.
11 ..ib3 ll'les 12 h4 h s). Here White b) 12 ..ih6 ll'lxds 13 exds ll'lb4 14
must decide how to proceed: ll'lc3 c6 (14 ... ..ixh6!? lS 'ilfxh6 c6 is a
good alternative) looks good for Black
A: ulCds as well.
• :u: lhf4,
:C: 11.i.h6

A) 11 ll'lds
With this move White removes the
f6-knight from the defence of the black
king. However, this costs time and
Black can create counterplay.
11 ... bxc4 12 ll'lxf6+
Consistent. Beating other moves
used to be a specialty of John Nunn:

2 76
P a n n o Va ria t i o n : 9 h4 h 5 1 0 o - o - o

White has tried: ample, 1 3 g 4 hxg4 (or 13. . .ll'lb4 1 4 ll'lc3

bl) lS ..ixc4 cxds 16 ll'lxds ..ixh6 17 cs lS dxcs ..ie6 16 a3 ll'ld3+ 17 ..ixd3
�xh6 ll'lxds 18 ..ixds .ifs 19 ..ie4 ..ixe4 cxd3 18 gxh s) 14 hS gxf3 lS h6 fxe2 16
20 fxe4 'ilfc7+ 21 'iii>b l l:txb2+ 0-1 was ..ixe2 and White has a strong attack.
H.Hurme-J.Nunn, Helsinki 1983. 13 g4 ll'lb4
b2) lS dxc6 was suggested by Nunn, A thematic move. The knight takes
but after 1S .....ixh6! (worse is lS ... ds 16 aim at both d3 and a2, while the rear c­
g4) 16 �xh6 ..ifb6 (with the idea of pawn is freed to advance, which will
...ll'la2) 17 �d2 ..Wxc6 Black has the up­ both strike at the centre and open a
per hand. path for the black queen. other moves
b3) lS g4 �as ! (1s .....ixh6!?) 16 are ri sky but are not necessarily bad:
..ixc4 (both 16 ..ixg7 ll'lxa2+ and 16 a) 13 ... es 14 gxh s ll'lb4 (after
'iii>bl ll'lxds should be winning for Black) 14... exd4 lS ll'lxd4 ll'les 16 hxg6 cs 17
16 ... cxds 17 ..ib3 ..ie6 18 ..ixg7 'iii> x g7 gxf7+ 'iii> h 7 18 ll'lfs? ..ixfs 19 exfs :xb2!
19 'iii>bl (19 gxh S l:tfc8! 20 'iii>bl ll'lxa2! Black won in J.Herbert-V.Walsh, corre­
19 l:tdel l:tfc8 20 'it>d1 ll'lc6 21 gxh s l:tb4 spondence 1989, but 18 f4! looks very
was much better for Black in S.Kitte­ strong) lS ll'lc3 exd4 16 ..ixd4 ..ie6 17
M.Zulfugarli, Szeged 1994) 19 ...l:th8 hxg6 fxg6 18 'ilfh6 ..ixd4 19 'ilfxg6+ ..ig 7
(19 ... hxg4 and ....ifs is winning as well) 20 ..Wxe6+ 'iii>h7 21 ..ixc4 gave White the
20 l:tdgl ll'lc6! (with the idea of ... l:tb4 upper hand in J.Sammour Hasbun­
and ...�6) 21 ..ia4 l2Ja7 22 l:th 2 l:tb4 23 J.Stopa, Pawtucket 2008 .
..ib3 ll'lc6 24 ll'le2 ..ifb6 2s gxhs .ifs+ 26 b) 13 ... hxg4 14 hS is risky but may
'iii>a l l:txb3 27 axb3 ll'lb4 28 ll'lc3 ll'lc2+ be playable:
29 'iii>a2 l:tb8 0-1 was M.Petursson­ bl) 14... es lS hxg6 fxg6 16 ll'lc3
J .Nunn, Lucerne Olympiad 1982. ll'lxd4? 17 ..ixd4 exd4 18 ..ixc4+ l:tf7 19
12 ..ixf6
... ..ixf7+ 'iii>xf7 20 ll'lds ..ig7 21 fxg4 d3 22
l:th 7 �g8 23 ..Wgs 1-0 was Nguyen Anh
Dung-Wang Rui Budapest 2000.
b2) 14 ... gs!? lS ..ixg s es 16 ..ixf6
'ilfxf6 with an unclear position has been
tried with some success.
b3) 14 ... gxf3 ! ? (the 'grab everything'
policy is sometimes seen in similar
structures in the Dragon) lS ll'lc3 gS!
16 ..ixg s ..ixgs 17 'ilfxg s+ 'iii>h 7 was un­
clear in K.Cedikova-D.Vismara, Estensi
2001, although Black did go on to win.
Here 12 ... exf6 looks too risky: for ex- 14 ll'lc3

2 77
A ttacking C h e s s : Th e Kin g 's I n dian, Vo l u m e 1

play according to Soloviov.

c) 18 f4 ll'lc6! 19 �g2 (after 19 dS
l:txb2! 20 'iii> x b2 :b8+ 21 'iii> c2 'ilfxa3 22
l:tb1 ll'lb4+ Black is winning and 19 es?
ll'lxd4 20 ..ixd4 cxd4 2 1 1Wxd4 dxes 22
..Wg1 .ifs 23 ..ixc4+ 'iii> g 7 24 l:.ds 'ilfc7 2 s
J:tcs �b7 gave him a winning attack in
J.Tarjan-R.Gunawan, Indonesia 1983),