Sei sulla pagina 1di 208

Dvoretsky / Yusupov .

Secrets of Creative Thinking


P[2gress inCHess

Volume 26 of the ongoing series

Editorial board
GM Victor Korchnoi
GM Helmut Pfleger
GM Nigel Short
GM Rudolf Teschner

2009
EDITION OlMS

m
Mark Dvoretsky and Artur Yusupov

Secrets of
Creative Thinking

School of Future Champions 5

Edited and translated


by Ken Neat

2009
EDITION OlMS

m
4

Books by the same authors:

Mark Dvore1sky. Artur Vusupov. School of Future Champions


Vol. 1: Secrets of Chess Training ISBN 978-3-283-00515-3 Available

Vol. 2: Secrets of Opening Preparation ISBN 978-3-283-00516-0 Available

Val. 3: Secrets of Endgame Technique ISBN 978-3-283-00517-7 Available

Vol. 4: Secrets of Positional Ploy ISBN 978-3-283-00518-4 Available

Vol. 5: Secrets of Creative Thinking ISBN 978-3-283-00519-1 Available

Mark Dvore1sky. School of Chess Excellence


Vol. 1: Endgame Analysis ISBN 978-3-283-00416-3 Available

Vol. 2: Tactical Ploy ISBN 978-3-283-00417-0 Available

Vol. 3: Strategic Play ISBN 978-3-283-00418-7 Available

Vol. 4: Opening Developments ISBN 978-3-283-00419-4 Available

Bibliographic information published by Die Deutsche Bibliothek

Die Deutsche Bibliothek lists this publication in the Deutsche


Nationalbibliografie; detailed bibliographic data is available in
the internet at http://dnb.ddb.de.

2009 Edition Olms AG

Willikonerstr. 10 . CH-8618 Oe1wil a. S./Zurich


E-mail: info@edition-olms.com
Internet: www.edition-olms.com

All rights reserved. This book is sold subject to the condition that it shall not. by way of trade
or otherwise, be lent. re-sold, hired out or otherwise circulated in any form of binding or
cover other than that in which it is published and without a similar condition including this
condition being imposed on the subsequent purchaser.

Printed in Germany

Editor and translator: Ken Neat

Typeset: Arno Nickel . Edition Marco, D-10551 Berlin

Printed by: Druckerei Friedr. Schmucker GmbH, D-49624 Loningen

Cover: Eva Konig, D-22769 Hamburg

ISBN 978-3-283-00519-3
ftJ 5

Co nte n ts

Preface (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

PART I THE CALCU LATION OF VARIATIONS


The Tech nique of searching for and taking Decisions (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . 8
Wandering through the Labyrinth (Mikhail Krasenkow) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Visual I magi nation and the Calculation of Variations (Beniamin Blumenfeld) . . . . . . . . . . . . . . . 35

PART II INTUITIVE DECISIONS


The Development of Chess I ntuition (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
I n Jazz Style (Sergey Dolmatov) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

PART III PRACTICAL EXP E DIENCY IN THE TAKIN G OF DECISIONS


Practical Chances in a Chess Game (Beniamin Blumenfeld) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Does it pay to sharpen the Play? (Vladimir Vulfson) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Thoug hts about a Book (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

PART IV ATTACK
Missed Brilliancy Prizes (Artur Yusupov) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Long-d istance Dispute (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 33
Attacks with opposite-sided Castl ing (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 38

PART V DEFENCE
Practical Exercises in the Taking of d ifficult Decisions (Igor Belov) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 53
Virtuoso Defence (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 65
What l ies behind a Mistake (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 77

PART VI
Analysis of a Game (Mark Dvoretsky) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 82
Creative Ach ievements of Pupils from the School (Artur Yusupov) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 91

I ndex of Players and Analysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204


I ndex of Openings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
6

Mark Dvoretsky

P reface

Vou now have in your hands the con chess and the ways to overcome the m . To
I clud ing, fifth book in the series School of demonstrate the main d i rections and meth
Future Champions, based on material from ods of chess improvement. And that is a l l .
the Dvoretsky-Yusu pov school for talented Not s o m u c h , b u t also n o t s o l ittle. The
young chess players. pupils' results confirm that this was the
Our small school fu nctioned for only th ree correct approach and that on the whole we
years ( 1 990-1 992 ). Between ten and fifteen solved our objective successfu lly. I nciden
youngsters attended the sessions. Nearly all tally, it was at a session of the school that I
of them began studying with us at the age of advised Peter Svidler to seek Lukin's help.
1 2 -1 5. I can mention with pride that five Our books reflect the same approach . We
years later eight of our pupils became have not tried to write textbooks, with a full
grandmasters - some of them very strong and exact coverage of a particu lar topic.
and world-renowned . Here are their names: The aim was to provide readers with high
Alexey Alexandrov, Vasily Emel i n , I n na qual ity material and a va riety of ideas for
Gaponenko, l Iakha Kadymova , Sergey Mov i ndependent th inking and independent work
sesian , Ella Pitem, Peter Svidler and Vadim in the g iven directio n . Moreover, not only
Zviagintsev. I am sure that in the near futu re our own ideas, but also the ideas of other
Vladimir Baklan and Peter Kiriakov will also experts (in particular, trainers working to
become grandmasters. ( They have! Trans
- gether with us at the school). Clearly, such a
lator. ) Hardly any other junior chess school way of presenting the material demands of
can boast of such a high 'pass rate'. the readers a creative (and at times critical)
I n listing the ach ievements of the school, I attitude to the text being stud ied and is not
nevertheless clearly real ise that the pupils' su itable for those who like ready-made
successes have been forged mainly by the prescriptions. To judge by the popularity of
players themselves and their permanent our books, such an approach su its very
trainers. For example, did we have time to many players.
teach much to the futu re three-times Rus Not all the problems discussed in this book
sian champion Peter Svidler during those are purely chess problems - they lie some
th ree ten-day sessions of the school (the where between chess and psychology.
2nd, 4th and 6th ) in which he participated? Th inking at the board and the ways of taking
Of cou rse, the main components of Svidler's decisions i n a variety of situations - this, in
successes are his enormous talent and the brief, is its main content. Many of the
aid of his splendid trainer Andrey Luki n . examples offered are very complicated and
Yusu pov a n d I saw o u r role as being t o give not straig htforward , and demand a deep
an impetus to the fu rther development of the penetration into the position, ingenu ity, and
young players. To help them to understand bold, risky actions. Therefore, compared
themselves, their virtues and deficiencies, with the preced ing vol umes, the present
and to outl ine plans for the futu re. To book is less instructional and more problem
discuss the problems they encou nter in atic and creative.
Preface lLl 7
The arrangement of the lectu res and arti whereas successful observations and con
cles in the d ifferent parts of the book is to clusions relating to chess playing in general
some extent arbitrary, si nce their topics are retain their value for many years. The reader
closely interconnected . For example, the will be able to see this for himself by reading
discussion of the accu rate and deep calcu two articles by the Soviet master Beniamin
lation of variations i n the fi rst part of the Blumenfeld, a subtle analyst of chess psy
book is merely a prelude, and it will be chology, wh ich were written several decades
continued right to the end of the book. ago. Don't be put off by his writing style,
The calculation of variations is not every which is somewhat archaic by present-day
th ing - during the course of a game a player standards - it is the author's thoug hts that
is obl iged not only to calculate , but also to are most important, and they are still modern .
guess. The problem of developing intuition One of my earlier books School of Chess
has hardly been discussed seriously in Excellence 2 - Tactical Play was devoted to
chess literatu re . I am not a professional problems of attack and defence. But these
psychologist and do not claim to have topics are inexhaustible and I hope that the
written anyth ing scientific, but I hope that fresh material analysed here in appropriate
my practical ideas and recommendations chapters will be useful to you .
on this will prove useful to the readers.
I n the traditional concluding chapter Yusupov
Many players make the serious mistake of analyses some games by pupils from the
devoting all their free time excl usively to the schoo l . In previous books he mainly fo
study of opening theory. After all, errors cused on instructive mistakes, but this time
made in the later stages of play have as the g randmaster decided to demonstrate
much influence on results as poor i n itial some creative achievements by the j u n iors.
organ isation of the game. Specific playing The book concludes with a brill iant game by
deficiencies which , given desire and persist Vadim Zviagintsev, which the experts judged
ence can and should be elimi nated , are to be the best of all those publ ished in
typical of players of any standard . I n order to Informator No.62. It is extremely ra re for
emphasise this idea , the book critically young players to have such an honour
analyses the play not only of young masters conferred on them , since the opin ions of the
and candidate masters, but also of such jury members are strongly influenced by
top-class g rand masters as Artur Yusupov names and titles. I wish our readers the
(he does this h imself in the chapter 'M issed same competitive and creative successes
brilliancy prizes' ) and Garry Kasparov. as those ach ieved by our best pupils. I hope
Opening theory develops very rapidly and that you will be hel ped by the ideas derived
therefore opening books are sometimes out from the books in the series School of
of-date even before they are published , Future Champions.
8

PART I

The Ca l c u l ati o n of Va ri ations

Mark Dvoretsky

T he Tec h n ique of search i n g for a n d tak i n g


Decisions

W chess? We look for promising possi


hat d o we d o throughout a game of example, the question of 'prophylactic think
ing'), and have covered others only in very
bil ities, compare them, calcu late variations, general terms. I am unable to formulate an
endeavour to neutral ise the opponent's accu rate scheme for optimal thinking at the
counterplay, and so on. All this is a creative board (I am sure that in principle it does not
process - here there are no ready-made exist), but I will give you certain pieces of
prescriptions. And yet there are rules and advice which , I hope, will come in useful in
ways of thinking which somehow help us to futu re competitions.
organ ise this process and i ncrease its You should fi rst try to solve the exa mples
reliabil ity, avoid simple mistakes, save think yourself - this will be qu ite good practical
ing time - in short, improve the qual ity of the training and at the same time you will gain a
decisions we take . better feeling for the benefit of using the
Qu ite a lot has been written on this topic. recom mended tech nique for taking deci
For example, in his book Think Like a sions.
Grandmaster Alexander Kotov shared his
The ideas which we will examine can be
ideas on the tech nique of calculating varia
arbitrarily d ivided into two parts:
tions. I would also draw you r attention to an
interesting article by M i khail Krasen kow 1 ) Methods of search ing for a move and
the calculation of variations;
included in the present book, and also the
far-from-obsolete articles by the Soviet 2) Ways of saving time and effort, and
master Beniamin Blumenfeld, a subtle re rational thinking .
searcher into the psychology of chess
th inking . The technique of searching for a
move and the calculation of variations
The problem of contemplating a move has
always interested me. Many of my articles 1. Candidate moves. G rand master Kotov
are devoted to it, and also many chapters in was probably the fi rst to single out this way
earlier books . I have made a detailed study of calculating variations. He recommended
of certain methods of taking decisions (for that you should immediately decide on all
The Technique of searching for and taking Decisions ctJ 9
the possible candidate moves, do this 34 "e4? ':'xb4 with the threat of 35 . . ... b 1 + ,
not only on the first, but also the sub or 3 4 "a7? ':xb4 3 5 ':xg7 ll b 1 + 3 6 c;t>g2
sequent moves, and not only for your fld5+, but 34 'tli'c6 ! is perfectly possible
self, but also for the opponent. If you read (34 . . . .l:l xb4 35 l:[xg7 ! , 34 .. :iIi' b 1 + 35 c;t>g2
the afore-mentioned article by Krasenkow, "xb4 36 "xf6 ! , or 34 . . . d3 35 c;t> g2!? d2 36
you will see that this rule (and , however, this l:tcd7).
also applies to su bsequent ru les) is by no I do not think it is so necessary to calculate
means always appl icable . Nevertheless, for all these variations accu rately - it is suffi
many situations this is very good advice. cient merely to realise that the opponent
Why is it so important to use the 'candidate retains possibil ities of a defence. The point
moves' ru le? Firstly, it helps you to is that Black has one more resou rce
organise rationally the analysis of varia available: simply to advance his passed d
tions, to accurately pick out those con pawn , al lowing l:! xg7, and defend the h7-
tinuations which should be calculated. point with the q ueen from b 1 . This is the
most forcing and therefore the most tempt
Alexa nder - E uwe
ing path - clearly it is the one that should be
considered fi rst of a l l . It is important to verify
Notti ngham 1 936
whether or not the opponent has a per
petual check. If not, then this is what should
be played , since here Black gains an
overwhelming material advantage . And if it
does not prove possible to avoid perpetual
check, then he can retu rn to a more
thorough study of the other conti nuations.
33 . . . d3!
It is important to choose the accu rate move
order. In the event of 33 . . . 'ii' b 1 +? 34 g2 d3
Wh ite acq u i res an additional possi bil ity: 35
1:[cd7! d2 36 J:ixd8 (or even 36 lixg7!? ':xd7
37 l:txg8+ c;t>xg8 38 'ifc8+) 36 . . . l:txd8 37
'ifd7 ! , lead ing to a draw. But now 34 ':'cd7 is
pointless, since after 34 . . . d2 the pawn
queens with check.
Black is two pawns up, but the enemy
34':xg7 1:txg7
pieces are very active. The g7-pawn is
attacked . If he defends it by 33 . . . 'ii' b 1 + 34 35':xg7 'ili'b1 +
c;t>g2 'ifg6, then after 35 ];tcd7 d3 36 'ifd5 The next move will be 36 . . . d2. It is here that
(36 'ii'x a6 !?) Wh ite regains the d-pawn and one must concentrate on determining the
soon succeeds in transposing into a drawn candidate moves . Wh ite has two ways of
ending with three pawns agai nst two on one conti nuing his attack: 37 :g4 (th reatening
wing . 33 .. :if d5 34 'ili'xd5 1:[xd5 35 l:[ed7 has mate on g7) and 37 'iff? (with the idea of 38
roughly similar conseq uences. The attempt ':g8+ or 38 lixh7+). I n addition , his king can
to play for an attack with 33 . . . .l:.d5 (hoping move to either g2 or h2. Four branches are
for 34 ':xg7?l:!xg7 35 l:!xg7 'ifb 1 + 36 c;t>g2 obtained , and each of them must be
':h5) is refuted by 34 %:tc8 ! . Finally, after calculated before 33 . . . d3 is played .
33 . . Jlb8 it is unfavourable for Wh ite to reply We will beg in with the queen move to fl.
10 The Tech n ique of searching for and taking Decisions

a) 36 h2 d2 37 'iff7 'iff5 ! . Now 38 l:txh7+ is 39 'iff7 d 1 'if 40 'ii'xf6+ h7 4 1 'ii'e 7+ (after


not possible, since the rook is captu red with 41 'iff7+ h6 42 'ii'f4+ g7 43 'ilfe5+ Black
check, while 38 l:tg4 is met by 38 . . . 'ii'xf2+ 39 has both 43 . . . h7 44 'ili'e7+ h6 - cf.
h3 'iff1 + 40 h2 'ifh 1 + ! (or 40 . . . 'ii'e 2+ 4 1 below, and 43 . . . f7 44 'ikf4+ g8 45 'ii' c4+
h 3 'ifxg4+ ! ) . 'ifd5) 4 1 . . . h6 42 'ii' h 4+ (42 'ilfe3+ h5! 43
b) 3 6 g 2 d2 37 'iff7 . Now 3 7 . . . 'iff5? is bad : 'ife5+ g5, and the checks come to an end , or
38 l:[xh7+! 'ifxh7 39 'ifxf6+ 'ifg7 40 'ifxd8+ 43 'ife7 'ifd4! ) 42 . . . 'ifh5, and the rook
and 41 'ifxd2 with two extra pawns for cannot be taken because the queen is
Wh ite , while 37 . . . d 1 'if? 38 :g8+! ':'xg8 39 pinned. Now it is clear that after 36 g2 d2
'ifxf6+ leads to perpetual check. Black wins 37 .1:.g4 the reply 37 . . . 'ili'g6? is insufficient for
by interposing the check 37 . . . 'ii'e 4+ ! . I n the a win - only 37 . . .'tIi' h 1 + ! is correct.
event of 38 h2 'iff5 we transpose into the It remains to add that in the game after 36
previous variation . If 38 f3 the simplest is h2 d2 Wh ite resig ned .
38 . . . 'ifxf3+! 39 xf3 d 1 'if+ with a rapid Thus by determining the candidate moves
mate, although it is also possible to play beforehand we ensure that our calcu lation
38 . . . 'ii'e 2+ 39 h3 'ili'f1 + 40 h2 'ii' h 1 + ! 4 1 of variations is accu rate and rel iable. But
xh 1 d 1 'ii' + 4 2 h2 l:td2+ (or 4 2 . . . 'ilfd2+ 43 the 'search function ' of this proced ure is
g 1 'ife3+ 44 g2 l:.d2+ 45 h3 'ifh6+) 43 even more important. It enables a typical
h3 'ifh 1 + 44 g4 h5+! 45 f4 .:td4+ 46 mistake to be avoided , one wh ich is repeat
e3 'ii'g 1 + . edly made by nearly all players - delving
Now let us examine the rook move to g4. immed iately into the calculation of those
c) 36 g2 d2 37 :g4 'ii' h 1 + ! 38 xh 1 d 1 'ii' + conti nuations which first come to mind. I n
and 39 . . . 'ilfxg4. this case some strong possibil ities may be
d ) 36 h2 d2 37 ':'g4 missed , resulting in a mass of time and
effort being spent in vai n . By concentrat
ing on a search for all the available
candidate moves, we sometimes find
resources, the existence of which we
initially did not even suspect.

Najdorf - Kotov
Mar del Plata 1 957

The h 1 -sq uare is controlled by the wh ite


queen, and Black fails to win with 37 . . . 'ifg 1 +?
38 h 3 ! 'iff1 + 39 h4 . The only possibil ity
is 37 . . . 'ifg6! 38 l:[xg6 hxg6. Let us verify
whether White can g ive perpetual check
with his lone queen .
The Tech n ique of searching for and ta king Decisions 11

It is immed iately apparent that the h7-pawn It turns out that he has two more ways of
can be captured with check: 21 xf6 xf6 conducting the attack:
22 'ii'x h7+ 'itf8 . Here there is noth ing to a ) 21 d 1 (with the idea of 22 ..i h5);
calculate - it is a matter of assessing the b) 21 ..i c2 (with the threat of 22 ..ixh7+ and
resulting position . It is not possible to give the key variation 2 1 . . . l:txc2 22 ..ixf6 ..ixf6 23
mate (the bishop on f6 securely defends the
'iWxh7+ and 24 'iVxc2).
kingside), and Black retains some positional
The second way is more forcing, and is the
compensation for the lost pawn in view of
one which must be checked in the fi rst
his control of the c-file and the weakness of
instance .
the d4-pawn .
Another, more tempting continuation , is 2 1 2 1..ic2 ! ! :xc2
ttJ g4 . We easily find the variation 2 1 . . . xb3? 21 . . . g6 22 ..i xf6 and 2 1 . . . h6 22 ..ixh6 are
22 ttJ xf6+ xf6 23 'iVxh7+ 'itf8 24 'iWh8+! both bad for Black, while if 2 1 . . . 'itf8 , then
e7 25 'iVxg7 and wins. If 21 . . . h6?! there either 22 ..i h6! or 22 ..ixh7 ttJ xh7 23 'ii' h 5! is
follows 22 ttJ xh6+, and the captu re of the decisive.
knight leads to mate (22 . . . 'itf8 ! is more 22 ..ixf& h&
tenacious, although after 23 xd5 'iVxd 5 24 2 3 'iWh 5 ! ..ixf&
ttJ g4 ttJ g8 25 ttJ e5 Black's position is
23 . . . :f8 2 4 ..i xg7 .
difficult).
24 'ili'x f7+ 'it'h7
But we will not jump to conclusions -
candidate moves should be sought not only 24 . . . 'it h8 25 11xh6+ ! gxh6 26 ttJ g6 mate .
for ourself, but also the opponent, and this 25 .l:r.xh &+! xh &
mea ns we must check whether we have 2 & 'iii' g & mate
taken all the defensive resou rces into
account. We find the only defence: 2 1 . . . 'itf8! . The winning combi nation (pointed out by
The h7-pawn can b e captu red in various Igor Zaitsev) immediately resolves the ques
ways , but noth ing is completely clear. For tion about the strongest conti nuation of the
example, in the variation 22 ttJ xf6 ..ixf6 23 attack, and none of the remaining conti nua
..ixf6 'ii'xf6 24 'iVxf6 gxf6 25 ..ixd5 exd5 26 tions needs to be analysed . You see that it
.l:[xh7 'itg8 the activity of the black rooks on is important not only to determine the
the open c- and e-files is a concern . complete list of candidate moves, but
also to establish the optimum order in
Of course, no one has given us a guarantee
which they are considered.
that we can achieve more than the win of a
pawn . The knight move to g4 looks very It would have been much simpler to fi nd the
strong, especially if we notice the possibil ity combi nation if there had been an obvious
after 21 . . . 'itf8 of conti nuing the attack by 22 lack of promising possibilities for Wh ite . But
i. h6!? (however, it is sti ll an open question in the g iven instance there were such
whether it is possible to checkmate the possi bil ities, and they immed iately drew our
opponent in the variation 22 . . . ttJ xg4 23 attention . In such conditions, even if you are
..ixg7+ 'itxg7 24 'iWxh7+ 'itf6 25 'ili'h4+ 'itf5). an excellent tactician, it is easy to miss the
Nevertheless, after a slight delay (perfectly move 21 ..ic2 ! ! . A well-developed search
excusable - the moves 21 ..ixf6 and technique ('candidate moves') significantly
especially 21 ttJ g4 are really too tempting) improves our chances of success.
let us remember about the 'candidate But no technique will save a player if he
moves' principle and look for new possibili does not possess sharp combinative
ties for Wh ite . vision. This quality must be trained and
12 The Tech nique of searching for and ta king Decisions

developed, by regularly solving appro Fortunately for me, my opponent made his
priate exercises. move without checki ng the variations.
I n the game Miguel Najdorf played the 16 . . . ttJe4 ?
weaker 2 1 i.d 1?1. The opponent could It is clear that the exchange h a s to be
have parried Wh ite's threat by playing his sacrificed ( 1 7 ttJ 4f3? ttJ xd2 1 8 ttJ xd2 dxc4 is
king to a safer square: 21 . . . f8 ! (22 i.h5 completely bad), but in what way? If a list of
ttJ e4! ) , or by defending the weak f7 -point
all the candidate possibilities is established ,
beforehand with 21 .. J:tc 7! (22 i.h5?! ttJ xh5 it is not at all difficu lt to find the strongest of
23 'iYxh5? i.xg5). But Kotov carelessly them :
repl ied 2 1... iia5 ?, and after 22 i.h 5 ! The
a ) 1 7 l:.xe4 dxe4 1 8 ttJ xe4;
attack became irresistible. There followed
22 . . . l:.ed 8 (22 . . . ttJ xh5 23 'it'xh5; 22 . . . l:.f8 23
b) 1 7 ttJ xe4 i.xe 1 1 8 'iVxe 1 dxe4 1 9 iixe4;
i.xf6 i.xf6 24 i.xf7+ l:txf7 25 iixh7+) 23 c) 1 7 ttJ xe4 i.xe 1 1 9 cxd5.
i.x f7+ f8 24 i.h6 ! ttJe 8 25 iif4 i.f6 26 I n the fi rst two cases one can speak only
i.xg7 + e7 27 i.xe 8 i.xg7 28 lIxh7 Black about some compensation for the lost
resig ned . exchange, but in the third case a double
edged position arises and it is not clear to
Dvorets ky - Butnoryus
which side preference should be give n .

Dubna 1 970 A s you see, sometimes w e look n o t for


candidate moves, but for candidate pos
sibilities - short variations, the first
moves of which may coincide.
17 ttJxe4! i.xe 1
1 8 cxd5!
The wh ite pieces dominate in the centre,
and in add ition the d5-pawn prevents the
normal development of the knight at b8.
18 . . . i.b4
1 8 . . . i.a5!?
19 ttJf3 ! l:.e 8
2 0 iid4 i.f8
What does Black want now? Of course, to
bring out his knight to d 7 . Can this be
I had played the opening stage badly and prevented?
now Black could have gained an excellent 2 1 ttJe5 ! iib6?!
position with the simple 1 6 . . . dxc4 ! 1 7 bxc4
ttJ c6, forcing 1 8 ttJ 4b3 or 1 8 ttJ 4f3 ( 1 8 2 1 . . . ttJ d7? 22 ttJ xd7 iixd7 23 ttJ f6+! would
ttJ xc6? i s bad : 1 8 . . . bxc6 1 9 l:.e2 'iVd3 with have lost immed iately. 21 .. .f5 came into
consideration, but after 22 ttJ g3 'iVf6 23 f4
the th reats of 20 . . . 'it'xh3 and 20 .. .l:I d8).
Wh ite would have retained the advantage .
However, this is not yet a win, but 'merely'
an excellent position . Black can try for more 22 'iVd3
by playing his knight to e4 , but then he is There is noth ing that Black can move . If
beh ind in development, and therefore the 22 . . . ttJ a6, then 23 ttJ d7 and 24 ttJ f6+ is
consequences must be carefully calculated . decisive .
The Tech n ique of searching for and ta king Decisions ttJ 13

It should be said that, despite the obvious twenty-five m i n utes , I worked out a winning
virtues of my position, my mood was far combination .
from optimistic. At the tournament of young I ncidentally, i t was only one of two possible
masters in Dubna I was playing terribly combi nations. The other, perhaps even
badly, making constant oversig hts, which more spectacular one: 23 ttJc4! "dS 24
were explained mainly by a lack of self ttJxf6+ gxf6 25 Wxg6+ ..t.g7 26 ..t.xf6 'ifxf6
control and a q u ite u njustified haste i n the 27 .xeS+ 'iifS 2S l:te 1 with four pawns for
taking of decisions. In the previous game the piece , in fact remai ned unnoticed . So
with one hasty move I had thrown away a that the candidate moves were neverthe
practically winning position , and in the less not determined as well as possible.
present game I had messed u p the opening.
23 d6! l:txe5?
It was pure l uck that my opponent had so
cheaply conceded the in itiative . If 23 . . .fxe5 I was intending 24 ttJf6+! (no win
is apparent after 24 ttJg5 ttJd7) 24 . . . gxf6
'Yes' , I thought, ' I stand wel l , of cou rse, but
(24 . Af7? 25 ttJxeS xeS 26 Wxg6+ dS
I'll probably have a 'fit' and make some
27 .g5+ ! ) 25 .xg6+ hS 26 WxeS Wxd6
blunder. Alright, whatever he plays, on my
27 l:tc1 We7 2S WcS . Black is tied hand and
reply I will spend at least five m i n utes ! I have
foot, but it is not easy for Wh ite to make
an enormous reserve of time, and I should
progress. After 2S . . . gS (defending against
make use of it. '
g4-g5) he does not have 29 ':c7? because
After a long think Alg is Butnoryus played : of 29 . . . ttJa6. Even so, objectively Black's
22 . . . f6 position remains d ifficult, as is shown by the
fol lowing variation : 29 l:td 1 a5 (what else?)
30 ..t.c1 ! (of course, not 30 ':dS? ttJc6 ! )
3 0 . . .c 5 (the th reat was 3 1 ..t. h6 a n d 32
..t.xfS ; if 30 . . . f7 Wh ite has both 31 ..t.a3!?
ttJc6! 32 .xaS .xa3 33 'iix b7+ ttJe7 34
.e4 with advantage, and also the unhurried
31 ..t.e3 ! ) 31 We6+ g7 32 ..t. h6+! xh6 33
'iixf6+ h7 34 .f7+ ..t.g7 (34 . . . hS 35
:dS ; 34 . . . h6 35 h4) 35 .f5+ gS 36
UdS+ ..t.fS 37 "g6+ hS 3S 'ili'f7 .
24..t.xe5 fxe5
25 'ili'c4 + h7
2 6 .c 8 .d4
26 . . . ..t.xd6 27 ttJg5+ with mate .
I even felt u pset. ' It's all clear: I take on g6, 27 ttJg 5 +
he develops his knight on d7, and then I can I n h i s joy Wh ite also conceives a rook
have a think. But now, why do I need to sacrifice. And although this leads to a forced
spend these five m i n utes?' But there was mate , in principle such a 'combi nation for
noth ing to be done: I 'd g iven my word . the sake of a combination' (an expression of
In order not to be bored, I began analysing grandmaster Vladimir Pavlovich Simag i n )
other possibil ities apart from 23 ttJxg6 hardly meets with approva l . After all, the
(there you are - candidate moves! ) . And as elementary 27 l:.e1 (with the th reats of .xfS
a result, after th inking not for five but for and 'ili'xb7) would have forced Black to
14 The Technique of searching for and taking Decisions

resig n . The best way to the goal is always A. Wotawa


the simplest way! 1 938
27 . . . h6
2 8lDf7+ h7
2 9 Wxf8 'it'xa 1+
3 0 g2 lDc6
3 1 Wxa 8 g5
32 lDxg5 + g6
33 We 8+1
Black resigned .

2 . What might I not have seen? Some


times during the calculati ng process a
search again has to be made for candidate
moves (someti mes we simply forget to
determine them at the right time, and also
there is not always confidence that the l ist of Yusupov thought for a long time, unsuc
possibil ities found is exhaustive). If the cessfully trying to fi nd a saving line for Wh ite
variations are not working out in your favour, in the labyrinths of the d ifficult rook end
it makes sense to go back and ask yourself: game.
'What else can there be in the position ; what 'Stop calculating!' I said to h i m finally. ' Look
idea might I not have noticed?' A similar at the position and think what you might not
additional check, but this time search ing for have seen here . '
resou rces for the opponent, is worth carry Artur i m med iately found the solutio n .
ing out if the prospects, by contrast, seem fxe4
1 e4 11
too rosy. Try as often as possible to cast
2 g7 lIh5
off the burden of variations calculated
earlier and look at the position with new 3 g6 :e5
eyes . Such a method often gives excel 4 f6 lle 8
lent results . 5 f7
There is another aspect to this rule. Don 't The king makes a perpetual attack on the
be in a hurry to go too deeply into your rook.
calculations . If problems arise, demand
ing a deep verification, don 't be in a 3 . Should you check your calculations?
hurry to start this . First ask yourself how Another of Kotov's principles - that you
essential it is, and whether it is possible should go through each of the branches of
to improve your earlier play or that of the the 'calculation tree' only once - is dubious,
opponent. New ideas at the start of a in my view. This is possible only for the
variation are far more important than purely tech n ical checking of variations, but
subtleties at the end, which have far less we are not only calculating them, but also at
of an influence on the play. the same time looki ng for the strongest
I once invited Artur Yusupov to try and solve moves. They do not always come to mind
the fol lowing study. immed iately, and sometimes this is i n
principle not possible, without a prel iminary
The Technique of search ing for and ta king Decisions ctJ 15

analysis of the positio n . Suppose that the We will fi nd the solution if we think about 3
variations do not work because of some l:[xc6 (instead of 3 b8) i n the last variation.
deta i l , and here it dawns on us that an U nfortu nately, it does not work, but the idea
apparently poi ntless i ntermediate move can ca n be improved.
be included , adding this deta i l . 1l::td 5 +! ! f5
We already know that this is the only move .
2 l::td 1! ..txa4
F. B o n dare n k o AI. K u zn etsov
,

Here too , as we established in our prelimi


1 977 nary calculation , Black has no choice - if the
a4-pawn is left alive, Wh ite easily wins the
bishop endgame.
3 l::tc 1 ..tc6 +
4 .:txc6! h 1'it'
5 ..tf7+ g5
6 f4+1 gxf3
7 l::tg 6 + h5
8l::tg 8+ h6
9 l:th 8+
Thus it is not often that one consistently has
to calculate one variation after another right
to the end. I recommend another order of
action. A fter deciding on the range of
candidate moves, first make a rapid
How to stop the enemy pawn? I n the event
appraisal - check them supe rficial/y. The
of 1 ..td5 ..txd7 2 a5 Black has time to play
preliminary conclusions will almost cer
his bishop to f3 : 2 . . . ..tb5 3 b7 ..te2 4 a6
tainly come in useful in the subsequent
i.f3 5 ..txf3 gxf3 6 a7 h 1 'it' with the
calculation . Possibly you will be able to
advantage. I nterposing the check 1 l::t d 5+
assess how promising this or that con
(with the idea of 1 . . . g6? 2 l::t d 6+ and 3
tinuation is, and establish a rational
i.d5) will be met by 1 . . .f5! 2 l:bf5+ g6. For
order of the subsequent analysis. Per
example: 3 lIh5 xh5 4 ..td5 ..txa4 fol
haps (as, for example, i n the Najdorf-Kotov
lowed by the bishop manoeuvre to f3 , or 3
game) one move will prove so strong,
i.d5 xf5 4 a5 e5 5 ..tb7 ..tf7 6 a6 ..td5,
that the others will simply not have to be
and Black wins.
calculated.
There only remains 1 lid 1 , but then 1 ... ..txa4!
2 I1c1 ..tc6+ 3 b8 h 1 'it' with a drawn 4. Reg ister the res ults of your ca lcula
bishop ending. tions , and end the variations with a
This is apparently the best available to defi n ite concl usion. Sometimes an abso
Wh ite , but only apparently. Let's not be in a lutely clear concl usion is needed , such as
hurry to agree a d raw, but try to devise we made when analysing the move 33 . . . d3!
someth ing. Here, of course, we have to use in the Alexander-Euwe game. There the
our imagination, but our accurately per exact result (wi n or draw) had to be
formed preparatory calculating work will establ ished - term inating the calculation
also be used . half way with the conclusion 'unclear' would
16 The Technique of searching for and taking Decisions

have prevented the correct decision from or b7. For this he must first defend his knight
being reached . by . . . b6-b5.
But an exact evaluation is by no means I n this way we find the solution of the
always req uired. For example, you come to position for Wh ite.
the concl usion that a position is reached by 21 a4!
force, but it is d ifficult to evaluate - Th reatening, among other things, 22 .tf1 . If
additional calculation is required . If it will be 21 ... b5 there follows 22 .tf1 ! (22 axb5
necessary, you ca n carry it out later, directly .txb5 23 .tf1 is also not bad ) 22 . . . .tc8
from the critical position , without repeating (22 . . . lIc8 loses after 23 liJd4 liJxe5 24 f4
work that has been done earlier. This is the liJxg4 25 l:ixe7) 23 'iif4 a6 24 .txc4 bxc4
point of reg istering in your mind the conclu 25 'iix c4, and Wh ite emerges a pawn up.
sions on variations that have already been
The question arises , is it not possible to
stud ied .
reach the same position by playing 2 1 .tf1
(and if 21 . . . b5, then 22 a4)? Wh ich move
5. Prophylactic thinking. Often it is order is more accurate? Here we have to
useful to begin considering a position concentrate on a search for resources for
with the question: 'What does the oppo the opponent. We will probably g ive prefer
nent want; what would he play if it were ence to the pawn move after in reply to 2 1
him to move ? ' .tf1 we d iscover the unexpected sortie
Readers who are familiar with my earl ier 21 . . . 'ii a 3!.
books will probably need no convincing I n the game Konstantin Lerner did not think
about the exceptional value of the abil ity to about prophylaxis, and simply played 21
th ink prophylactically. Even so, I will g ive lIad 1 ? ! . His opponent replied 2 1 . . . b5! , not
one more example. fearing 22 lId7 .tc8 ! 23 ':xe7 (23 e6 .txd7
24 exd7 'iix e 1 + 25 liJxe 1 lIxe 1 + 26 .tf1
Lerner - Lukin lId8) 23 . . . .txg4 24 lIxa7 .txf3 25 .txf3
USSR 1 977 liJxe5 with approximate equality.

6. What is the drawback to the oppo


nent's move? If the opponent makes an
unexpected move which is uncomfort
able for you, ask yourself this question.
A logical answer sometimes helps you to
understand where to look for a chink in
the opponent's idea, and how best to
combat it.

Ciocaltea - Liberzon
Netanya 1 983

What does Black want? The advance of the


f-pawn will only weaken his position . It is far
better to include his bishop in the play via c8 (see diagram)
The Technique of search ing for and taking Decisions 17

Threatening 1 9 'iWh5 and 20 lLlg6 mate.


18 . . . g 6 ?1
1 8 . . . l1f6 is better. Now the a 1 -h8 d iagonal is
weakened . How can White exploit this?
19 c4 1 d4
1 9 . . . dxc4? is not possible on account of 20
xc6. Having forced the opponent to seal
the queenside, White has freed his hands for
active play on the kingside. His advantage is
now undisputed .
2 0 'ilke2 lLlb6 21 b3 %1be 8 22 'iWf2 lLlc 8 23
f3 :xe1 24 :xe1 :e 8 25 ':'xe 8 'ii'xe 8 26
g4! lLld6 27 gxf5 lLlxf5 (27 . . . gxf5 came into
16 . . . :a b 8 consideration , intend ing 28 'ii'g 2 lLlb4! 29
Let u s apply 'prophylactic th inking' a n d ask 'ii' h 3 lLlxd3 30 lLlxf5 lLlxf5 3 1 'ii'xf5 'iWg6 32
ourselves what Black wants. Obviously, to 'ii'g 4 'ii'f6 ) 28 e4 lLle3 ? (28 . . . lLlce7 was
play . . . b7-b5 and press on the b2-pawn. more tenacious) 29 xc6! bxc6 30 lLle4
How can th is be prevented? lLlg4 ?! 3 1 'ii'g 2 lLlh6 32 'ii'g5 'ii'f8 33 'ii'xc5
17 l1c11 'iVxc5 34 lLlxc5 Black resigned .

Now if 1 7 . . . b5 there follows 1 8 axb5 .l:txb5 1 9


7. W hat do I wa n t to ach ieve? Also a
:c2 . Then the bishop will go to c1 , securely
usefu l question. Clarify your aims: do you
defending the pawn , and the rook will switch
want to exchange a couple of pieces,
to the open file, to e2. An excellent idea!
seize an important square, prevent some
Let us now consider for Black what may be a
active possibility of the opponent, or
minus featu re of the opponent's plan. With
something else? A logical analysis of the
the rook on c2 there is . . . d5-d4 with the
position may suggest the direction for
th reat of . . . b3. However, Wh ite replies c3-
further calculation.
c4; the sealing of the queenside is probably
to his advantage.
Kotkov - Dvore tsky
There is another drawback: after :c2 the
a4-pawn is vulnerable. How can this factor Moscow Championship 1 972
be exploited? If we refrain from . . . b7-b5, the
rook is doing nothing on b8. 1 7 . . . l:tbe8
suggests itself, and if 1 8 :c2 - 1 8 . . . lLld8 ! .
And i f Wh ite tries 1 8 e 3 b 6 1 9 l:lc2?, then
our first idea goes into operation: 1 9 . . . d4!
(with gain of tempo) and then . . . b3.
17 . . . lLlc 8?1
Another way of attacking a4 ( 1 8 :c2 lLlb6),
but a less successful one. What is the
drawback of Black's move? The knight
moves away from the kingside, where it was
needed for defence.
18lLlg51
18 The Tech nique of searching for and taking Decisions

Black is a sound pawn to the good . It is 60 . . . tDa5 !


tempting to beg in active play immediately 6 1 e7 tDc4 +
with 54 .. .1:[a3+. This move is probably good 62 c;t>e2 tDd6
enough to win , but even so it seemed
This entire episode is also instructive as
un method ical to me. The black king is stuck
regards the tech nique of converti ng an
on the edge of the board and for the moment
advantage. It is clearly more comfortable
is not taking part in the play. Here is a sample
playing with the king on b7. Since the
(although, of cou rse , not forced) variation , in
opponent does not have the right to ex
which this factor tells: 55 c;t>d2 tDxd4? 56
change rooks, by choosing 54 .. J:tb5 Black
tDxd4 l1d3+ 57 c;t> c2 l:txd4 58 1:Ie1 ! , and it is
makes life easier for hi mself. For the sake of
Wh ite who wins.
this it is worth exerting yourself and calculat
Black would like first to bring his king closer, ing a forcing variation , especially if you are
so that if necessa ry it ca n su pport the c able to do this qu ickly and accu rately (the
pawn or stop the enemy e-pawn . Therefore I calculation took me about th ree min utes). I
began checking the exchange of rooks . think that now the followi ng idea will become
54 . . . l:tb5 ! clearer: good technique is largely based
If 55 :a 1 +, then 55 . . . c;t> b 7 - the king has on concise and accurate tactical play.
approached the centre, whereas Black will It only remains to show how the game
always have the check along the 3rd rank. concluded .
The only question is what happens if the 54 . . . l:tb5 1 55 lIa 1+ c;t>b7 56 c;t>f2 l:tb2 57
opponent exchanges on b5 and then cap c;t>e3 Itb3 + 5 8 c;t>f2 lIb2 59 c;t>e3 l::tb 3 + (the
tures the d5-pawn with his knight. Try to fi nd sealed move).
the answer. Moreover, this answer should
Another element i n the conversion tech
be convincing and comparatively straig ht
nique - in winning positions you should not
forward . The initial position is too good for
force events before the adjournment of the
Black, for h i m to afford the slig htest risk.
game. However, with the switch ing to time
55 1:ixb5 c;t>xb5 controls without adjournments, natu rally
56 tDc3 + c;t>b4 this rule has lost its sign ificance.
57 tDxd5+ c;t>b3 6 0 'ifilf2 h4 61 e 6 c;t>c7 (th is is where the
5 8 e6 conseq uences of Black's 54th move are
I n the event of 58 'itd2 tDxd4 the outcome is felt! ) 62 ]::ta 8 h3 63 l:tg 8 tDe7 64 lIh 8 c;t>d6
obvious . Now 58 . . . tDe7!? is possible, but 65 tDg 1 c3 66 c;t>e3 c2 + (another way was
first it is better to examine a more forcing 66 . . . l:t b 1 67 tDe2 h2 68 :xh2 c2 69 c;t>d2
continuation - the win of the knig ht. c1 'i1i'+ 70 tDxc1 ltb2+) 67 c;t>d2 l:tg3 ! (but not
58 . . . c3 67 . . . l:t b 1 ? 68 tDe2 h2 69 tDc1 ! ) Wh ite
resigned .
59 tDxc3 c;t>xc3
60 d5
Let us now tu rn to another, no less well
After qu ickly reaching this position in my
stud ied aspect of the tech nique of ta king
calcu lations, I experienced a moment of
decisions.
fright, since I couldn't see how to stop the
pawns. 'But where should my knight go to?'
Of course, to d6, and if possible with gain of
tempo. The route for it immed iately became
clea r.
The Tech n ique of searching for and taking Decisions CZJ 19

Principles of rational, economic weakest replies (again remember the


thinking Alexander-Euwe game: this was the order
i n which we operated , when studying the
First, the most general ru le. When you are
position after Black's 35th move). After
thinking about your move, your ob jec
qu ickly checking the simplest variations and
tive is not to calculate all the varia tions
thereby restricti ng the extent of the calcula
to the end and obtain an exhaustive
tion , perhaps red ucing it to just one or two
impression of the position. You have
directions, it is then psychologically easier
only one ob jective: to take the correct
to concentrate all your efforts on them. You
decision, to make the best move. Try, as
will already know for sure that the opponent
far as possible, to minimise the expendi
has noth ing else.
ture of time and energy. You should
calculate only the minimum number of But if you suspect that the combination
varia tions needed for the taking of the does not work, immediately concentrate
correct decision. on the best defence. If it refutes the
combi natio n , this is sufficient, and the
How can this be ach ieved? Here are a few
opponent's other resou rces do not have to
considerations.
be studied.
Obviously, these recommendations are not
1. With what to begi n the calculation? In set in stone. Va rious situations arise, and
Krasen kow's article you will fi nd some sometimes you have to act q uite d ifferently
interesti ng ideas about this, although some in them . But as a guideline, applicable to the
times they d iffer from my recommendations. majority of cases, these pieces of advice
You have the right to choose those which make sense.
seem to you to be more correct.
If there is a tempting continuation , forcing 2 . 'Emergency exit'. H ere I will not give any
the play, then, of course, it makes sense to examples, but will simply explain the idea.
beg in your calculation with it (as, for Su ppose that a complicated position has
example, in the analysis of the Alexa nder arisen , and it is unclear whom it favours.
Euwe game). Generally speaking, it is After beg i n n i ng the calculation of a compli
a dvisable to immediately begin consid cated combi natio n , you notice that at some
ering forcing moves, such as exchanges, point you can , if you wish, force perpetual
win of ma terial or, on the contrary, check, or, say, an equal endgame. You have
sacrifices. It is often easier to calculate a the right to stop here and decide - very well ,
specific variation than to assess the I ' l l go in for the combi nation! I have a
consequences of a quiet continuation. perpetual check, and there are also some
I was once chatting with M i khail Kats , a attacking possibil ities. I will not bother to
famous draug hts trainer, who had prepared ca lculate them - this ca n be done later, and
a whole constellation of lady world champi at the worst I will force a draw.
ons. He told me that he trains his pupils in The 'emergency exit' , enabling you to avoid
the fi rst insta nce to check all possible piece the prel iminary calculation of the most
sacrifices. As a result, it is extremely rare for critical conti nuation, does not necessarily
them to miss unexpected combinations. have to be a forced draw. It is sufficient to
If you sense tha t a combinative idea you notice for yourself the possibil ity at some
have foun d is probably correct, it makes point of making a move, even a second-rate
sense first to examine the opponent's one, but one which according to your
20 \it The Technique of searching for and ta king Decisions

assessment is nevertheless qu ite accept of imperceptible inaccu racies.


able. A few moves later it will be far easier This is what happened , for example, in the
for you to decide whether to go in for the very important game Yusupov-Zapata ,
main variation , or satisfy you rself with the played in the 1 977 World J u n ior Champion
'minimum wage'. ship in I n nsbruck. The Col umbian Alonso
Zapata played splendidly in the second half
3. The method of elimi nation. Often i t is of the tournament (6 poi nts out of 7) and
not necessary to make an accurate won the si lver medal. The only defeat that
calculation of the intended continuation; he suffered towards the fi nish was against
it is sufficient merely to sa tis fy yourself the future champion . But see how easily
that it makes sense, it cannot be immedi Yusupov gained this win .
ately refuted, the remaining moves are 4 . . . d 6 S d 4 0-0 6 g2 liJbd7 (if Black i s
bad and all the same you have nothing intending . . . e7-e5, i t is better to play this
better. In this way you can sometimes immediately, and to answer 7 dxe5 with
save a mass of time and energy. Players 7 . . . liJfd7) 7 0-0 eS?! (7 .. J:teS!?; 7 . . . c6 !?) 8
sometimes forget about this method of dxe S liJg4 9 liJc3 dxeS (9 . . . liJgxe5 is better)
choosing a move or do not use it properly. 1 0 liJd2 :e 8 (if 1 0 .. .f5 there would have
followed 1 1 e4 , but this would possibly have
In 1 973-75 I worked with Valery Chekhov,
been the lesser evil ) 1 1 liJc4 liJ b6 1 2 liJxb6!
preparing him for the world j u n ior champion
cxb6 ( 1 2 . . . axb6 1 3 'iix dS l1xdS 1 4 J:r.fd 1 ,
ship. We devoted a lot of attention to
and 1 4 . . . :eS is not possible because of 1 5
improvi ng his method of taking decisions,
liJd5) 1 3 'it'xd8 :xd8 1 4 l:tad1 fS (it is
and in particu lar on savi ng thinking time (at
already hard to offer Black any good advice)
that time Chekhov invariably used to get into
1 S xb7 l:tab8 1 6 e4 d7 1 7 dS, and
ti me-trouble). An instructive episode oc
su bsequently Wh ite converted his extra
cu rred in one of our training games.
pawn .
4... 0-0
Dvoretsky - Chekhov
Sometimes even such a natural move as
Moscow 1 974
castl ing can turn out to be a loss of time. The
Reti Opening
most accu rate move order is 4 . . . d6! (intend
1 liJf3 liJf6 ing 5 . . . e5) 5 d4 c5! . The point is that Wh ite
2 g3 g6 has to reckon seriously with 6 . . . cxd4 7 liJxd4
3 b3 g7 d5, and after 6 c4 it is possible to play either
4 b2 6 . . . cxd4 7 liJxd4 d5 S g2 dxc4, or 6 . . . liJe4
with the unpleasant th reat of ? . . . 'it'a5+ (if the
It was no accident that Valery asked me to
moves 4 . . . 0-0 5 g2 are included , White
employ this particular variation (the theory
can simply castle here). Also, nothing is
of which I did not then know). This was how
given by 6 dxc5 'iWa5+ 7 liJbd2 'it'xc5
Alexander Koch iev, his main rival in the
(th reatening S . . . liJg4 ; bad is S a3 0-0 9 g2
forthcoming elimi nation tou rnament, l i ked to
liJg4! 1 0 liJe4 liJxf2 ! ! ) S d4 'it'h5 (S . . . Vc7
play with Wh ite. It should be said that,
is also good ) 9 g2 liJc6 1 0 b2 h 3 .
although apparently unpretentious, such
opening set-ups are qu ite venomous. If Another version o f t h e same idea is 4 . . . c5!
Black does not have a clear pla n , he can (th reatening 5 . . . d5) 5 c4 d6! (intending
q u ickly end up in a strategically d ifficult 6 . . . e5) 6 d4 liJe4 ! .
position - it is sufficient to commit a couple S g2 cS
The Tech n ique of searching for and taking Decisions ttJ 21

6 c4 lDc6 exchange on f5) secures Wh ite a positional


7 0-0 d6 advantage. This means that there remains
I n such positions . . . e7-e5 is an unpleasant only 1 0 . . . lDxd2 .
positional threat, si nce the bishop at b2 is When I played 1 0 lDh4, I was intending to
shut out of play. It can be activated only by consider whether it was possible to spoil the
e3-e3 and d2-d4, but this is d ifficult to carry black pawns by interposing the exchange
out. on f5. Suppose the complications after 1 1
8 d4 lDe4 lDxf5 lDxf1 1 2 lDxg7 should prove to be in
my favour thanks to the strong bishop on b2 .
9 lD bd2
But, of cou rse, I did not bother to calculate
The game Korch noi-Gl igoric (USSR v. Yu the variations beforehand and I merely
goslavia Match 1 956) went 9 e3 .i.g4 1 0 satisfied myself that the simple 1 1 'ili'xd2
'iVc1 .i.xf3 1 1 .i.xf3 lDg5 1 2 .i.xc6 bxc6 1 3 was qu ite possible ('emergency exit' ! ) .
dxc5 dxc5 1 4 .i.xg7 xg7 1 5 f4 lDe4 1 6
Generally speaki ng, t h e idea o f 1 1 lDxf5
lDc3 ( 1 6 'ili'c2 was stronger) 1 6 . . . 'iVd3 with
looks d u bious, and, as we have already
equality.
mentioned , this means that we should
9. . . .i.f5 immed iately concentrate on looking for the
In the event of 9 . . . lDxd2?! 1 0 'ili'xd2 (Keres simplest refutation . Don't be in a hu rry to
Szabo, Hastings 1 954/55) the pin on the study the position where Black is the
a 1 -h8 d iagonal is eliminated , and Wh ite exchange up (even if it is in his favour after
aims to gain a spatial advantage by d4-d 5 . 1 2 . . . cxd4 ! ) - first ask yourself the q uestion
1 0 lDh4 1 ? about candidate moves. I saw the reply
1 1 . . . lDxc4 ! , with which Black retains a sound
extra pawn , and I immediately terminated
my calculation.
Thus 11 'ili'xd2 is forced. Most probably the
opponent should not al low d4-d5 (although
1 1 . . . .i.d7 is in fact possible). He has to
choose between 1 1 . . .cxd4 1 2 lDxf5 gxf5 1 3
.ltxc6 bxc6 1 4 .i.xd4 and 1 1 . . . lDxd4 1 2 lDxf5
lDxf5 1 3 .i.xb7. Which of these positions is
preferable for Black? Here it is not calcula
tion that is requ i red, but a correct positional
assessment.
I think that the second continuation is
correct. Wh ite will have to lau nch an attack
on the kingside by h2-h4-h5 or possibly f2-
The knight on e 4 is attacked . H o w would f4-f5. But the chances of such an attack
you conti nue as Black? What position would succeeding are du bious - there are not
you go in for? many pieces left on the board . At the same
Sharp attempts such as 1 0 . . . d5? 1 1 cxd5 time Black is ready for an immediate
'i'xd5 1 2 g4 or 1 0 . . . lDg5?! 1 1 lDxf5 gxf5 1 2 counterattack on the queens ide by the
e3 (with the idea of 'ili'h5) are obviously advance of his a-pawn. Objectively the
inferior. 1 0 . . . lDf6? is also bad - the seizu re chances here are roughly eq ual.
of space by d4-d5 (immed iately or after the The capture on d4 with the pawn is less
22 The Technique of searching for and taking Decisions

good , since pawn weaknesses are created point you spend too much time and energy,
in Black's position. Wh ite's superior pawn you do not have sufficient for the solving of
structu re guara ntees h i m a slight but endu subsequent problems.
ing advantage . 1 6 Wf4 Wd7
B u t what happened in t h e game? After I 1 7 :ad 1
played 1 0 li:)h4, Chekhov sank i nto thought. Threatening 1 8 %:tdS with the win of a pawn .
For a couple of min utes I checked the
17 . . . 'ire6
consequences of 1 0 . . . li:)xd2 1 1 li:)xfS. Then,
1 8 I:tfe1
having convinced myself (in his time ! ) that
this could not be played , I simply began The opponent has to reckon not only with
strolling around - after a l l , there was l:[dS, but also with the opening of l i nes: e2-
noth ing more to th ink about, and it was my e4 . If 1 8 . . . Wf6 there follows 1 9 lidS eS 20
opponent who faced a choice . Wd2 (from this variation it is understandable
why 1 8 e3?! would have been less accu
Ten mi nutes passed , then a fu rther ten
rate).
minutes, and I began to grow slig htly
i rritated. What was he th inking about? By 18 . . . f6
the method of elimination it is not d ifficult to
establish that the exchange of knig hts is
obligatory - why then waste time?
10 . . . li:)xd2
1 1 'ili'xd2
Obviously I made my move instantly, after
which Chekhov again sank into thought for
some twenty-five m i n utes. Apparently he
had still not come to a defin ite conclusion, or
else new doubts had arisen and he again
began trying to solve a problem over which
he had racked his brains earl ier.
It stands to reason that to spend almost an
hour on a not very complicated problem was
pure wasteful ness. And also he did not 1 9 l1d5
solve it in the best way. This previously plan ned move, win ning a
11 . . . cxd4? ! pawn , I made without thinking, and this was
a mistake. Stronger was 1 9 e4! fxe4 20
1 2 li:)xf5 gxf5
:xe4 Wf? (20 . . . Wd7 21 We3 with two
1 3 xc6 bxc6
threats: 22 :xe7 and 22 WxcS) 2 1 Wg4+
14 xd4 i.. x d4?! h8 22 Wd 7 .
14 . . . eS 1 S i. b2 We7 was preferable. 19 . . . a5?
1 5 Wxd4 c5? 1 9 . . .We4 20 l:[xfS Wxf4 21 llxf4 as was
Now Black's position becomes d ifficult. He more tenacious. In the endgame the ad
should have chosen between 1 S . . . Wd7, vance of the a-pawn promises Black some
intending ... e7-eS, and 1 S ... e6, preparing cou nterplay. But in the m iddlegame it is of
. . . d6-d S . However, weak play on the part of l ittle use - in reply Wh ite launches an attack
my opponent was not surprising - if at some on the king.
The Tech n ique of search ing for and taking Decisions ctJ 23

20 ':'xf5 a4 I n t h e 3rd part o f t h e book School o f Chess


21 e4 axb3 Excellence 1 - Endgame Analysis I de
22 axb3 ':'fb8 scribed an interesting form of training - the
playing of stUdies. In this way you can
23 :e3
develop you r imagination and tech nique of
White has a strateg ically won position. He calculating variations , and, in particu lar,
intends to open l ines on the kingside by g3- learn to make skilful use of the method of
g4-g5 . elimi nation.
After t h e g a m e I , naturally, asked Chekhov
what he was th i n king about for so long on
the 1 0th move .
H . Kasparian
'What do you mea n ' , he replied, ' I was
1 950
choosing what to do next . '
'But wasn't 1 0 . . . tL'lxd2 forced?'
'Yes. '
'After making i t , was there a n y risk that you
could lose q u ickly?'
'No.'
'Well then, make a move and then th i n k
later, when you are indeed a t t h e cross
roads. It is far easier to choose if the
position is already on the board - there is
less likelihood of an oversight. Besides, on
the way it is possible that you will also be
able to th ink during my time - if you had
captu red on d2 immediately, I would still
have had to check 11 tL'lxf5 . '
It is u n l i kely that the wh ite knight will escape
It was strange, b u t a t t h e time these ideas alive from the trap. The attempt to save it
were new to Chekhov; he was completely with the help of the bishop is easily parried :
unfamiliar with ways of saving time, such as 1 ..i d5? tL'lf4 (not immed iately 1 . . . e6? 2 e5+
the method of elimination. However, such a or 2 ..ixe6 xe6 3 tL'lg6) 2 ..i c4 e6 3 g3
mistake in thinking is very common, and it is ..id6 and 4 . . . g7. It looks tempting to play 1
repeated by many players. xh3 g7 ( 1 . . . ..i g7 2 e5+) 2 g4 tL'lf6+ 3
An analysis of the typical mistakes made by f5 , but Black can react fa r more strongly:
Chekhov in the taking of decisions hel ped 1 . . . tL'lf4+ 2 g4 e5! followed by . . . ..i g7.
us almost to eliminate his previously typical Here Wh ite has no cha nces at all.
wild time scrambles and to g reatly improve What then can he do? Only 1 e5+ g7 offers
his practical playing strength . In the follow some hope, and now either 2 ..if3 tL'lf4 3
ing year, 1 975, he successfully came through g3, or 2 xh3 xh8 3 g4 tL'lg7 4 g5,
the qual ifying events and then won the trying to exploit the temporary lack of
esteemed title of world j u n ior champio n . harmony in the placing of the opponent's
pieces. Of course, we must qu ickly d iscover
l ittle traps which make l ife easier for us in
certain variations, for example: 2 xh3
24 The Technique of searching for and ta king Decisions

lLlf4+ 3 g4 lLld3 4 lLlg6! or 3 . . . lLle6 4 lLlf7 !


xf7 5 .i.d5 with a draw.
Thus our first move is clear.
1 e5+ 1 g7
What now? After 2 .i.f3? lLlf4 3 g3 the
black knight acquires the additional square
g2. After 3 . . . lLlg2! 4 xh3 lLle3 the defen
sive resources are exhausted . Black also
wins by 3 . . . lLle6 4 lLlf7 xf7 5 .i.d5 h2 6
xh2 .i.g7. This means that the pawn must
be captured immed iately.
2 xh31 xh8
2 . . . lLlf4+ leads to an immed iate draw after 3
g4 lLle2 (we already know the replies to
Si nce 6 f7? lLld8+ is not possible, Wh ite
3 . . . lLld3 and 3 . . . lLle6) 4 g5 xh8 5 g6
must move his bishop, renewing the threat
with the irresistible threat of 6 f7 (5 . . . g8
of f7 . The opponent repl ies 6 . . . lLld8, and
6 .i.d5+).
then we play 7 e6 ! . But where should the
3 g4 lLlg7 bishop be moved to?
4 g5 If 6 .i.c6? Black repl ies either 6 . . . lLld8 with
Now Black must have a long think, since it is gain of tempo, or 6 . . . .i.xe5 7 f5 lLld4+. 6
not at all easy to convert his extra knig ht. .i.a8? lLlc7, 6 .i.e4? lLlc5, 6 .i.f3? lLld4 and 6
During this time we too will calculate .i.g2? lLlf4+ are all unsuitable. There only
variations. In the event of 4 . . . g8 5 g6 the remains 6 .i.c8 or 6 .i. h 1 . But after 6 .i.c8?
only way to strengthen the position is by lLld8! 7 e6 g8 the black knight will come
5 . . . e6 6 f6 .i. b4 . Now 7 .i.c8? f8 is a into play via c6 or b7. The bishop must
mistake, since if 8 .i.xe6 .i.e7+ the bishop is defin itely remain on the long d iagonal .
lost, while after 8 .i.d7 .i.e 7 + 9 g6 .i.d8 1 0
6 .i. h 1 ! ! lLld8
.i.c8 .i. h4 (zugzwang) 1 1 .i.d7 Black places
7 e6! .i. b2
his king on e7 with gain of tempo, consoli
dating his forces. The e6-pawn must be 8 .i.d51
attacked along another d iagonal . As I Iya 8 .i.a8? lLlxe6 9 f7 lLlc7 .
Odessky pointed out, Kasparian's sugges 8. . . .i.a3
tion 7 .i.a6? loses to 7 . . . lLle8+! 8 xe6 9 h6 .i.e 1 +
lLlc7+, and therefore the correct way is 7
9 . . . g8 1 0 g6 f8 1 1 h7 does not help .
.i.f3! f8 8 .i.g4 .i.e7+ 9 g6 .i.h4 1 0 .i. h 3 ,
An amusing picture : Black's knight and king
when Black is unable t o strengthen his
are arrested i n their own camp!
position .
10 g6 .i.d2
Noth ing is changed by 4 . . . h7 5 .i.e4+ g8
1 1 .i. h 1 ! !
6 g6. It only remains to try and bring out
the knight from g7. Again the only safe square for the bishop.
4. . . lLle6+ 11 . . . .i.e3
5 g6 .i.g7! 1 2 .i.d5!
S . . . lLld8 6 .i.d5 (intending 7 e6) 6 . . . e6 7 The draw has become obvious.
.i.xe6! lLlxe6 8 f7 is hopeless for Black. This is roughly how a player can reason ,
The Tech n ique of searching for and taking Decisions ltJ 25

when playing the position from Genrikh any serious importance , it is clear what he
Kasparian's study. You will see that it is should play - why allow the opponent an
not essential to analyse lengthy and additional defensive resou rce?
complicated variations 'to the end' - it is 1 tLl h41
far more important to check accurately
When you are considering you r move in a
the necessary short variations, endeav
practical game, there is no need to ascerta i n
ouring in so doing to take account of all
h o w t h e g a m e should end. Based on a
the significant playing resources both
comparison of your two possibilities, you
for yourself, and the opponent.
qu ickly place your knight on h4, and let you r
opponent try t o fi nd a way t o save h imself.
4. Compa rison. This is a rather subtle During this time you will gain a better
method . Sometimes you quickly choose understanding of the resulting variations.
a move, only because you see that the
1 . . . 'iii' g 1
situation arising after it is nowhere
worse, and is in some places better, than 2 tLlf3+ 'iii' g 2
after another possible continuation. For 3 lLlxh2 'iii' x h2
example, in this way the accu rate move If now the wh ite king heads towards the a7-
order (33 . . . d 3 ! , rather than 33 . . . Wb 1 +?) was pawn, Black will shut it in by rush ing to c7
determ ined in the Alexander-Euwe game. with his king. How can this be prevented?
Let us consider the followi ng study. 4 e5! .i.xe5
5 'iii' e 6! ! 'iii' g 3
F. Bondarenko, M . Liburkin 6 'iii' d 7 'iii'f4
1 950 7 'iii' c 8
The bishop turns out to be in the way of its
own king.
I should mention that in the event of 4 . . ..i.c3
(instead of 4 . . . .i.xe5) 5 e6 .i. b4 6 'iii' e 5 'iii' g 3
7 'iii' d 5 'iii' f4 8 'iii' c 6 'iii' e 5 9 'iii' b7 'iii' d 6 it is now
the king that h i nders the bishop: 1 0 e7! and
wins.
From this last variation it is clear that the
tempo gai ned by Black after the i ncorrect 1
lLle1 ? .i.c3! is vitally important for h i m : 2 lLlf3
'iii' g 2 3 lLlxh2 'iii' x h2 4 e5 'iii' g 3 5 e6 .i.b4 6
'ii.? e 5 .iofa (or 6 . . . 'iii' g 4) 7 'iii' d 5 'iii' f4 a 'iii' c6
'iii' e 5 9 'iii' b 7 'iii' d 6 1 0 'iii' x a7 'iii' c7 with a draw.

Wh ite has two moves: 1 lLl h4 and 1 lLle1 . I n 5. Don't calculate ultra-complicated vari
the event o f 1 . . . 'iii' g 1 2 lLlf3+ 'iii' g 2 3 lLlxh2 ations for too long - i n these cases rely
there is no d ifference between them . How on i ntuition. Often we encou nter irrational
ever, after 1 lLle1 .i.c3! White must play 2 situations, in which it is practically impossi
tLlf3 , when 2 . . . 'iii' g 2 leads to the same ble to establish the truth with in a restricted
position, but with the black bishop on c3. time. Even if, after spending a mass of
Even without calculating whether this has energy, you find the correct move, the price
26 The Tech nique of search ing for and taking Decisions

may prove to be too high - for subseq uent I recommend that you should not restrict
decisions you will most probably not have yourself to the advice that has been g iven -
sufficient time and strength . develop it further, by working out new rules
In which cases does it make sense to for you rself. Analyse examples illustrating
spend a lot of time when considering a the rational tech nique of seeking and taking
move ? When you realise that an exact decisions, try solving special training exer
solution may be found to the problem cises, and analyse your own actions in the
facing you and that it will decisively cou rse of tournament battles. I hope that
innuence the further course of the game. you will become i nterested in this way of
In other words - at key moments (it is working on chess. And indeed , only on
very important to be able to determine chess? After all, rationa l , clearly organ ised
them). Or if you do not see any continua thinking comes in useful in any wal k of l ife.
tion that is in the least acceptable, and
you must devise one.
I n conclusion , here are a few rather d ifficult
The time of our lectu re has come to an end . exercises for independent solving .

Exercises

1 . Wh ite to move 2. assess 1 'ifh5


The Technique of search ing for and taking Decisions ctJ 27

3. White to move 4. Black to move

Sol utions

1 . Berg - Hort (BieI 1 985). c6 are already known to us, and 2 d2


Black has the advantage, since his king is c6 is no better, while if 2 e1 , then 2 . . . c3
more active. The main threat is . . . d6-c6- is decisive . The game went 2 g4 hxg4 3 fS
b5. A very important cou nter-chance for g3! 4 fxg6 g2 S f2 e6 6 g7 f7 , and
Wh ite is the pawn break g3-g4 ! . But at the Wh ite resigned in view of 7 b5 c3 8 b6 c2 9
present moment it is clearly inappropriate . It b7 g 1 'i1i'+ ! 1 0 xg 1 c 1 'ii' + .
can be carried out only with the king on the A n d yet a way t o save t h e g a m e does exist.
e-file, from where i n one move it ca n stop 1 e1 ! 1 c6
either of the opponent's passed pawns - on 1 . . . d5 2 e2! leads to the familiar
the c- or the g-file. zugzwang position , but with Black to move.
The most natu ral king move 1 e3? turns After 2 . . . c6 there follows 3 g4 ! , while if
out to be the weakest, since after 1 . . . c6! 2 2 . . . d4 , then not 3 g4? f5 4 gxf5 gxf5 5 d2
94 hxg4 3 f5 gxf5 4 h5 f4+ ! , as it is easy to c3+ 6 c2 c4 7 b5 xb5 8 xc3 c5 , but
see, Black is the fi rst to queen a pawn. It simply 3 d2! with a draw.
also does not help to play 2 d2 b5 3 g4 2 g4! hxg4
(3 c3 f5) 3 . . . hxg4 4 f5 g3 or 2 e2 b5 3 Noth ing is g iven by 2 .. .f5 3 gxf5 gxf5 4 d2.
g4 f5! .
3 fS gxfS
Thus the drawback to the king's position on 4 hS g3
e3 is that it comes under the tempo-gaining
4 . . . c3 5 h6 c2 6 d2 is similar.
check . . . f5-f4+. Having established this,
Wh ite chose 1 e2? (intending the drawi ng S h6 g2
variation 1 . . . c6? 2 g4! hxg4 3 f5). But after 6 f2 c3
1 . . . dS! he unexpectedly found himself in 7 h7 c2
zugzwang. The conseq uences of 2 e3 8 h8'i1i' g1 'i1i'+
28 \t> The Tech nique of searching for and taking Decisions

9 xg 1 cH W+ the ideas beh ind the move you choose .


1 0 f2 The game went 1 9 axb5 axb5 ( 1 9 . . . exf4?
The position is a d raw. 20 bxc6 is bad for Black) 20 ..tg5 b4! 21
Pawn endings are an excellent testing ..txf6 bxc3 (otherwise 22 lLld5 with advan
ground for training in the technique of tage) 22 ..txe7 1i'xb3 (22 . . . xe7? 23 'iVxc3)
calculating variations! 23 cxb3 xe7 24 bxc3 ':a3! 25 b4! (25
.l:!.b1 lib8) 25 .. Jlxc3 26 ':c1 lixc1 27 ':xc1
d7 with equal ity.
2. Bobrov - Platonov (Chernovtsy 1 963).
I n stead of 20 ..tg5 Wh ite should have
1 1i'h5
considered 20 ..td2 ! ? , for example, 20 . . . b4
This was played in the game. (20 . . . 0-0 2 1 ..tg5!?) 2 1 lLld5 (the b4-pawn is
1 . . . ..txg5 under attack) 2 1 . . .lLlxd5 22 exd5 ..ta4 23
2 ..te4 ..t h61 1 1i'xb4 'ii'x b4 24 ..txb4 l:tb8!? 25 ..ta3 ..txc2
Black can not get by without this move: if 26 ':c1 ..tf5 27 ':c7 (or 27 . . .f4 f6) 27 . . . ..td7!?
2 . . . 'ii'c 7(a7), 2 . . . ':c7 or 2 . . . nd7 there follows with a favourable endgame for White.
3 f7 . It is very important to clarify immedi Now let us analyse the i mmed iate 19 ..tg5.
ately which continuation of the offensive 1 9 ..tg5!? b4? !
Wh ite will choose: 3 'ii'x h6 or 3 1i'f5 .
20 ..txf6 bxc3
I n the event of 3 1i'f5 it is hopeless to play
21 ..txe7 'ii'x b3
3 . . . .l:!.c7? 4 f7 llxf7 5 'ii'xf7 . However, Black
fi nds the pretty defence 3 . . . l:td7!! 4 'ii'x d7 (4 No better is 2 1 . . .xe7 22 .xc3 ..txa4 23
f7 llxf7 5 1i'xf7 'ii' c7! , but not 5 . . . ':'c7?! 6 1i'a3 .
"f6+ ..tg7 7 1i'f5 and 8 ..txc2) 4 . . . ':c7 ! , and 2 2 cxb3 xe7
Wh ite's attack peters out (whereas after 23 bxc3
4 . . . 'ii'c 7? 5 li d 1 ! ..tf8 6 1i'xc7 ':xc7 7 ':c1 In contrast to the variation which occu rred i n
the game becomes equal). t h e game, t h e a-file is closed a n d Wh ite
3 1i'xh6 ':c7 ! ! retains his extra pawn .
O f course, not 3 . . J:td7? 4 f7 ':xf7 5 l:txf7 . Let us check whether the avoidance of the
But now if 4 f7 Black wins by 4 . . . ':'xf7 5 pawn exchange allows the opponent new
':xf7 .l:!.d 1 + 6 g2 'ii'g 8+. After 4 ..txc2 possibil ities. He can try 1 9 . . . lLld7 in the
1i'b7+ Wh ite resigned . hope of 20 ..txe7? lLlc5! 21 'ii' a 2 (2 1 1i'b4
Conclusion : the tempti ng combination 1 a5!) 2 1 . . . xe7 with a good game. However,
1i'h5? is incorrect. Wh ite should have simply interposing the exchange 20 axb5! places
played 1 l:th5, retaining a strong attack. Black in a d ifficult position: 20 . . . ..txg5 2 1
bxc6 o r 2 0 . . . axb5 2 1 ..txe7 lLlc5 2 2 'ii' b4 .
3. Jansa A.Sokolov ( I nterzonal Tourna
-
He should probably choose 1 9 . . . bxa4! 20
ment, BieI 1 985). 'ii'x b7 ..txb7. It is hardly possible to claim
Wh ite wants to place his bishop on g5, and any advantage after 21 lLlxa4 ':c8 22 lLlc3.
the only question is whether to do this It looks more promising to play 21 ..txf6 gxf6
i m med iately or after a prel iminary exchange 22 lLlxa4 , when White's chances in the
of pawns on the queens ide. There are endgame are preferable (22 . . . .:c8 23 lLlc3
arg u ments in favour of both decisions, and followed by ':d2 and lLld 1 -e3). He is given
therefore the main thing here is not so much more problems by a bold pawn sacrifice:
the choice itself, so much as its justification, 21 . . . ..txf6!? 22 ':xd6 ..te7 23 ':'b6 0-0-0 24
The Technique of searching for and ta king Decisions ttJ 29

ttJxa4 l:td4 (worse is 24 . . . .:.d2 25 f1 ! ) 25 l:th8 28 lhf2+. Black resigned in view of


l:ta 1 l:td2 (the wh ite rook is no longer 28 . . . 'ilfxf2 29 ttJh7+ and 30 'iVxf2 .
defending the e4-pawn ), or 25 b3 c7 26 c3 Also bad was 20 . . . h7? 2 1 :e5! with two
l:td2 , and the position is not easy to threats: 22 l:txd5 and 22 : h 5 .
evaluate. There were two roughly equ ivalent possibili
ties available to Black:
4. Timoshchenko - Vaganian (USSR
Championship, First League, Baku 1 977).
20 . . . g7 ! ?
The h6-pawn is attacked . It can be de
fended by the king, but the active continua 2 1 e5
tions 20 . . . .i.xh3 and 20 . . . 'it'g3 should also 21 :ad 1 ? xh3.
be considered . 21 . . . c6
Rafael Vaganian decided on a combination, II
without fully calculating all its consequences.
20 . . . 'iVg3 ! ?
20 ... xh3? 2 1 'ii'x h6 (th reatening both 22
'iVxh3, and 22 ttJg5) 21 . . . 'it'g3 22 'it'xh3 21 ttJd41
'iVxf2+ (22 . . . .ltxf2+ 23 f1 xe 1 24 ':'xe 1 is 2 1 'ilfxh6? does not work because of
unsatisfactory for Black) 23 h1 g7 2 1 . . . xf2+ 22 h 1 xe 1 23 ttJg5 'ilfh4 or
(23 . . . 'it'xc2 24 ttJg5!). The th reats 24 . . . l:.h8 23 I1xe1 ':fe8.
and 24 . . . 'ilfxc2 look dangerous, but Wh ite 21 . . . 'ilfh4
launches a counterattack. In both cases Black stands worse , but it is
24 l:tf1 ! 'fi'xc2 25 ttJg5 f2 (25 . . . ':h8 26 perfectly possible for h i m to put up a
xf7+ g8 27 l:th7!) 26 'fi'h7+ f6 27 'it'h41 successfu l defence.
30
Mi khai l Krasen kow

Wa nderi n g th ro u g h the Labyri nth

W chess , one of the main elements of a


hen choosing a move in a game of i.e. decide on the moves which need to be
considered in every specific position , Kotov
player's reasoning is the calculation of formulates the following ru le:
variations. The abil ity to carry this out deeply 'When we beg in the calculation of varia
and accurately in conditions of restricted tions, first we must mentally list and pre
time on the clock is an important component cisely establish all the possible candidate
of playing strength . Here a major role is moves in the g iven position . . . After deter
played by the organ isation and discipline of m i n i ng and listing them, we then beg in
the calculation. A major expert in this field, calcu lati ng in tu rn one variation after an
the author of the theory of calculating other. ' Of course, this appl ies not only to the
variations, with which any strong player initial position, but also to any position
should be familiar, was grandmaster Alexan arising during the calculation , where one of
der Kotov (all the subsequent citations will the sides has a choice. Thus all the possible
be from his book Think Like a Grandmaster) . candidate moves are determined before
The basic concept of Kotov's theory is the hand , once and for all, before the calcula
analysis tree. 'In any position we depict all tion of the given position . Here the i n itial
the possible variations in the form of an data is provided by general considerations,
"analysis tree ", in which the variations and i ntuition , and so on.
sub-variations are presented in the form of On the whole, these two rules of Kotov are
branches and twigs '. Kotov d isti ngu ishes appl icable to the majority of practical calcu
different types of calculation trees : bare lations, and the abil ity to follow them, until
tru nk (one variation or branch ), bush (many they become automatic, sign ifies a colossal
short variations) and thickets (mass of long, advance in a chess player's thinking .
complicated variations). Accord ing to Kotov, However, these rules also have sign ificant
the main rule of calculation is that 'a player drawbacks, which appear mainly in compli
should. . . work through each branch only cated , unclear positions. I nternational mas
once. No turnings, no returnings! Only in ter Boris Zlotn ik, who has made a study of
certain isolated cases, in especially compli this question, has this to say:
cated positions, should a grandmaster check '1. In complicated positions it is extremely
once again a selected variation. But, as a difficult to determine immediately all the
rule, he does not rush up and down the candidate moves at the very start of the
calculation tree '. thinking process, and in reality they appear
I should mention that the analysis tree is when a player delves into the position.
essentially created during the analysis '2. The correct move - a nuance in the
process itself; i . e . we are talking about a calculation of one variation - often "sur
defi nite order in which it is produced . faces" in the calculation of another, and
I n reply to the second important question : therefore in difficult situations a repeated
how t o construct the very nodes o f the tree, calculation is inevitable. '
Wa ndering through the Labyrinth l2J 31

Here is a very simple example, illustrating importance for the taking of a decision . And
Zlotn i k's first remark: th is can often be decided on the basis of the
calcu lation of other candidate moves. In this
Alexander - Marshall case the order of consideration plays a
major role.
Cambridge 1 928
I n the present article an attempt has been
made to suggest (and illustrate with an
example) a more complicated calculation
algorithm (true, a not very formalised one),
which is effectively used (sub-consciously)
by many players in complicated positions
(of the 'th icket' type).
1 . Decide on the aim of the calcu lation ,
i . e . the criteria by which we will assess the
variations we calculate, whether they satisfy
us or not. This may be, for example,
achieving a decisive material advantage;
enhancing a positional advantage; gaining
equal play; putti ng up resistance in a bad
position , and so on. The aim should be
After calculating the variation 1 l:tf4 exf4 2 realistic, i . e . based on an assessment of the
gxf4, to which Black repl ies 2 . . . dxc3 , con position and intu itive considerations. If there
trolling the g 1 -sq uare, we arrive at the idea is sufficient time for calculation , the aim may
of first moving the knight: 1 ttJd5 or 1 ttJa4! be raised somewhat; if there is little time, it
(and then 2 .l:.f4). The move 1 ttJa4! is in fact may be lowered.
the quickest way to wi n . But without the 2 . Search for ideas to ach ieve the aim,
calculation of the (albeit elementary) varia choose appropriate candidate moves
tion with 1 lif4, it is not clear for what reason and (very important) determine their order
it should be included in the list of candidate of priority, i . e . select those which are most
moves . likely to prove successfu l .
The second remark characterises such 3 . Calcu late variations ( a s deeply a s
featu res of human thinking as the work of possible) i n thei r order o f priority (begin
the su bconscious and association . Another ning with those which seem most appropri
significant defect of Kotov's theory becomes ate for achieving the aim). Here each time
apparent: he ignores the problem of the there is a choice, the calculation order is
order in wh ich candidate moves should be also determined by the priority of the
considered , assu ming that this 'depends on possible moves (in connection with the
the character and habits of each player and aim).
the peculiarities of the position. ' Of cou rse , 4. If a contin uation leading to the set aim
if, as Kotov impl ies, all ca ndidate moves is found, what happens next depends on
must be examined , the order in which this is how much time there is on the clock: if there
done is not of great sign ifica nce. But in fact is insufficient, the main part of the calcula
in many cases, in the interests of economy, tion may be concluded here (sti ll necessary
the calculation of a number of candidate is the 'Blu menfeld check' - cf. point 8
moves may be omitted , if this of no below); but if there is still time in hand , the
32 Wa ndering through the Labyrinth

aim can be refi ned (raised ), and the set of However, a s we have seen above , this is not
candidate moves which have not yet been always possible.
examined also refi ned , and the calculation 7. It may happen that, while calculating one
continued; if the new aim is not achieved , of the variations, a new idea appears, a
then stick with the continuation fou n d . candidate move u n related to this variation .
5 . I f , as a result o f t h e calculation a path I n this case its priority should be estab
leading to the aim is not found, the fu rther lished , but you should not start examining it
actions also depend on the clock situation . until you have completed the calculation of
I f time i s short, the a i m should b e lowered, the variation you were working o n . An
the set of candidate moves corrected, and exception may be made when it is immedi
the calculation continued . In this case the ately obvious that the new idea is better
new aim will often be satisfied by one of the than the conti nuation being exami ned (but
variations already exami ned or it will be not simply of higher priority).
comparatively simple to fi nd an appropriate 8 . One of the major th i n king deficiencies of
conti nuatio n . The only thing you must avoid many players is 'chess bli ndness', the
is making a move 'in the dark' , without overlooking of elementary replies by the
calculation. opponent of 1 -2 moves. To tackle this evil in
6. If, however, there is sti l l plenty of time, the calculation of variations there exists the
and your intu ition suggests that the aim 'Blumenfeld ru le' (Kotov also talks about
should be achievable (a strong player it): after concluding your calculation and
should trust his intuition more, since it is an taking a decision , pause and look at the
accu mulation of his chess understanding), position with the fresh 'eyes of a novice': is
then you can (and shou ld) deli berately go in the plan ned move a blunder, lead ing to
for a ' repeat' calculation of certain varia immed iate d isaster? Only after ascertaining
tions. For this you have to fi nd new ideas for that it is not a blunder can it be made on the
achieving your a i m . I n accordance with th is, board . But if a mistake is d iscovered , the
new ca ndidate moves and 'candidate varia calculation of variations will have to be
tions' are fou n d . I should explain what is renewed . In this case you should normally
meant by this. During the fi rst stage of the lower the aim of the calculation , and aim for
calculation , against many conti nuations you simpl ification, since the bad oversight is
will already have found the only or the evidence the player is un prepared for a
strongest repl ies for the opponent, and complicated battle .
forced series of moves. Often a new idea , a Most strong players are well aware o f the
candidate move, is found not in the i n itial Blu menfeld ru le, but. . . in the heat of the
position , but after a series of moves, which battle they often forget about it.
together with it comprises a 'candidate
I should l i ke to illustrate what I have said
variation ' . After this there begins the calcu
with an extract from one of my own games. I
lation of new possibil ities (point 3) - the
regard the winning manoeuvre found in it as
second stage of calculation (it can happe n ,
one of my best creative achievements.
although rarely, that t h i s cycle w i l l also
proceed a third time).
Generally speaking, a repeat calculation is
a shortcoming, therefore it is desirable to
encompass all the ideas for ach ievi ng the
aim i n the fi rst stage of the calculation .
Wa ndering through the Labyrinth CD 33

Lagunov - Krasen kow I I I . 23 . . . 'ii' a4 24 ..td2 (but not 24 b3? 'iVa5),


Dnepropetrovsk 1 985 and nothing is apparent.
IV. 23 . . . 'ii'g 4 24 ..td2.
V. 23 ... 'ii' e 5 24 ..te3.
I n the last two variations Black's possibil ities
are obviously exhausted .
Thus the first stage of the calculation did not
g ive the desired solution . If Black had been
short of time here , he would have had to
make a correction to his goal (for example,
look for the best version of winning a second
pawn). I n cidentally, then the candidate
move VI . 23 . . . 'ii'x b2 (wh ich clearly does not
satisfy the i n itial maxi mum goal) would have
been included in the exami nation , and B lack
would have had to choose from 1 .2a, 1 .2b
and VI . But, fortunately, there was sufficient
The poor placing of the wh ite pieces and the
time (in the tou rnament the 'good old' time
weakness of the back rank suggested to
limit of 2Y:z hours for 40 moves was being
Black that he should look for a forced wi n ,
used ), and Black decided to look for new
i.e. a major w i n o f material (the a i m o f the
ideas. He conceived the idea of rearranging
calculation). The idea is to combine attacks
the queen beh ind the rook; in this case the
on the bishops with a threat of invad ing on
move ..t b 1 -c2 , defending the d 1 -point, must
the back rank. In order of priority the
not be allowed . The following candidate
candidate moves were determined : I. 23 . . .
variations emerged :
'ii'd 1 ; I I . 2 3 . . . 'ii' b4 ; I I I . 2 3 . . . 'ii'a4 ; IV. 2 3 . . . 'ii'g 4;
V. 23 . . . 'ii'e 5, and the fi rst stage of the 1 ' . 23 . . . 'iVd 1 24 '>itf1 'ii' a4 25 ..td2 .:td6; I I ' .
calculation was beg u n . 25 . . . l:td5; I I I ' 25 . . . .:td4; IV'. 2 3 . . . 'ii' a4 2 4 ..td2
:Id6 ; V' . 24 . . . .l:.d5; V I ' . 24 .. Jd4; VI I ' .
I . 2 3 . . . 'ii'd 1 2 4 '>itf1
2 3 . . .'iIi'd 1 2 4 '>itf1 'ili' h 5 25 ..td2 'ii' b 5+ 26
1 ) 24 . . . 'ii'g 4 25 ..td2 'ii' c4+ 26 'ii'e 2, and '>itg 1 l:td6; VI I I ' . 26 . . . l:td5; IX'. 26 . . . l:td4 (in
there is nothing; each case with the threat of .. .'it'd7). The
2) 24 ... 'ii' h 5 25 ..td2; fi rst three were qu ickly rejected i n view of 26
'iVe2 , the last three - if only because of 27
a ) 25 ... 'ii'x h2 26 f3! 'ii' h 1 + 27 '>itf2
c3, and in the remaining cases this was
'ii' h 4+ 28 '>itf1 'ii'c4+ 29 'ii'e 2;
found:
b) 25 . . . 'ii' b 5+ 26 '>itg 1 'ii'x b2 27
IV'. 23 ... 'iVa4 24 .id2 .l:td6 25 'ii'e2 ! (25
..te3 'ii'c3 28 'ii'f 1 or 27 ... 'ii' b4 28
'>itf1 ? 'iVb5+ 26 'iIIe2 'ili'xb2 27 .id3 'iIId 4)
'ii'c 1 .:tc7 29 'ii' d 1 ;
25 .. .'iIi'd7 26 ..te1 or 25 . . . 'iIi'a6 26 'ili'e 1
3) 24 . . . 'ii'a4 25 ..td2 'ii' b 5+ 26 '>it g 1 - cf. 2b). (weaker is 26 'ii'e 3 'ii' b 6 ! ) , intending ..tc3
In these variations Black merely wins a and .ic2 , when everyth ing is defended .
second pawn - the goal is not ach ieved . V' . 24 . . . l:td5 25 .ie4! and then 26 .if3, si nce
I I . 23 . . .'iIi'b4 24 'ii'c 1 'ii'c5 (24 . . . :tc7 25 'ii' d 1 ) 25 . . . 1:[e5? 26 xh7+ is not possible.
25 ..tc2 .:tc7 26 'ii' d 2! f6 27 ..tb3! or 27 ..tf4 ! , VI ' . 24 . . . l:td4 25 'iVe2 (25 '>itf1 is weaker in
and Black does not ach ieve anyth ing. view of 25 .. :iIi'b5+ 26 'ili'e2 'ili'xb2 27 ..td3
Wa ndering through the Labyrinth

'ifxa2 , or, more accu rate, 25 .. :ii'a6+! 26 impossible to find, without first calculating
g 1 'ifd6) 25 .. .'ii' d 7 26 e1 or 25 .. .'ii' a 6 26 nu merous variations and delving i nto the
'ili'e3! . labyrinth of d ifferent attacks and defences .
I n none o f these variations is i t apparent On the other hand, the fu rther calcu lation
how Black can win . Thus the second stage (already the third stage!) is not complicated .
of the calculation also failed to produce a With the wh ite q ueen on e2, the invasion of
result. Black's on d 1 is decisive: 26 e 1 'ifd 1 27
'ili'e4 f5 or 26 'ii' b 5 l:td8 27 e 1 'ili'd 1 28 'ili'a5
So, should the calcu lation be cu rtailed and
b6.
the maximum goal abandoned? I neverthe
The entire th ree-stage calculation (together
less decided to keep looking. And like a
with the verification) took exactly an hour.
flash of lightening an idea occu rred to me. U nfortunately, in the game after 23 . . . 'ifa41
23 . . . 'ii' a4 24 d2 .l:[d6 25 'ilke2 'ilkd4 ! ! . 24 d2 :td6! the player from Novosibirsk
played 25 f1 (?) and after 25 ... 'iIi'b5+ he
resigned. Alas, Black's main idea remai ned
off-stage . . .
I am proud to say that, of the many strong
players (grandmasters and masters) to
whom I have shown this position, only one
has been able to find the solution independ
ently.
I am not suggesting at all that my proposed
algorith m is appl icable to all complicated
positions. At the same time, like any other
method of organ ised th inking, it can g ive
good resu lts when it has entered a player's
subconscious and he follows it automati
cally. But this can be achieved only by
Yes, the idea o f return ing with the queen i s special training, to which , alas, few players
not a t a l l obvious. I think i t would have been g ive sufficient attentio n .
ttJ 35

B e n iam i n B l u m enfeld

Visual I mag i nati on a n d


t h e Calcu lation of Va riations

M time studying the opening. A knowl


ost chess theoreticians spend their Blu menfeld - Zhivtsov
Moscow Championship Semi-Final
edge of the opening certainly plays an
important role in over-the-board play, but
nevertheless not a decisive one. Even in the
last Alekhi ne-Euwe match ( 1 935), despite
Euwe's enormous amount of opening prepa
ration and Alekh ine's d u bious opening ex
periments , there were only a few games
where a loss was exclusively the result of a
badly-played opening. At any event, in most
games the outcome depended , or could
have depended , on insufficiently deep play
in the later stages of the game by one
player, and often by both .
Hence the conclusion: the elimination of
shortcomings in the thinking process is
at least as important as a complete I n this position Black placed his bishop en
knowledge of the opening. prise, by playing 1 ... .i. e4?
In expressing the hope that the chess press This blunder particularly staggered me,
and chess organ isations will raise the because my young opponent's play in this
question of investigative work in this field , game, and in the tournament in genera l ,
we assume that the thoug hts of over-the created a good impression, he was not in
board players about methods of elimi nating time-trouble, and he thought about his faulty
mistakes and improving the qual ity of chess move for qu ite a long time. As I d iscovered
th inking may be important, if only as from a conversation with him after the
material for futu re research . game, he rejected 1 . . . .i.h3 (1 . . . b7 2 ttJg4)
For a fi rst test we have chosen the fol lowing 2 'ii'x h3 'ii'x e5 3 'ili'xh6 'ii'e 1 + 4 :xe 1 gxh6,
topic. when the advantage is rather with Wh ite,
who has bishop for knight with a pawn
majority on the queenside. I n making the
move in the game, he thought that the
bishop could not be taken in view of the
variation 1 . . . .i.e4 2 .i.xe4 'it'xe5 3 .i.h7+
cJ;lxh7 4 'ii'x e5 ttJd3+ regaining the q ueen ,
36 Visual I m agination and the Calculation of Va riations

but he overlooked that after 2 .i.xe4 the rook Sergeev - B l u menfeld


on a8 was attacked . Moscow Championship Semi-Final
At first sight it seems inexplicable how the
player with Black saw qu ite a long way in the
two g iven variations and at the same time
he overlooked an obvious attack after
Wh ite's very first move.
As fa r as I can judge from my own
experience in similar cases, the cause of
this mistake was as fol lows: when the player
with Black calculated the variation 1 . . . .i.e4 2
.i.xe4 etc. , in his mind he did not place the
bishop from c2 onto the e4-square, but as
though held it in mid-air, aiming for the h7-
point, in order to give check and win the
queen by a d iscovery.
The mental movement of the bishop can be This position was reached in a game played
expressed as follows: from c2 the bishop is in the last round of the same tournament. A
aiming for h7, where it will be released , and draw was sufficient for me, in view of my
it is realised that on the way to h7 there is a tournament position . It is well know that
halt at e4 , but this halt at e4 is not made in playing for a draw is not so easy: aiming
the visual imagi nation . Such mistakes are for simplification and a fear of complica
not uncommon in practical play. tions can have an adverse effect.
This occu rs especially often whe n , i n con Here I played 1 . . . tbd4 . I spent more than
sidering a variation, i n your visual imagina half an hour on this move, since I was
tion you forget to move a piece or remove a uncertain how to conti nue: Wh ite is th reat
captu red piece, and in your mind the piece ening to exchange the fianchettoed bishop
incorrectly remains on its in itial square. by .i.h6. I did not want to waste time movi ng
It should be mentioned that in most cases my rook from f8 , especially since it may
such mistakes are the result not of weak come in usefu l on the f-file to su pport the
ness of visual imaginatio n , but either of . . . f7-f5 advance. In general I realised that
nervous haste, or of insufficient intensity of Wh ite has a clear plan of attack on Black's
wil l . After mentally making a move , you castled position involving the advance of his
ought promptly to fix the change in your f-pawn , and possibly also his g-pawn , and I
visual imagination , but instead you often did not see any sufficiently convincing way
make the move as the result of a conversa of opposing this. Not knowing what to do, I
tion with yourself, or by reproducing the deferred the decision by playing 1 . . . tbd4.
move notation in your mind, or else, When I made my move, I was sure that, in
although you make the spatial movement in view of the threat of . . . tbxf3+ followed by
you r mind, it is without fi rmly fixing the . . . tbe5 etc. , my opponent would reply 2
position after the move in your visual h2, and to myself I gloomily thought: what
imaginatio n . am I going to do then, since 2 . . . a5 3 .th6
.txh6 4 'iVxh6 tbxc2? 5 tbg5 tbf6 6 tbd5 is
Here is another example. bad for me. Here I several times kept
Visual I magi nation and the Calcu lation of Va riations
CD 37

retu rning to the thought: ' It's a pity that after At any event, as far as I can judge from my
1 . . . lL'ld4 2 'ifi>h2 lL'lxf3+ 3 xf3 lL'le5 4 g2 own experience, moments occu r when the
xh3 he captu res on h3 with the king (and impression created by visual imagination
not the bishop), and I ca n't derive anyth ing displaces reality.
from the exposed position of his king ' . Although such occurrences are rare , it can
I n the game (after 1 . . . lL'ld4) Wh ite replied 2 be considered a regular phenomenon that
lL'ld1 ? and here I thought for five whole moves made mentally when considering
min utes before I saw that with 2 . . . lL'lxf3+ etc. one variation h i nder the correct visualisa
I cou ld win a pawn . I spent these five tion of a position reached in another
minutes hesitati ng over what plan to choose, variation. It is clear that the greater the
without reach ing any conclusion , and to number of variations and the longer they
take a rest from these gloomy thoughts I are, the greater the possibility of a
retu rned to the previous one: ' It's a pity that mistake.
after 2 . . . lL'lxf3+ 3 .i.xf3 lL'le5 4 .i.g2 .i.xh3 he The followi ng should also be borne in m i n d .
captu res with the king ' , when suddenly I I n a l o n g variation , each move m a d e in the
saw that he couldn't capture with the king , mind leads to a position which is increas
since it was at g 1 , not at h2. ingly removed from real ity, and therefore the
Thus during these five min utes, in my mind impression becomes fainter and fainter.
his king was not at g 1 , where I could see it Even though a player with an especially
with my own eyes, but at h2, i . e . the square strong visual imagination is sure that he can
to where I had earlier moved it in my visual picture correctly in his mind a position
imag ination, in anticipation of my oppo reached as a result of a long variation , he
nent's reply. It is qu ite possible that if after 2 can not be sure that the defin ite weakness of
ttJd 1 I could have easily decided on some the impression will not influence the correct
thing else, and had not retu rned to thoug hts evaluation of the position, reached at the
of 'it's a pity' etc. , I would not in fact have end of the variation. With every player it
played 2 lL'lxf3+ with the win of a pawn .
...
happens that, after calcu lating a variation
The especially interesting point about this correctly, he cannot decide whether or not it
case is that I played 1 . . . lL'ld4 based on the is advantageous for h i m , which , as fa r as I
threat of winning a pawn , but after movi ng can judge from myself, is mainly explained
the wh ite king in my mind when considering by the fact that the pictu re in his mind is
the variation , I forgot to put it back, and then insufficiently clear. A chess player's think
the impression created by my visual imagi ing involves his visual imagination.
nation h i ndered the objective perception of Therefore, the clearer and more vivid the
the square occupied by the king. visual picture, the easier and more
Th is explanation of the above occu rrence is accurately his thinking works and the
not the only one. The following explanation more fruitful it is.
is also possible: when I was considering There is another danger involved with
1 . lL'ld4 , I decided that si nce Wh ite replies 2
. . long variations: the mental strain of
h2, it means that 2 . . . lL'lxf3+ does not give working out a long variation is so great,
anyth ing, and this prepared conclusion because of the need, move after move, to
remai ned in my m i n d , although the prereq record the changes with the visual
uisite move (2 'ifi>h2) was not made. imagination, that tiredness resulting from
Of course, it is hard to decide which the strain may tell later in the game.
explanation is correct in a specific instance. Every over-the-board player should be
38 Visual I mag ination and the Calcu lation of Va riations

clearly awa re of the role of visual imagina The establishment of the order of con
tion and the dangers inevitably involved in sideration should be based on aiming for
the calcu lation of variations, and he should a possible reduction in the number and
d raw appropriate conclusions, taking ac length of variations. First you should
cou nt, of cou rse, of his degree of visual examine wha t seems on first impression
imagi natio n . to be the most dangerous reply to the
F o r our pa rt, w e can d raw t h e followi ng proposed move and only if a defence is
concl usions: found against this dangerous reply
should you examine whether or not there
A fter a move by the opponent you should
is a more veiled reply. In exactly the
begin thinking not with prepared conclu
same way, if within a few moves a
sions, made beforehand, but as though
variation should give a clear, decisive
anew, beginning by visually impressing
advantage, it is pointless to lengthen in
the resulting position on your mind.
your mind the details of converting the
However strongly developed your visual
advantage.
imagi nation , it is qu ite obvious that the
impression in you r mind will be weaker than If your next move is absolutely forced,
the visual perceptio n . Therefore, when and the branches (variations) begin after
your opponent makes a move, even one your move and the opponent's reply, for
that is expected, you should never the moment it is too early to delve into
(except, of course, in extreme time the variations: after your forced move
trouble) without thinking immediately and the opponent's reply the visual
make the prepared reply to the expected picture will be clearer, and it will be
move: after all, this reply was prepared easier to calculate varia tions. This also
when the given position was in your applies to a case where in a variation tha t
imagination; it is qu ite possible that now, is, say, eight moves long, after the first
when after the opponent's move this posi few moves a forced return to the initial
tion is d i rectly perceived with all its featu res, position is possible (repetition of moves).
i . e . including the opponent's move , as a In this case you are recommended,
result of the greater cla rity of the picture without thinking for long, to make the
there will also be new ideas. first few moves, in order then to work out
the variations to the end, and if they
A strict internal discipline should be
prove unfavourable, then return to the
observed when considering variations;
initial position by repeating moves.
in particular, you should not rush men
tally from one variation to another, In positions which are not sharp, where
returning several times to the same one, there cannot be forced varia tions, calcu
but first establish an order for consider lation should be restricted to a few short
ing the varia tions applicable to the variations for better revealing the fea
specific situa tion, and then gradually tures of the position.
move in your mind from one variation to If there is a possible choice between two
another; moreover, when considering continuations, producing roughly the
each varia tion, after each move make the same effect (equality, advantage, deci
appropria te move in your mind, fix it with sive advantage), you should prefer the
you visual imagination, and at the end of continuation which involves less vari
the variation make a summary, and only ational calculation, and hence, the smaller
after this turn to the next variation. danger of a mistake. This principle should
Visual I magination and the Calculation of Va riations ttJ 39

be fi rmly adhered to, rejecting any kind of should aim to ach ieve the desired result with
'romanticism' . If, for example, there is the maximum certainty. This is why we
choice between liqu idating into a pawn consider our argu ment to be correct.
endgame with an extra pawn and a certain The importance of visual impressions for
wi n , and a mu lti-move mating combi nation chess thinking is so great, that a defi n ite role
with branches, it is more sensible to choose is bound to be played by factors aiding
the fi rst continuatio n : there have been visual perception, such as: appropriate
instances in tou rnaments where a player lighting during play, the correct correlation
annou nced mate in a few moves then lost between the board and the pieces, and a
the game, since the mate proved to be colouring of the pieces that is easy on the
fictitious. eye . From my own experience I know that if
Our arguments, especially the last one, will during a simu ltaneous d isplay the lighting is
certainly be opposed by supporters of chess poor, the pieces are pai nted an irritating
'beauty' . In our opinion, the calculation of colour, or the board is not correctly propor
variations is only a necessary technique, tioned , the result of the d isplay even against
and if this technique can be simplified or weaker players will be worse than in a
made easier, so much the better. The d isplay against stronger players but with
beauty of chess lies in its inner logic and more favourable cond itions for visual per
richness of ideas, for the revealing of which ception . I th i n k that chess organ isations
in most cases a deep penetration into the should consult with experts on physiology
position is sufficient, calculation being and psychological testing, and, in accord
needed only to check the correctness of the ance with their d i rectives, develop a stand
ideas. Chess is a pu rposefu l game: you ard type of chess equ ipment.
40

PART I I

I ntu itive Dec i s i o n s

Mark Dvoretsky

T he Devel opment of C h ess I ntu iti on

I players with a un iversal style, who perform


t i s extremely rare to come across chess Tigran Petrosian a n d Anatoly Karpov, have
a keen feeling for the slig htest nuances of a
with identical success in any type of position and possess sharp combinative
position . One such player was Bobby visio n . They are weaker in the planning of a
Fischer, and - in his best years - Boris game, in strategy, they do not especially like
Spassky. Normally all players, including calculating variations, and they make mis
outstanding grand masters, have various takes in calculation .
playing defects . It is very important to try At the opposite extreme we fi nd, for exam
and elimi nate them in good time, to 'tig hten ple, Akiba Rubinstein, M i khail Botvinnik,
up' the backward aspects of your play, Lajos Portisch and Garry Kasparov. They
without, of course, abandoning your natu ral find deep plans in the opening and the
style. su bseq uent stages, their thinking is disci
Players are trad itionally divided into combi pli ned , and they calcu late variations accu
native and positional . At one time it was rately. But occasionally they miss unex
comparatively simple to disti nguish players pected tactical ideas, sometimes they are
by this criterion , but now things are d ifferent excessively direct, and they sense insuffi
- hardly any purely positional or purely ciently keenly the turni ng-poi nts of a game.
combinative players rema i n . Besides, such Of cou rse , all this is merely an approximate
a d ivision talks only about the manner of scheme. Usually the 'diagnosis' I give to a
play, and not about thinking pecu liarities. It player with whom I am working (i rrespective
is insufficiently informative as regards choos of whether it is a candidate master or a
ing the direction and content of training g randmaster) includes many more different
req uired by a player. parameters . Even so, from the methodologi
To me it seems more productive to cal point of view this classification seems to
distinguish a player by the type of me to be very usefu l .
thinking which dominates in his ap F o r a chess player i t is a great stroke of
proach to the taking of decisions - fortune to possess a natu rally well-devel
intuitive or logical. oped intuition . But, as Alexander Alekh ine
Grand masters with an intu itive approach , poi nted out, this can also harbour a serious
such as Jose Raul Capablanca , M ikhail Tal , psychological danger.
The Development of Chess I ntuition ltJ 41

Along with the obvious advantages given by assess the promise of particular conti nua
a quick grasp of situation, the ability to see tions. I ntuitive insight enables the lengthy
almost simultaneously the whole array of and complicated calculation of variations to
tactical features contained in any compli be avoided , makes our searches easier,
cated position (economy of thinking and, as and suggests where the solution may be
a consequence, self-belief), almost insepa h idden .
rably linked are temptations: a player may A serious study of chess, of its playing
easily arrive at the faulty opinion that those methods, and a thoughtful analysis of
good moves, which on acquaintance with various specific situations significantly
the position he sees immediately - or develops and enriches our intuition. I will
almost immediately - are definitely the best, not attempt to demonstrate this argument -
and as a result of this his play loses just as it is illustrated in the first part of my book
much in depth as it gains in ease. This School of Chess Excellence 1 - Endgame
gradual rejection of seeking the absolute Analysis, in the chapter "The benefit of
best, and being satisfied only with good 'abstract' knowledge". I also recommend
moves, is unfortunately (for the art of chess) that you read the article by Eduard Gufeld
characteristic of the present phase of ' How to develop intuition' from his book My
Capablanca 's career. (From a famous arti Life in Chess.
cle by Alekh ine 'The 1 927 New York
Throughout a game we rely (to a g reater or
tournament as a prologue to the battle in
lesser extent, and with varying degrees of
Buenos Aires for the world championship' . )
success) on our intuitio n . It d isplays itself in
For players with an intuitive type of th inking the most varied forms. Think of certain
it is advisable to do training in the solving of concepts which we constantly use: 'positional
strategic problems (for example, involving feeling', 'spirit of the position ' , 'combinative
choice of plan at the transition from opening vision', 'sense of danger' , 'feeling for the
to middlegame). It is usefu l for them to test in itiative' - even from their verbal expres
their strength in exercises with the compli sion it is obvious that these are d ifferent
cated calculation of variations, demanding man ifestations of the intu itive perception of
perseverance and concentratio n . At one the game. I n principle, it would be useful to
time I suggested that Alexander Chern i n discuss each of these sepa rately, but this is
should work in t h i s directio n . Soon he made a topic for a special investigatio n .
sign ificant prog ress , qu ickly progressing
Stra ngely enoug h , in chess l iteratu re intui
from an ordinary master to a strong g rand
tion is often simply taken to mean the abil ity
master, and a participant in a Candidates
to decide on a sacrifice of material that does
tournament for the world championship.
not lend itself to exact calculation . Essen
Things are more complicated with the tially this confuses the concepts of risk,
development of i ntuition . Sometimes play because of the impossibility of calculating
ers and even their trainers do not know how the variations to the end, and intuition .
to approach this problem. In this lectu re I will
share certain thoug hts, based on my train
ing experience.
Chess intuition is the abil ity easily and (see diagram)
qu ickly - and sometimes immed iately - to
grasp the essence of a position, the most
important ideas contai ned in it, and to
42 The Development of Chess I ntuition

Sueti n - Bagi rov But some players would probably have


3 1 st USSR Championship, Leningrad 1 963 approached the problem d ifferently, indeed
intu itively. For example, M i khail Tal , after
assessing a few variations, would al most
certainly have qu ickly decided : the sacrifice
was promising (not correct, but promising)
and decided to go in for it. Or, on the
contra ry, he would have judged its conse
quences to be in Black's favour and played
differently.
1 8 'ii'x g7+ ! ? xg7
1 9 il.xf6+
'Now came the turn for my opponent to
think. As was later discovered, this was
perhaps the decisive point of the game.
Where should the king move to: h6 or g6?
8agirov thought for a whole hour, and also
Vlad imir Bagirov has just played 1 7 ... il.e7- played most probably by intuition. '
d6! . ' Normal ' conti nuations lead to piece
A strange conclusion, wou ldn't you agree:
exchanges and Black obtains an excellent
'thought for a whole hour, and played by
position. Neither 1 8 il.xd6 l1xd6 ( 1 9 'iVxd6??
intu ition '?! In fact Black tried to calculate
il.xg2+ ) nor 1 8 lIxf6 il.xe5 1 9 .:txc6 'iVxb2 !
everyth ing exactly, but he was unable to do
is dangerous for h i m . Alexey Suetin writes:
so and he made a mistake . Where does
'What was I to do ? I did not wish to go in for
intuition come in here? We see that Suetin
simplification. It was here that I was at
talks about it, clearly having no idea what it
tracted and as though entranced by a queen means.
sacrifice. I feverishly calculated the varia
I n cidenta lly, as was shown by Andre
tions. The hands on my clock inexorably
advanced, but the calculations became Lil ienthal, the sacrifice was i ncorrect and
would have been refuted by 1 9 . . . g6 ! '
more and more complicated. I had to
reconcile myself to a draw, or, relying on my If 20 il. d 3 there is t h e strong reply 20 . . . il.e7! ,
intuition, take a risk. ' attacking the bishop on d3. For example, 2 1
il.xe7 ':xd3 2 2 :g5+ h 6 2 3 cxd3 ':e8! 24
From the commentary it is clear that Suetin
il.f6 ':'e6 (or 24 . . . 'iVf2).
spent a long time trying to calculate the
sacrifice exactly, but he was unable to do The main variation is 20 l:1af1 'ife3! 2 1 il.d3
so. Of course, his bold decision contains an h6 ! ' If White now follows Tal's recommen
element of intu itive assessment, but only an dation 22 ltJd 1 'iVd2 23 .:t5f2 (23 ltJf2 ':g8),
element. I n principle, he acted not intui then 23 . . . 'iVxf2 ! 24 .:txf2 .:tde8 25 il.c3 f5!
tively, but by calculation . And this was with advantage to Black.
probably correct - Black obtains too g reat a 22 il.xd8 ':xd8 23 ':xf7 (23 l:tf6+ g7 24
material advantage (queen for just one l1xf7+ g8 25 il.xh7+ h8, and the attack
minor piece). Any unforeseen defensive is parried) 23 . . Jtd7 24 ltJd5!? (24 : 1 f6+
resource , enabling the immed iate threats to g5 25 .:tf5+ h4 ! ) 24 . . . il.xd5! 25 l:txd7
be parried , may immed iately decide the il.xg2+! 26 xg2 'ifg5+ 27 f3 'iff4+ 28
outcome in Black's favour. e2 'ifg4+ , or 25 l:t 1 f6+ 'iti>g5 26 l:1f5+ 'iti>h4
The Development of Chess I ntuition ltJ 43

27 l:[xd7 'ii'c 1 + 28 f1 (28 :tf1 xg2+ 29 25 .td3+ l::t x d3?!


xg2 'iVg5+) 28 . . . xg2+ ! 29 xg2 'ii'x c2+, In time-trouble Vlad imir Bag irov h u rries to
and Black wins. simpl ify the position. A q u ite understandable
19 . . . h6? decision , although by no means forced.
20 l:taf1 ! Suetin g ives the variation 25 ... g7 26
Wh ite i ntends 2 1 l:[h5+ g6 22 l:[h4 with the ':xf7+ g8 27 .tc4 h8 28 l:.7f5 .te8 29
th reats of 23 d3+ and 23 h5+. lDd5 with dangerous threats, but instead of
28 . . . .te8? Black has the stronger 28 . . . h4 or
28 . . . l:ld2.
26 cxd3?
I nterposing the check 26 l1f6+! would have
placed Black in a d ifficult position. He would
have had to allow the captu re on f7 with
check, since after 26 . . . g5?! 27 :t 1 f5+ g4
28 cxd3 he can not play 28 .. :iix d3? 29 ':f4+
'it>g5 30 h4 mate.
26 . . . 'ii'x d3
27 1:.f6+ g5
28 ltxf7 h4
29 g1 'iVe3+
30 l:l7f2
20 . . . 'ii'e 3?
A fu rther mistake, after which Black's posi
tion becomes anxious. In the event of
20 . . . .te7 White could either force a draw by
24 l:th5+ g6 25 l:tg5+, or play on with 2 1
i.xe7 f6 (2 1 . . . :tg8!?) 2 2 i.xf6 l:[xf6 23
l:txf6+ g7 24 1tf7+ h8 25 d3 'iVxb2 with
chances for both sides.
The sharp battle would qu ickly have ended
peacefully after 20 . . . .txg2+!? 21 xg2
1:g8+ 22 h 1 l:tde8 23 :h5+ (23 .tf3 'ii' a 6)
23 ... g6 24 :tg5+ h6. The immediate
20 . . . :tde8 !? was also possible.
21 l:th5+ g6
22 l:th4! .tf4!
30 . . .
The only defence.
Now White gains a decisive advantage. It
23 llhxf4
was essential to open up the position of the
23 l:tfxf4 'iVc1 + 24 i.f1 did not work because enemy king by 30 . . . h3! 31 gxh3 i.f3 or
of 24 . . . h5! 25 lDe2 'iVxc2. 31 . . . h4 with a probable draw.
23 . . . h5 31 lDe2 ! h6 32 lDf4 a5 33 l:[d1 a4 34 h3
23 .. :ifh3 24 lDe4! h7 35 lDd5? (35 lDh5) 35 ... 'ii'c 5? (35 . . .
24 xd8 l:[xd8 i.xd5 3 6 l:[xd5 'iVc1 + 3 7 l:tf1 'ii'x b2 ) 36
44 The Development of Chess I ntuition

liJf6+ g7 37 a3 g6 3S liJg4 h7 39 :e1 player, but in complicated tactical situations


'ii'd 6 40 liJe3 g6 41 liJf5 "dS 42 %1e6+ (in he was usually much weaker.
Sueti n's opin ion , 42 liJe7+ g7 43 :e6 was I was once able to exploit this factor.
even stronger) 42 . . . f7 43 liJd4+ 'it>g7 44
:1e4 d7 45 liJf3 f5 46 l:td4 'ili'eS 47
Dvoretsky - Bagi rov
':xh4, and Wh ite grad ually converted his
material advantage. USSR Championship, First Leag ue,
Tbilisi 1 973
Let us retu rn to the problem that Wh ite
Alekhine Defence
faced. Deliberating over this type of
1 e4 liJf6
irra tional problem is one way of develop
2 e5 liJd5
ing intuition. Think a little about the
position and try to 'guess' whether or not the 3 d4 d6
sacrifice is correct, and whether it should be 4 c4 liJ b6
made . Clearly, here you can 't get by without 5 exd6 cxd6
calcu lating some variations. When training 6 liJc3 g6
your intuition, you should aim not to 7 h3 g7
calculate everything 'to the end', but,
S liJf3 0-0
after checking some minimum number of
9 e2 liJc6
varia tions, come to a definite conclusion
as soon as possible. After then checking 1 0 0-0 f5
you opinion with the 'answer', you will 1 1 e3 d5
see whether you were searching in the 12 c5 liJc4
right direction, and whether or not at the 1 3 xc4 dxc4
very start you missed some ideas impor 14 'ili'a4 d3
tant for the taking of the decision -
This is one of the well-known variations of
evaluative or specifically tactical.
the Alekh ine Defence, on which Bag irov
I n just this way you can try to choose the was an expert. Later ga mes convinced me
correct square for the black king on the 1 9th that Black ach ieves equal ity by 14 . . . e51.
move .
1 5 l:[fd 1 'ili'a5!
You will fi nd several examples of this type Now if 1 5 . . . e5? there follows 16 d5 liJd4? ! 1 7
(with the help of the index of themes) in the liJe 1 I . Bad is 1 5 . . .f5? 1 6 d5 liJe5 1 7 liJg5! (or
afore-mentioned series School of Chess 1 7 liJe 1 ! ) 1 7 .. .f4 1 8 d4! (but not 1 8 xf4
Excellence. l:txf4 1 9 liJe6 'iVb8 20 liJxf4 liJf3+ 2 1 gxf3
The success of Wh ite's attack in this 'iVxf4). 1 6 . . .f4 (instead of 1 6 . . . liJe5) also
example was mainly based not on purely does not help: 1 7 .l:l xd3 ! ! cxd3 1 8 dxc6 fxe3
chess factors (objectively the queen sacri 1 9 cxb7 exf2+ 20 'it>f1 (now it is clear why
fice was incorrect), but psychological fac the exchange was given up) 20 . . . l:tb8 2 1
tors , which must be taken into account when 'iVc4+ h8 2 2 c6.
you intu itively assess how prom ising a 1 6 'iVxa5 liJxa5
problematic decision is. What told was the
1 7 liJe1 f5
surprise effect (Bagirov had stud ied the
position after 1 7 . . . d6 in his home prepara 1 S l:r.ac1
tions, but had not noticed the q ueen 1 8 d5! is stronger, as I later played against
sacrifice). But the main thing was Bag irov's W. Martz (Wij k aan Zee 1 975).
style of play. He was a strong positional 1S . . . liJc6!
The Development of Chess I ntuition ltJ 45

1 9 g4 iLd7 26 . . . b5 27 ':'xb3 a5? 28 lLld3, and after the


20 dS lLl b4 rook moves - 29 lLlxb5.
27 nxb3 bS
2s lLld3 l:tSe7
29 na3 hS
29 . . . a5 30 lLla2 l:txc1 + 3 1 lLldxc1 a4 came
into consideration . The pawns are block
aded , but at least they would have advanced
a little further.
30 gxhS
30 f3 was safer, or even 30 lLle2 ! ? hxg4 3 1
hxg4 iLxg4 3 2 l:txc7 ':xc7 3 3 lLld4 iL d 7 34
':a5.
30 . . . gxhS
31 h4
Here too 31 lLle2 or 31 lLla2 came into
If 2 1 a3, then 2 1 . . . lLla6 followed by 22 . . . l:tac8, consideration .
and the c5-pawn is very weak. Generally
31 . . . iLfS?!
speaking, now Black is excellently placed .
Taking account of this factor, and also my It is not clear why Black avoids 3 1 . . . a5! 32
opponent's style of play, I decided to go in lLla2 ':'xc1 + 33 lLldxc1 a4 (with the th reat of
for great complications, by provoking Bagirov 34 . . . .:tc4) 34 lLld3 iLf5 (34 . . Jc4? 35 lLle5).
into making a piece sacrifice. By playing 35 lLle5! Wh ite retains some
21 b3 1 ? lLlxa2 ! ? winning chances, but no more than that. We
see that in a complicated position Bagirov
21 . . . cxb3 2 2 axb3 e 6 ! was simpler, with a
acts unsurely. Usually he avoids situations of
roughly equal game, but the temptation
this type, he has insufficient experience in
proved too strong .
them, and so here his intuition lets him
22 lLlxa2 exb3 down .
23 lLle3 lIfeS?! 32 lLle2 lIxe1 +?
My reckoning proves justified - Black Now Black's position becomes hopeless.
immed iately commits a serious inaccu racy. 32 . . . iLxd3 33 ':xc7 l:txc7 34 l:[xd3 a5
Bagirov was hoping to prevent 24 iLd4, but suggests itself. I n the endgame, passed
he fails to ach ieve this aim. Stronger was pawns should be advanced !
23 . . . a5 24 iLd4 a4 (24 . . . e5 25 dxe6 iLxe6 is
33 lLldxe1 l:te7
also possible) 25 iLxg7 'iti>xg7 26 ':'b1 l:tfc8
27 lLld3 :a5! (i ntend ing . . . .:.xc5 or . . . iLb5) 34 lLl b3 iLe4
with chances for both sides. 3s lLl bd4 iLxdS
24 iLd41 iLxd4 36 lLlxbS lle4
2S :xd4 ':xeS 37 lLlbd4! :e7
26 l:tb4 37 . . . e5 38 l:ta51 .
26 lLld3 seemed less accu rate to me on 3S 'iti> h2 eS?!
account of 26 . . . b2! 27 lLlxb2 l:tac8 . 39 l:taS! exd4
26 . . . :aeS 40 ':'xdS l:te2
46 The Development of Chess I ntuition

41 tDxd4 lIxf2+ 'oii> g 4.


42 'oii> g 3 1:td2 47 . . . 'oii> g S
43 llg5+ 'oii>fS 4S l::t h 6 a4
44 tDf5 a5 49 .l:ta6 .l:.a2
45 l:txh5 'oii> g S 50 'oii> g 4 a3
46 l:tg5+ 51 'oii> h 5 f6
The sealed move. This was an appropriate 5 1 . . . : a 1 52 'oii> h6 would have come to the
moment to adjourn the game: Wh ite's same thing . 51 . . . 'oii> h7 52 .l:ta7 l::t f2 ! was the
position is certainly won . but now he needs most tenacious. but even then Wh ite would
to decide on a plan for converting his have won by 53 .l:.xf7+ 'oii> h 8 54 'oii> g 6 (54
advantage. and this is best done i n home 'oii> h 6!? 'oii> g 8 55 'oii> g 6 :g2+ 56 'oii> f6 with the
analysis. th reat of 57 tDe7+) 54 . . .1lg2+ 55 'oii> h 6 'oii> g 8
46 . . . 'oii> fS 56 :a7 a2 57 h5 1:tb2 58 tD h4 'oii> f8 59 tDg6+
'oii> e 8 60 tDe5! (preparing 61 'oii> g 5) 60 . . . .l:.g2
61 tDd3 and 62 tDb4 .
52 l:ta7! l::t a 1
53 'oii> g 6 :g 1 +
54 'oii> xf6 1:ta1
55 tDh6+ 'oii> h S
56 tDf7+
Black resig ned .

I rrational problems. with wh ich you can


check and sharpen your i ntuition . do not
necessarily i nvolve material sacrifices.

Kaspa rov - Karpov


47 llh5 World Championship Match . Moscow
I n itially it seemed to me that 47 h5 would 1 984/5. 6th Game
decide matters more simply. for example:
47 . . . l:.d 1 48 'oii> g 2 .l:td2+ 49 'oii> h 3 lId1 50
tDg3 (the a5-pawn is attacked ) 50 . . . a4 51 h6
l:td6 52 h7. But then I d iscovered that after
47 .. .1ld 1 48 'oii> g 2 Black has 48 . . .f6 ! 49 l:tg6
(49 h6 l:td7 ! ) 49 . . . .l:td7! (but not 49 . . . 'oii> f7? 50
1:[g7+ 'oii> e 6 51 h 6 ! ) . The exchange of rooks
after 50 .l:txf6+ .l:.f7 leads to a draw. but
otherwise. by playing 50 . . . .l:ta7. Black gains
counter-chances . I ncidentally. Black needs
to interpose 47 . . . .l:.d 1 ! . since after the
immed iate 47 .. .f6 48 l::t g 6 l::t d 7 White wins
easily by 49 l:txf6+ :If7 50 :Ixf7+ 'oii> xf7 5 1
tDd4 (or 5 1 tDd6+) 5 1 . . . a 4 5 2 tDb5 'oii> f6 53
The Development of Chess I ntuition ltJ 47

Black is a pawn u p , but the activity of the and suggests when he needs to concen
wh ite pieces more than compensates for trate and carefully check variations, or
this small material deficit. It is clear that now where , on the contrary, for one reason or
the knight must be advanced . But where to : another there is no point in delving into a
f5 or c6? detailed calculation .
On c6 the knight attacks the a7-pawn ,
restricting the black rook's mobil ity. From f5 , Tal - Dvoretsky
on the other hand, it controls the d6-square 42nd USSR Championship,
and prepa res the advance of the passed Len i ngrad 1 974
pawn . Wh ich is more important? To calcu
late the variations at the board is completely
impossible - after some approximate esti
mations you have to trust your intuition.
In his book The Test of Time Garry
Kasparov several times draws the attention
of the readers to the fact that in complicated
situations his intu itive perception of the
position proved correct. He is obviously
proud of his own intu itio n , and considers it
one of his strong poi nts . But it is clear that
any top player can boast of n umerous
examples of the correct solving of d ifficult
problems. In order to make an objective
judgement about the degree to which
intuition is developed , it is more important to 21 . . . .tfS ! ?
follow how often it lets a player down . For
'The move in the game involves a clever
example, in sharp positions the young
trap ' (Tal). I was very much hoping that the
Mikhail Tal nearly always acted in the
ex-world champion would be tempted by the
strongest way, fi nding the attacking re
possibil ity of beg i n n ing an offensive against
sou rces that were most dangerous for his
my king by 22 .txe5 .txa2 23 .t a 1 !
opponent. Whereas, as a careful study of
(threatening not only the capture of the
Kasparov's play revealed to me, his i ntuition
bishop, but also the deadly 24 'iVc3) 23 . . . 'iVb3
is far from faultless. Even in his best games,
at some point he often 'miscued ' and gave (the only defence) 24 'iVd2 . The variations
would appear to be in his favour. Such an
his opponents additional chances (wh ich ,
attack would have been fu lly in keeping with
however, they did not always exploit).
Tal's style.
That was also what happened i n thi
example. Kasparov 'guessed wrong' and 'After some hesitation, I decided not to open
missed a wi n . Later he did not sense the the sluices for the black pieces', writes Tal .
moment when it was now time to force a 'A nd I acted correctly: after 2 2 .txe5 Black
draw, and in the end he lost. You will find the had prepared 22 .tb3!!, not only securing
. . .

game in an addendum to the lectu re. opposite-colour bishops, but also regaining
the pawn!'

A correct i ntu itive perception of the situation 22 llb1 ! 'iVd7


helps a player to spend his time rationally, 23 l:ed 1 .txd6
48 The Development of Chess I ntuition

24 cxd61 essence of the position: to determine the


Subsequently Wh ite method ically converted most important problem (positional or
his positional advantage. tactical) which has to be solved, to sense
the correct direction of our searches,
Many years later I retu rned to the analysis of
and to perceive the desirability or unde
the position and ca me to the conclusion that
sirability of a particular operation. It is
22 . . . b3 (in reply to 22 xe5) was not as
clear that a well-developed intu ition assists
strong as I thought. The pawn is indeed
in perceiving th ings rapidly and correctly.
regained , but Wh ite still retains the adva n
tage in the middlegame with opposite You may have heard the classic story of how
colour bishops, by continuing 23 axb3 ':'xe5 some grand masters, absorbed in a d ifficult
24 f4 ':'xc5 (or 24 . . . :ee8 25 e5) 25 c4. analysis, asked the advice of Smyslov. After
a l ittle thought, Vasily Vasil ievich remarked :
On the other hand, in the variation 22 . . .
'The rook should be placed on the fifth rank'.
xa2 ! ? 2 3 a 1 'ii' b 3 24 'ii'd 2 it i s not
The recommendation seemed tQO abstract,
possible to demonstrate an advantage for
but with in a short time Smyslov repeated :
Wh ite . I thought that 24 . . . ':'xe4 was refuted
'Ah , if only the rook was on the fifth ! ' They
by 25 .tf3 1be 1 + 26 ':'xe 1 (26 'ii'x e 1 ! 'ii'e 6
began looking in this d i rection and soon
27 'ii' b4 is stronger) 26 . . . g7 27 xg7
real ised that Smyslov was absol utely right.
'it>xg7 28 ':'a 1 , and Black loses a piece, but
Examples of this sort of assessment can be
instead of 26 . . . g7? he has 26 . . . xc5 ! '
found by studying the commentaries to
Even more important is the fact that after
games, especially those written by players
24 . . . xc5! 25 'ii'g 5 Black parries both
with an intuitive style.
th reats 26 'ifxc5 and 26 'iff6 with the simple
move 25 . . . e7 ! .
N i mzowitsch - Capablanca
Tal d i d not delve into these variations,
because for this there was no need . The New York 1 927
basis of his decision was a correct evalua
tion of the situation on the board . I ndeed ,
why go in for complications, in the calcula
tion of which it is easy to go wrong, if all the
opponent's pieces are condemned to pas
sivity and Wh ite's obvious advantage can
be retai ned by simple means?

It is now time to move away from wild


combinative complications and talk about
quiet positional problems or the compara
tively simple tactical tasks which we are
obl iged to solve at al most every step.
I n some books you can read that the
process of assessing a position consists in
determining and weighing up all the positional Jose Raul Capablanca writes:
factors wh ich affect it. This is ru bbish ! - in 'White is at last ready to liberate his position
fact the greater part of such work is by means of b2-b4 followed by b2. Black
performed subconsciously. The art of as on the other hand, as a result of simple and
sessment is the ability to grasp the logical development, has the control of both
The Development of Chess I ntuition ttJ 49

the open files with his rooks and is also (Alekh ine) 2S . . . aS! he would have had to
ahead in time. It is now time to turn his seek salvation in a heavy piece endgame a
advantage to account before White is able pawn down . A sample variation is 26 bxaS
fully to develop his game. ' bxaS 27 l:txc8 l:[xc8 28 :d 1 .i.xa3 29 .i.xa3
20 . . . 'iWe51 'iWxa3 30 'iWa6 'iWc3 .
'A finesse to gain time in bringing the queen 2S 11ad 1 ! as 26 .i.d4! is stronger: 26 . . . axb4
into the battle. Black wants to take posses 27 axb4 .i.xb4 28 .i.xb6, or 26 . . . 'iWxa3 26
sion of the second row with one of his rooks bxaS 'iWxaS (the reply is the same after
and to do that he needs the co-operation of 26 . . . bxaS) 27 :a 1 , regaining the pawn .
the queen. The text move aims at prevent 25 . . . :c2
ing b2-b4 at once, which would be an 26 'iWa6?1
swered by 21 . . . .i.d6 22 g3 "fie4, and Black Another error by N imzowitsch in his percep
will obtain possession of the second rank. ' tion of the position : he does not sense that
As you see , Black's main aim is formulated his queen should be participating in the
- the occu pation of the 2nd ran k (it is also defence of the kingside. 26 "fif1 or 26 'iWd 1
clear what Wh ite wants - to complete his (with the idea of 27 ':e2) suggests itself. The
development and beg i n exchanging rooks). move i n the game allows Capablanca to
Without specific analysis it is d ifficult to include his second rook in the attack along
foresee wh ich of the two sides will be more the 2nd rank.
successfu l in carrying out their plans. But at
least it is clear what they need to a i m for.
21 93 'ii' d 5!
22 b4 .i.f8
23 .i. b2 'ii'a 2!
24 l:ta1 ?!
Alexander Alekhine suggested 24 J:[bd 1 !
::lxd 1 (if immediately 24 . . . aS, then 2S J:.xd8
J:txd8 26 .i.d4 ! ) 2S ':xd 1 . After 2S . . . aS 26
bxaS bxaS (26 . . . .i.xa3 27 'ii' a 6 ! ) Alekh ine
conti nues 27 'ii' a 6? l:tc2 28 l:td8 l:txb2 29
l:I.xf8+ with perpetual check, or 28 . . . 'ii'x b2
29 'ii'd 6 with a draw. As Harry Golombek
poi nted out, in this variation Black wins by
28 . . . 'ii' b 1 + ! 29 g2 "fixb2 . 27 l:.d2! is 26 . . . e5!
correct, and if 27 . . . .i.xa3, then either 28 27 .i.xe5 l:tdd2
'ii'd 1 ! , or 28 'ii' a 6! J:[f8 (28 .. Jlb8 29 'ii' x aS ! , 28 'ii' b 1?
and the bishop a t b 2 is immune) 2 9 'ii'x aS
By this point all the commentators had
'ii'b 1 + 30 g2 .i.xb2 (30 . . . 'ii'e 4+ 31 f3) 3 1
already written Wh ite off, but to me his
'i'b4 .
position seems defensible, despite the
24 . . . 'ii' b3
inaccuracies committed earl ier. The queen
25 .i.d4?! should have been retu rned to the defence:
It is surprising , but Aaron N imzowitsch does 28 'ii'f 1 (in the event of 28 :f1 ? Black
not real ise that he should seek salvation by spectacu larly decides matters with 28 . . .
exchanging rooks. However, after 2S l:.ac1 'ii'x e3! 2 9 .i.f4 l:.xf2 ! ) . Alekh ine gives the
50 The Development of Chess I ntuition

variation 28 . . . "d5 29 d4 "h5! 30 h4 At one time Sergey Dolmatov and I played


(otherwise 30 . . . l:txf2) 30 . . ...f3, and after 3 1 an i nteresting type of game, aimed at
l:.ac1 ':'xf2 the g3-pawn is under attack. developing endgame i ntuitio n . An issue of
This is why instead of 29 d4? Wh ite Informator was opened (the 'endgame'
should play 29 f4! 'ii' h 5 (29 . . ...f3 30 section) and some random number was
l:.ec 1 ! ) 30 h4 "f3 31 l:tec 1 ! or 30 . . . h6 3 1 named . The ending with this number was
e4 ! , retaining good chances o f a d raw. set up on the board . Dolmatov played for the
28 . . . :txf2 side which with correct play could (accord
ing to the assessment given in the book)
29 g4 'ili'e6
gain a d raw in an inferior position, or a win in
30 g3 l:txh2 !
a superior one. He would fi rst ponder over
30 . . ... xg4 was less good because of 3 1 the position for five m i n utes, then the clocks
l:.f1 . would be started and we would play a blitz
3 1 "f3 game, with the trainer using the analysis
3 1 xh2 'ili'xg4+ 32 h 1 "h3 with unavoid published in the book. Sometimes an
able mate . additional ru le was introduced : at one point
of the game, which Dolmatov considered to
31 . . . 1:1 hg2+
be the most important, he had the right to
32 'ii'x g2 ':'xg2+ stop the clock and think for a further five
33 xg2 "xg4 minutes. If desired , you can try this with a
34 l:tad 1 h5 friend (using two d ifferent issues of Infor
35 1:1d4 "g5 matr) - in this case each of you in tu rn
36 h2 a5 performs the role of the trainer.
37 .:te2 axb4 However, the best way of training the rapid
38 axb4 e7 perception of position proved to be the
game which I will now describe. U nfortu
39 l:te4 f6
nately, it cannot be played without a trai ner
40 l:tf2 "d5
or partner and without a previously pre
41 1:[e8+ h7 pared selection of special exercises. (How
Wh ite resigned . ever, now this problem is nevertheless
resolvable with the help of a computer
For the development of intu ition , various training program I have prepared ).
types of training games are usefu l , forcing The clock is set, and you have , say, 1 5
you to ta ke a decision quickly, without minutes to the time contro l . During this time
careful thought. you have to fi nd the correct solutions to five
At the fi rst session of our school, g randmas d ifferent positions. The first position is set
ter Yusu pov recommended that you should up on a board , and the clock is started . After
play 'guess the move'. Take a good game by taking a decision, you make a move on the
a grand master, with detailed comments by board and stop the clock. The second
him, and after the opening start trying to position is set up, and so on. All five
guess his moves, al lowi ng you rself very positions have to be solved before your flag
little time for the whole game (for example, falls. The exercises (positional or tactical )
half an hour). Then compare your sugges are not too d ifficult, a n d do not demand
tions with the g rand master's moves and his deep calculation . Some of them are on the
comments. easy side, some are rather more d ifficult.
The Development of Chess I ntuition ltJ 51

You must use your time in the most under such rules you can win even if you
economical way, to avoid reaching the last make one mistake . With two mistakes, this
exercises already in severe time-trouble. is u n l i kely (too l ittle time for thought re
But it is dangerous to play too qu ickly - it is mains) and with three mistakes it is simply
easy to make a stupid mistake. You win, if impossible.
you correctly solve all five exercises - Play stops as soon as the time li mit is
otherwise you fail to a g reater or lesser exceeded . It is also possible to win 'ahead
extent. of schedule' - if for the last one or two
Another form of the same game, which I in positions you have a time reserve which is
fact used with Dolmatov, Yusupov and other g reater than the possible penalty for an
grandmasters whom I was training, is even incorrect but instant answer. I n this case it is
more effective . Slightly more time is al no longer necessary to solve them.
lowed : 20-25 m i n utes ( 1 5 m i n utes only for But play can also be continued after losing
grandmasters and strong masters). We play on time - u ntil you have gone through all
in exactly the same way, but if an exercise is five positions. It makes sense to do this if
solved incorrectly the clock hand is ad the ru les of the game envisage (with the a i m
vanced by one third of the i n itial time o f raising t h e seriousness and responsibil ity
reserve (with a 1 5-minute control - by 5 of the decisions take n ) some kind of
minutes , with a 20-min ute control - by 6% 'penalty' for a loss, depending on the
min utes, and so on). Success i n the 'series' nu mber of additional m i nutes used .
means getting through all five positions
without losing on time. You will see that Now try solving one such 'series' .

Exercises

1. White to move 2. Black to move


52 The Development of Chess I ntuition

3. Wh ite to move 4. White to move

5. White to move
The Development of Chess I ntuition lD 53

Sol uti ons

1 . Smyslov - Gurgenidze (34th USSR He also has the advantage after 21 . . . 'ii'x b4
Championsh i p , Tbilisi 1 966/67). 22 axb4 lLlf6 23 e3 lIe7 24 lLle2 g5 25 i.. c8 ! .
4S h4! I n the game there followed 21 . . . aS!? 22
The black pawn must be fixed on the 'ii'x bS lLlxc3 23 'iVxc4 dxc4 24 bxc3 l:.ab8
vul nerable h5-sq uare , in order then to 2S i.. d 71 l:1e7 26 i.. a 4 i.. d S 27 g4! g6 28 f3
attack it with the bishop, and possibly create fS 29 gxfS gxfS 30 f2 f7 31 g3 f6 32
a dangerous passed pawn on the h-file. It is f4 i.. f7 33 1:[g1 i.. g 6 34 h4, and Wh ite
hardly possible simultaneously to hold two converted his extra pawn .
weaknesses - on a7 and h5. Wh ite is bound
to win . 4. M i les - Makarychev (Oslo 1 984) .
I n the game there followed 4 S dS? h 4 ! 46 Wh ite would like t o attack t h e opposing
i.. e2 lLlf8 47 e4 (if the a7-pawn is queenside pawns with his q ueen , but fi rst he
captu red , Black shuts the king in the corner must suppress the opponent's counterplay
by . . . c7) 47 ... gS 48 dS f6 49 i.. g 4 on the kingside. 37 'ii'c6? i.. x g3 38 xg3
lLlg6 with an obvious draw. After a passed g 'ii'g 1 + would be premature. 37 lLlf1 ? 'iVb2 is
pawn is created, Black can give u p his pointless, while if 37 lLle2? there follows
kn ight for it, if his king is then able to retu rn 37 . . . 'ii' e 1 ! , and 38 'iVxe5+?? loses to
to b8. 38 . . . i.. f6 .
37 lLlh1 ! !
2. lohlesen - Belavenets (correspondence Threatening to gain an advantage b y 38
1 974-79). 'ii'c6. For example, 37 . . . i.. e 7 38 'iVc6 i.. d 6
2S . . . :t8xe6! 39 lLlg3, intend ing h4-h5 . In the game there
26 dxe6 'it'f3 ! ! followed 37 . . . 'iVb2 38 'iVc6 'iVb1 ? (38 . . . i.. g 5!
White resigned . 39 f3 'iVb 1 40 lLlg3 'iVd 1 + 4 1 g2 'ii'd 8
Usually the answer consists of just one was necessary) 39 'iVxc7 'iVe4+ 40 h2 hS
move , but sometimes the solution contains 41 'ii' c 6 'iVc2 42 gxhS 'iVfS 43 'iVg2 ! ? 'ii'x hS
44 cSI , and Wh ite won .
two or more moves. I n such cases I make
my reply, again press the clock butto n , and
so on, u ntil the entire solution is reprod uced S . P i nter - Larsen ( I nterzonal Tournament,
on the board . Las Palmas 1 982).
Wh ite is planning action on the kingside.
3. Beliavsky - C hern i n ( I nterzonal Tourna However, the hasty 25 f2? ru ns into the
ment, Tu nis 1 985). exchange sacrifice 25 . . . .l:.xg5! 26 fxg5 lLlg6,
In the event of 2 1 'iVxc4 dxc4 Black would when the position becomes unclear.
not stand badly. 2S i.. h41
21 'iVb4! The threat of the exchange sacrifice is
Wh ite has in mind 21 . . . lLlxc3 22 %:txc3 (22 neutralised . If 25 . . . lLlg6 Wh ite has 26 i.. f6 ,
'iVxc3 !?) 22 . . . 'iVxb4 23 axb4 llxe2 24 %:tc7 and otherwise he plays f2 and i.. f3 ,
1:lb8 25 l:Ixb7 and wins, or 22 . . . 'ii'xe2 23 concentrating his forces on the kings ide and
l:!.c7 l:Iab8 24 'iVxd6 with strong pressure. preparing g3-g4.
54 The Development of Chess I ntuition

25 . . . e8 26 f3 1tJd7 27 'itf2 g6 28 l:. h 1 INS TRUCTIONS


'it f7 2 9 g5 " a 5 3 0 g 4 1 'itg8? ( 3 0 . . . fxg4
was more tenacious) 31 f6 .l:tf7 32 gxf5
Recommendations for exercises aimed
exf5 33 h51 1tJxf6 34 exf6, and Black had
at the development of intuition
no defence.
1 . Carefu lly fol low your feeli ngs and try
Experience has shown that, if it is taken as often as possible to predict the reply
seriously, such training is exceptionally beforehand. To learn to guess, you must
usefu l , simultaneously developing several constantly try guessing .
skills that are important for a chess player: 2. Don't be restricted by you r fi rst
It improves i ntuition, the abil ity to qu ickly i mpression - follow the changes in your
and correctly grasp both the tactical and the feeli ngs as your delve i nto the position.
strateg ic details of the position . The truth can be sensed at various stages of
It cu ltivates the procedure for consider the th i n king process. Even so, try to do this
i n g a move - the habit of i m mediately as soon as possible.
determining the candidate possibil ities avail 3. After ascertai n i ng the objective truth,
able, and also the opponent's main th reats . don't forget to compare it with your
Without this, success when playing is not guesses. It is usefu l to record which ideas
possible - with time restricted, errors will be and rules were the most importa nt and
inevitable. decisive for the given position , and to what
It develops resol uteness. There is simply extent you took account of them in you r
no time for the careful checking of variations prel iminary assumptions.
- you have to trust yourself and boldly take
4. A very wide range of i ntu itive feeli ngs
decisions.
is possible. Not necessarily the best move ;
It helps i n the battle with time-trouble, perhaps some evaluative considerations,
since you constantly have to mon itor the the desirability of this or that operation,
expenditure of time. sense of danger, and so o n .
It assists the acq u i ring of good form
5. Comparative assessments are usually
before an event. Your reactions and q u ick
more valuable than absol ute ones. Con
th i n king are improved , without emotional
clusions such as 'the position is drawn ' or
fatigue setting in, since this type of game is
'the opponent is hopelessly placed ' are
lively and spontaneous.
rather crude and are often no help when
seeking a solution . More subtle conclu
I n conclusion I should like to un ite the main sions, relating to a comparison of d ifferent
ideas expressed in this lectu re i nto a kind of moves , plans and ideas, possible pros
instruction gu ide for independent work in pects , eval uation of d ifficu lties and dangers
this field . on the path to the goal, are fa r more
important.
6. In your calcu lation ta ke i nto account
not only purely chess factors , but also
competitive ones. Tournament position ,
reserve o f time a n d strength , opponent's
personality, the probabil ity of him making
mistakes, and so o n .
The Deve lopment of Chess I ntuition lZJ 55

7. 'Meta-intuitive' decisions are very positions wh ich do not lend themselves to


important. For example, can you trust your accurate calculation . Devise training exer
intuition in the g iven instance; does the cises and games which demand that you
position lend itself to precise calculation and take intu itive decisions. It possibly makes
how advisable is such a calculation ; how sense to play games with a shortened time
much time should you supposedly spend on control, study the games of intu itive players ,
considering a move. and so o n .
8. Analyse your actions. If necessary, 1 0. Don't expect an i m mediate resu lt, but
correct the recommendations given and remain confident about your ultimate
work out new rules. success. Purposeful actions in this d i rec
9. Look for topics and ways of working tion will defi n itely held to develop your
on chess that have the maximum effect i ntuition . As a result your play will become
on the development of i ntuition. Try more spontaneous, confident, rapid and
quickly guessing the reply in comparatively assured .
simple situations, and, by contrast, i n

Adden d u m

Kasparov - Karpov
World Championship Match ,
Moscow 1 984/S , 6th Game
Queen 's Indian Defence
1 d4 liJf6 2 c4 e6 3 liJf3 b6 4 g3 i.a6 5 b3
i.b4+ 6 i.d2 i.e7 7 i.g2 0-0 8 0-0 d5
9 liJe5 c6 10 i.c3 liJfd7 11 liJxd7 liJxd7
1 2 liJd2 ':c8 13 e4 b5
The consequences of 1 3 . . . cS were exam
ined by Artur Yusu pov at the second session
of our school - cf. the game Yusupov-Sax,
Rotterdam 1 989, which is analysed in
Secrets of Opening Preparation p.4S.
1 6 cxb5?
Incidentally, our analysis of the clash be
tween Kasparov and Karpov will be based The fi rst (but by no means last) occasion
on a deep analysis by Yusu pov, published when Kasparov's positional feeling let him
the day after the conclusion of the game in down . 1 6 cS! would have led to an advan-
the newspaper Sovietsky Sport. tage for Wh ite : 1 6 . . . liJa4 1 7 'ii'c2 (with the
threat of 1 8 eS! ) 1 7 . . . eS 1 8 liJb3 (Karpov-
1 4 :e1 dxc4
Van der Wiel, Amsterdam 1 986), or 1 6 . . . b4
1 5 bxc4 liJ b6?! 1 7 i.b2 liJc4 1 8 liJxc4 i.xc4 1 9 'ii'c2 i.bS
1 S . . . bxc4 is better. 20 a3.
56 The Development of Chess I ntuition

16 . . . cxbS then either 26 "eS "f8 27 ..te4 , or 26


1 7 lIc1 ..ta3 ':e3!? lLlc3 (26 . . ...d6 27 ..th3 ':f8 28 ..tfS )
1 7 . . . b4 !? was also not bad . 27 ..txc3 "xb3 (the bishop is invulnerable
because of the weakness of the 8th rank) 28
1 8 lIc2 lLla4
..te4! lIxc3 (28 . . . g6 29 "eS) 29 ..txh7+ f8
1 9 ..ta1 ':xc2 30 ..td3! and wins.
20 'iVxc2 'iVaS?1 2S . . . 'iVxd4
20 . . .'iVe7! was stronger, not only preparing 26 lLlxd4 lLlxa2
21 . . . lIc8, but also preventing d4-dS.
26 . . . ..tf8 and 27 . . . b4 was more cautious.
21 "d 1 !
I n the event of the immed iate 2 1 dS Wh ite
would have had to reckon with 2 1 . . . lIc8 22
'iVd3 (22 'iVd 1 lIc1 ) 22 . . ...tf8 or 22 . . . lLlb2.
He wants to play lLlb3, and only then d4-dS.
21 . . . :tc8?
Black should have chosen between 21 . . . ..tb2
22 lLlb3 "b4 and 2 1 . . . lLlc3 22 ..txc3 ! ? (22
lLlb3 'iVb4 23 "c2 nc8 , but, of cou rse, not
23 . . . lLlxa2? 24 %1e3) 22 . . ... xc3 23 lLl b 1 "as
24 lLlxa3 "xa3 2S "b3 (2S dS!?) -
however, i n both these cases Wh ite would
also retain somewhat the better chances.
22 lLlb3 "b4
23 dS exdS Yusu pov showed that 27 lLlfS! would have
24 exdS lLlc3 led to a decisive advantage for Wh ite. I n
reply 2 7 . . . g6? 28 d 6 gxfS 2 9 d7 is bad for
Black. If 27 . . . ..tf8 , then 28 d6 is again very
strong , for example, 28 . . . l:td8 29 lLle7+ h8
30 ..teS (30 lLlc6 ':'xd6 3 1 ':'e8 is also good )
30 . . . lLl b4 3 1 .11 e 4 .
If 27 . . . ':'c1 , then 28 %bc1 ..txc1 29 d6 ..tgS
30 h4 ..tc8 (30 . . . ..td8 3 1 lLle7+ f8 32 lLlc6
e8 33 ..t h 3 ! ) 31 hxgS ..txfS 32 ..tc6 f8 33
..td4 lLlb4 34 ..txbS.
I nterposing 27 . . . ..t b4!? is more tenacious.
White simply replies 28 :e2. The exchange
sacrifice 28 ... lLlc3 29 ..txc3 ..txc3 30 lLle7+
f8 31 lLlxc8 ..txc8 does not save Black -
he also loses a pawn after 32 d6 ..tb4
(32 . . . b4 33 lIe7; 32 . . . ..te6 33 ..tdS) 33 ..tc6
2S .d4?! ..te6 34 d7 e7 3S ..txbS.
I th i n k that it was not essential to exchange It remains to check 28 . . . lLlc1 29 lIe4.
the q ueens - playing d i rectly for an attack a) 29 . . . lLlb3 30 ..tb2! (the tempting 30
with 2S 'iVhS! was stronger. If 2S . . . lLlxa2, ..txg7? ! , hoping for 30 . . . lIc1 +? 31 ..tf1
The Development of Chess I ntuition 'ZJ 57

l:txf1 + 32 g2 ! , allows Black to hold on by .i.b7?! 2 9 d 6 .i.xc6 3 0 .i.xc6 l:txc6 3 1 d7


31 . . . lLld2! 32 .l:txb4 .l:tc1 + 33 f1 .i.c8 ! ) .i.e7 32 .i.f6! gxf6 33 .l:txe7 l:td6 34 .l:te8+
3 0 . . . .i.f8 3 1 d6 lLlc5 3 2 lLle7+ xe7 3 3 g7 35 d8'it' 1:.xd8 36 .l:txd8 a5 37 f1
l::tx e7 with a won position; Black is in danger of losing the resulting
b) 29 . . . lLld3 30 f1 f8 3 1 lLle7+ (there is endings) 29 l::t a 1 lLl b4 30 lLlxb4 .txb4 with
another way to the goal: 3 1 d6!? lLlc5 32 eq uality;
lLle7+ .i.xe7 33 .l:txe7 .l:td8 34 .i.c3 ! ) 28 lLle7+ (probably the most dangerous try)
3 1 . . . .i.xe7 32 l::t x e7 b 4 33 d6 with the th reats 28 . . . .i.xe7 29 ':xe7 b4! (29 . . . .l:tc1 +? 30 .i.f1
of 34 .i.xd3 .txd3 35 d7, 34 l:txa7 and 34 f8 31 ':xa7 is bad for Black) 30 h4! lLlc3 3 1
l::te 3; d 6 (3 1 ':xa7 .i.c4 3 2 d6 lLlb5) 3 1 . . . .i.b5 32
c) 29 . . . .tf8 30 d6 b4 (in the event of ':'xa7 l:r.d8! (32 . . . f8? 33 .l:tb7! with the
30 . . . lLlb3 3 1 d7 l:tc1 + 32 .tf1 .l:td 1 both 33 th reat of 34 d7 and 35 .i.xc3 ; 32 .. J:tb8? 33
.i.d4 and 33 lLl h6+ wi n ) 31 lLlh6+! gxh6 32 d7 l:td8 34 .i.h3 with advantage to Wh ite) 33
1:.g4+ .i.g7 33 .l:txg7+ f8 34 .i.d5 lLle2+ 35 l:[b7 .i.e8 34 ::'xb4 lLlb5 35 .i.e5 f6!
g2 lLlc3 36 .l:txf7+ g8 37 .txc3 bxc3 38 (35 . . . f8 36 .i.d4 ! , and Black has a d ifficult
l:txa7+ f8 39 l:txa6 c2 40 d7 c1 'iV 41 position) 36 .i.d5+ f8 (weaker is 36 . . . .i.f7
dxc8'it'+ 'it'xc8 4 2 .l:ta8. 37 ':xb5 fxe5 38 .i.e4) 37 .i.b2 ':xd6! with a
Now let us see what happened i n the game. draw.

27 lLlc6? .i.c51 28 . . . :a8


If 27 . . . f8? the simplest solution is 28 .i.d4! 29 d4 .i.xd4
(with the threat of 29 l:ta 1 ) 28 . . . lLlc1 29 d6. 30 lLlxd4 f8
27 . . . .i.d6? is also a mistake in view of 28 31 d6
i.e5! l:te8 29 l::t a 1 .txe5 30 .l:txa2 b7 31 1: a 1 lLlb4 32 lLlc6 lLlxc6 33 dxc6 .i.c8 is
(30 . . . .i.c8 31 .l:te2 f6 32 f4 .i.g4 33 l:te4) 3 1 advantageous to Black.
l:txa7 .i.xc6 3 2 dxc6 f8 3 3 .i. h 3 ! .l:te7 34
i.d7, and Black has to g ive up his bishop for
the mighty pawn .
28 .i.h3?1
A natu ra l , but in fact dubious move. It soon
transpires that the bishop has moved onto
an inferior d iagona l , whereas the black rook,
on the contrary, moves to a better positio n .
However, White no longer h a d a win:
28 .i.d4 .i.xd4 29 lLlxd4 (29 lLle7+? f8 30
ltJxc8 .i.xc8 31 d6 .i.f6 or 31 . . . .t c5 32 .tc6
i.e6) 29 . . . ':c1 !? 30 .l:txc1 lLlxc1 3 1 d6 f8
32 d7 (32 lLlc6 .i.c8 ! ; 32 .i.h3 .i. b7 ! ; 32 .i.c6
g6! ) 32 . . . e7 33 lLlc6+ xd7 34 lLl b8+ c7
35 lLlxa6 b6, and the two connected
passed pawns fully compensate Black for 31 . . . lLlc3 !
the lost piece; It was already possible to secure a draw by
28 .i.e5 .l:te8! (after both 28 . . . lLl b4?! 29 d6 3 1 . . . .l:td8 32 d7 .tb7 33 lLlxb5 c6. But
ltJd3!? 30 .l:te2 lLlxe5 3 1 lLle7+ f8 32 lLlxc8 Karpov keen ly sensed that as a result of the
i.xc8 33 l:txe5 .i.xd6 34 .l:txb5 and 28 . . . opponent's uncertain actions the situation
58 The Development of C hess I ntuition

had changed in his favour, and he decided to 34 l:ta 1 SLxc6 35 xc6 1:.e6 36 1:.xa 7 was
play for a win . more tenacious, although the endgame after
Kasparov, on the other hand , d i d not sense 36 . . . l:txd6 37 d7 1:.b6 can hardly be held.
the impending danger. He should have 34 . . . f6 1
forced a draw by choosing 32 SLg2 ! l:td8 33 35 d7
c6 (with the threat of 34 d7) 33 . . . SLc8 34 There is no longer any way of saving the
liJxb5. game: 35 SLxb7 l:txe5! 36 1:. a 1 b4 37 l:lxa7
It was also possible to play 32 d7 SLb7! b3 and 35 liJd7+ 'iti>f7 36 1:.a1 SLxg2 37
(defending against 33 liJc6 or 33 g2) and 'iti>xg2 'iti>e6 were equally bad .
now, for example, 33 liJf5 .l:r.d8 34 ':e8+ (if 35 . . . lld8
34 liJxg7? or 34 liJd6? there is the simple 36 xb7 fxe5
34 . . . c6) 34 .. Jbe8+ 35 dxe8'iV+ 'iti>xe8 36
37 c6 'iti>e7?
liJd6+ 'iti>d8 37 liJxb7 'iti>c7 - the two black
pawns are at least as strong as the piece. A time-trouble mistake. There was an easy
The unexpected move 33 l:t a 1 ! , found by win by 37 . . . e4! 38 lla 1 'iti>e7 (38 . . . liJe2+ 39
Vadim Zviagintsev, is safer. The point is that 'iti>f1 liJd4 is also possible) 39 llxa7 'iti>d6 40
if 33 . . . a6 there follows the pretty stroke 34 l:ta6 'iti>c7 41 'iti>f1 b4 .
liJc6! SLxc6 35 llxa6. The interesting try 38 .ltxb5! liJxb5?!
33 . . . a5 encounters the intermed iate move Ka rpov d id not have sufficient time to check
34 lla3! (weaker is 34 ':'xa5 'iti>e7) 34 . . . b4 35 the variation 38 . . . 'iti>d6! 39 SLd3 1:.xd7! 40
:txa5. Black does best to agree a draw after SLxh7 a5.
33 . . . 'iti>e7 34 :e1 + 'iti>f8 (but not 34 . . . 'iti>d6? 39 llxe5+ 'iti>xd7
35 1:.e8 'iti>c7 36 liJc6 ! ) 35 ':'a 1 . Of course, 40 l:txb5 'iti>c6
the order of the moves can also be changed :
41 1:.h5?
32 1:. a 1 SLb7 33 d7.
41 11e5! was stronger, and if 41 . . . 1:.a8, then
32 liJc6? b7!
42 1:.e6+ 'iti>c5 43 lle7 a5 44 1:.xg7, also
A draw results from 32 . . . b4 33 d7 (or 33 attacking the h-pawn .
liJxb4 SLb5) 33 . . . b3 34 d8'iV+ .uxd8 35 liJxd8
41 . . . h6
SLd3.
42 11e5 lla8
33 SLg2 :e8 !
The sealed move. 42 . . . 1:.d5 was also good .
Possibly Kasparov was hoping for 33 . . . b4? The ending is rather i nteresti ng, but here we
34 d7 b3 35 liJb8! 1:.xb8 36 SLxb7 b2 will cut th ings short, since from this point it
(36 . . . l:td8 37 c6) 37 c8, and White wins. was a contest not in the abil ity to fi nd the
But Karpov's sense of danger is equal to the strongest conti nuations at the board , but in
occasion . quality of adjournment analysis. Black won
34 liJe5 on the 70th move.
ltJ 59

Serg ey Dol matov

I n J azz Style

I games, in which a tense battle developed


should like to show you a few o f m y own Dolmatov - Lerner
All-Un ion Qualifying Tou rnament,
from literally the first few moves - already in Daugavpils 1 978
the opening or at an early stage of the Philidor Defence
middlegame. In them there was a sharp and 1 e4 eS
usually a very unconventional battle for the
2 lZJf3 d6
initiative.
3 d4 exd4
All the games were played many years ago,
4 tLlxd4 lZJf6
when I was making my fi rst steps i n the
world of top-class chess - this is no S lZJc3 i.. e 7
accident. Youth is typified by an absence of I suspect that this was already the extent of
stereotypes, by optimism, and by a bel ief i n my theoretical knowledge. But this factor did
one's own powers (sometimes excessive, not bother me: after all, in the resulting
involving an underestimation of the oppo position it is not hard to make common
nent). I nteresting, vivid games often result sense moves. For the moment a sound
from fl ig hts of imag inatio n , not burdened by cou rse can be followed : develop the pieces
experience and knowledge, from in ner and fight for the centre - there are no
freedom , not stifled by rules. With the yea rs, dangers to be feared .
alas, this 'flippancy' is usually lost. 6 ..ie2 0-0
The abil ity to think unconventionally is an 7 0-0 l:teS
important qual ity for achieving victory over a 7 . . . tLlc6!?
strong opponent. This is d ifficult to learn and S f4 ..ifS
probably impossible to teach . Try to develop
9 ..if3
this ability in you rself, by analysing the early
games of those outstanding players who Wh ite has obtained a strong centre . Even
made a name for themselves at a young now I have no complaints about his open ing
age. Players such as Boris Spassky, M ikhail strategy.
Tal and Alexey Shirov. . . Their ideas evoke 9. . . tLla6
naivety and spontaneity; they were gener 1 0 l:te1 c6
ated not in the q u iet of thei r study, but If 1 0 . . . lZJc5,then 1 1 lZJb3 tLlxb3 1 2 axb3 is
directly at the board . At times they do not good .
stand up to strict mathematical verification , 1 1 ..ie3
but it proved so d ifficult to refute these ideas
that the opponents went wrong .
I have taken the l i berty of comparing this (see diagram)
easy, improvised playing manner with jazz,
a type of music which is still popu lar today.
60 In Jazz Style

I placed my rook on e1 so that if 1 1 . . . lDc5 I 1 5 'iVxd4 i. c5 and was hoping to exploit the
could defend the e-pawn with the simple pin on the g 1 -a7 d iagonal (it is not clear,
bishop retreat 12 i.f2 . After 1 2 . . . lDe6 1 3 however, whether this is possible after 1 6
'iVd2 Wh ite brings his queen's rook to d 1 'iVd2 ) . Konstantin Lerner clearly underesti
and only then begins th inking about his mated my reply.
fu rther plans: whether to break through in 1 5 i.xd5!
the centre with e4-e5, or prepare a pawn
Now both pieces are invul nerable in view of
offensive on the kingside with h2-h3 and
1 6 xf7+, and 1 5 . . . 'iVxf6 1 6 lDe4 is bad for
g2-g4 .
Black. I n order to defend his queen , he must
My opponent did not want to defend
develop his bishop from c8 , but where? It is
patiently and he decided to beg in an
immed iately clear that any bishop move has
immed iate battle in the centre.
its drawbacks: 1 5 . . . i.d7 1 6 'iVh5 g6 1 7
11 . . . d5?! i.xf7+, 1 5 . . . i.e6 1 6 l:1xe6, or 1 5 . . .i.g4 1 6
12 e5 c5? 'iVxg4 dxe3 1 7 i.xb7.
It would have been better for Black to 15 . . . i.f5
restrict hi mself to the modest knight retreat 1 6 :e5 i.g6
1 2 . . . lDd7.
have removed one of my pieces from
Can you believe in the success of Black's
attack, but it is more d ifficult to deal with the
mil itary operation, begun with his bishop on
second - any knight move is answered by
c8 and his knight on a6? You can't? Then
1 6 . . . gxf6. However, as was shown by
you have to find a refutation .
Dvoretsky, it was nevertheless possible to
1 3 exf6 l:[xe3 play 1 7 lDe4 ! , since if 1 7 . . . gxf6 there is the
The 'point' of my opponent's idea! pretty stroke 1 8 11e8 ! .
14 ':'xe3 cxd4 1 7 fxg7 'ii? x g7
I n the event of 1 7 . . . i.xg7 1 8 'iVxd4 it all
(see diagram)
immed iately becomes clear.
For the moment I am the exchange u p , but I 1 8 lDe4! f6
have two pieces en prise. If either of them 1 9 'ii'x d41
should be captu red, the material advantage
will pass to Black. He was only expecting (see diagram)
In Jazz Style ttJ 61

23 . . . liJd7
24 'ith1 !
In such cases variations should already be
calculated to the end . To make things
easier, I recommend that you beg in your
calculation with moves to which the oppo
nent has only one reply. Thus the queen
check on e6 looks tempting, but you will
have to analyse not only 24 . . . 'itg7, but also
24 . . .liJf6 and 24 . . . 'iVf6 , and it is possible to
overlook . . . 'iVb6+. The king move, renewing
the threat of .l:!.e3, does not leave the
opponent any choice.

A pictu resque position! The centre is com 24 . . . .i.c5


pletely occupied by wh ite pieces . After the 24 . . . exf4 25 'ii'xf4+ is completely bad for
capture of the rook, even if a mate can not Black.
be found, Wh ite will later regain the material 25 .l:!.d1 !
by captu ring the b7-pawn with his bishop. The rook manoeuvre to h3 is agai n threat
19 . . . .i.xe4 ened , but there is the additional possibil ity
If 1 9 . . . fxe5 20 'iWxe5+ 'it h6, then either 21 g4 of captu ri ng the b7-pawn with gain of tempo.
or 21 liJf6 . An interesting variation was 25 . . . liJf8
found by Dvoretsky: 1 9 . . . liJb4!? 20 .l:!. d 1 26 .i.xb7
liJxc2 2 1 'iWf2 fxe5 22 ii'xc2 exf4 23 'it'c3+
Black resigned .
'it'h6 24 nd3 (24 liJf6 ! ? ) 24 . . . llc8 25 l:th3+
i. h5 26 'iVd2 'iVxd5! (the only defence
against the mating threats) 27 'ii'x d5 l:[c1 + The following example, like the previous
28 'it>f2 ':c2+ 29 'iVd2 ! (otherwise it is not one, is a fai rly lig ht-hearted game, with the
possible to h ide from the checks) 29 . . . .l:!.xd2+ rapid crushing of the opponent's position .
30 liJxd2, and Wh ite should be able to I ncidentally, don't be su rprised that I am not
convert his exchange advantage. showing you any of my lost games. Of
liJc5 course, they should be carefully studied, to
20 'iVxe4
disclose the causes of the mistakes made,
My opponent was counting on this intermedi
but at the moment why should I spoil my
ate move. 20 . . .fxe5 was hopeless: 21 'ii'x e5+
mood by remembering failures?
'ii'f6 (2 1 . . . 'it g6 22 'iWe6+) 22 ii'xf6+ 'itxf6 23
.i.xb7.
21 'ii'f3 fxe5 Dolmatov - Franzoni
22 'ii'g 4+ World J u n ior Championsh ip, G raz 1 978
Sicilian Defence
Black has nevertheless won a piece, but his
1 e4 c5
king can not escape from the mati ng attack.
2 liJf3 e6
22 . . . 'ith6
23 11e1 ! 3 d4 cxd4
All Wh ite's forces must take part in the 4 liJxd4 liJf6
assault! He is threatening both 24 l:[xe5 and 5 liJc3 liJc6
24 J:te3. 6 .i.e2
62 I n Jazz Style

This is rarely played (the usual continua of the game.


tions are 6 4Jdb5 and 6 4Jxc6 bxc6 7 e5). In 10 cxd4
offering to go into the Scheveningen (6 . . . d6), 1 0 iLxe4!? came into consideration, but I
White allows the bishop move to b4 , after was hoping to transpose into my analysis
which he has to sacrifice a pawn. I analysed after 1 0 . . . d5 1 1 iLa3.
this sharp variation with my trai ner Mark
10 . . 4Jf6
Dvoretsky and then I successfully employed
.

it a couple of times. I don't know why no one Wh ite is a pawn down , and for the moment
plays this now - in my opin ion , here Wh ite he also has no attack, but he has the two
obtains a very promising position . bishops and a defi n ite advantage in space
and development. I n add itio n , as I recall,
6... iLb4
there was a healthy optimism, a confidence
7 0-0 iLxc3 in my powers, which is of no small impor
8 bxc3 4Jxe4 ta nce in such situations. However, such a
9 iLd3 position is one that I would also happily play
now. Wh ite's in itiative is enduring , and it is
not easily neutral ised .
1 1 iLg5 'ii'a 51?

9. . . 4Jxd4
This was the extent of my knowledge. I
knew that 9 . . . 4Jxc3?! was dangerous in view
of 1 0 'ii'g 4 or 1 0 'ii'e 1 and I had only 12 f4!
analysed 9 . . . d5. I was aware of only one
A non-routi ne decision (with the bishop on
game on this theme, Geller-Khasin (25th
g5 it is not usual to place the pawn on f4),
USSR Championsh ip, Riga 1 958), which
but apparently the correct one. White
continued 1 0 iLa3 'ii' a 5 1 1 'ii' c 1 4Jxd4 1 2
should not hu rry with the exchange on f6 . By
cxd4 iLd7 1 3 ':' b 1 iLc6 1 4 iLb4 'ii'c7 1 5 'ii' a 3
advancing his f-pawn , he i ncludes his king's
a5 1 6 iLxe4 dxe4 1 7 c4 f6 1 8 iLd6 'ji'd7 with
rook in the attack. In the event of 1 2 . . . 'i!Vb4 1
roughly equal chances. I don't remember
would have g iven up a second pawn by 1 3
exactly how I was intend ing to improve
f5 .
Wh ite's play, but there was a way - you can
look for it you rself. Later I successfully 12 . . . b6
employed it against Sergey Gorelov, but, 1 3 iLxf6
unfortunately, I have not retained the score But now is an appropriate moment for the
In Jazz Style ltJ 63

exchange - thanks to it Wh ite will be able to the enemy q ueen from the long d iagona l .
gain a tempo by 1 4 Wf3 . 2 0 c41 Wxc4
13 . . . gxf6 21 fxe6 dxe6
1 4 'it'f3 lIbS 21 . . . Wxe6 was more tenacious.
14 . . . Wd5 1 5 Wxd5 exd5 1 6 1:[ae 1 + ..ti>f8 1 7 22 Wf41
Iif3 would have led to an inferior endgame A double attack on f6 and b8. But couldn't it
for Black. For the moment he is not ready so have also been made without the d iverting
openly to sound the retreat. pawn sacrifice?
1 5 f5 b7 22 . . . l::t b7
1 6 ..te4! 23 ':'c1 !
It is important to kill the opponent's hopes This is the point! Now all my pieces are
associated with counter-pressure on the g2- participati ng in the attack. Wh ite's threats
point. With just the heavy pieces on the are irresistible.
board , Black's position is d ifficult, since his 23 . . . 'it'd5
king is under attack and his rooks are
24 Wxf6 :e7
separated .
25 'ili'hS+
16 . . . xe4
Black resigned .
1 7 Wxe4 Wd5
1 S Wh4 ':gS
Dolmatov Flesch
For me there is someth ing mysterious about
-

this game: all the time Black seemi ngly Bucharest 1 98 1


makes good , logical moves, but his position Caro-Kann Defence
imperceptibly becomes hopeless. Why this 1 e4 c6
happens, where the defence can be im 2 d4 d5
proved , I myself do not know! 3 exd5 cxd5
1 9 :f2 Iig5 4 c4
How should Wh ite conti nue his offensive? Against the Caro-Kann I employ only the
Panov Attack, and qu ite successfully - I
have already scored nu merous wins with it.
4. . . liJf6
5 liJc3 e6
6 liJf3 ..tb4
7 ..td3
I n my game with Andrey Kharitonov (qualify
ing tournament for the World Ju nior Champi
onsh ip, Sochi 1 978) 7 cxd5 exd5! was
played . At that time the theory of the 6 . . . b4
variation was only just being developed ,
and the recaptu re on d5 with the pawn came
as a su rprise to me. I won a good game, but
from the opening I had noth i n g . From then
The pressure on g2 is restricting my forces . on I began playing 7 .td3, transposing into
I t is important, even for a moment, t o divert one of the variations of the N i mzo-I ndian
64 In J azz Style

Defence. The resulting positions suit me tioned game against Speelman) Wh ite has
perfectly well , so that I myself cannot an enormous lead i n development, giving
understand why I altogether avoid the q u ite him more than sufficient compensation for
reasonable move 1 d4. the sacrificed pawn .
7. . . dxe4 11 . . . lLl bd7
8 i.xe4 0-0 The opponent is hoping after 1 2 c4 b6 1 3
9 0-0 i.xe3?1 i.g5 i.b7 to arrange his pieces in accord
1 0 bxe3 "ike7 ance with Ka rpov's scheme, but I do not
al low h i m this opportun ity.
1 2 i.a31
An u nusual development of the bishop for
the g iven opening variation. I n his youth a
chess player has less dogma and more
energy - it can be easier for h i m to devise a
fresh idea . When he becomes older, he
already knows exactly what was played
earlier in similar cases, and this knowledge
sometimes prevents an unprej udiced ap
proach to the position .
I have managed to discover the main
weakness in the opponent's position - the
vulnerable d6-square. I ncidentally, also after
the normal development of his bishop at g5,
In the late 1 970s the world champion later White often tries to exploit the same
Anatoly Karpov successfully practised the weakness with the manoeuvre i.g5-h4-g3! .
plan involving the exchange of the b4-
12 . . . ':'e8
bishop for the knig ht, and the development
The usual square for the rook in this
of the knight at d7 and bishop at b7. Janos
variation (Karpov also used to place it here).
Flesch is aiming for a similar set-up, but he
It would have been better to play it to d8, but
carries it out i naccu rately - the premature
the opponent did not anticipate my idea .
exchange on c3 increases Wh ite's possibili
ties. He should have begun with either 1 3 lLld21
9 . . . b6 or 9 . . . lLl bd7. I n cidentally, the knight What to do now? The knight is aiming for d6,
move was made against me by Jonathan and after 1 3 . . . "ikxc3 14 lLlc4 Black is in
Speelman in a game which I demonstrated danger of losing his queen . He should
at the 2nd session of the school (cf. Secrets probably have chosen the cool-headed
of Opening Preparation p . 78). 1 3 . . . b6 , although after 1 4 lLlc4 i.b7 1 5 lLld6
1 1 i.d3! Wh ite has an obvious advantage.
A natu ral and logical move - the bishop was 13 . . . ltd8
under attack. I have to admit that at the time 14 "ikf3
I did not even consider the reply 1 1 . . . 'ilt'xc3 I n the event of 1 4 lLlc4 lLlf8 Black would
(now my optimism has dimin ished some have covered the d6-sq uare, and so I
what, and probably I would nevertheless try activate my queen , fi nally defending the c3-
to calculate it). After 1 2 i. f4! (but not 1 2 i.g5 pawn and preventing the development of
lLlbd7, transposing i nto the afore-men- the bishop at b7.
In Jazz Style ttJ 65

14 . . . lLlf8 1 8 h4 f5?!
What would you now play as W hite? It is probable that many would have played
this - it is hard to endure such intense
pressure for long. Even so, it would have
been better to be patient, and refrain from
weakening the position.
19 c2 lLlb6
20 b3 d7
21 g3
Before taking the f5-pawn it is useful to
improve the placing of the bishop. It is
amusing that i n the end it has nevertheless
moved to g3, its lawfu l square in this
variation.
21 . . . 'iVc6
22 'ifxf5 'iVxc3
1 5 lLle4!
23 e5!
A typical idea ! In such cases it is useful to
White defends the d4-pawn and parries the
exchange the opponent's few developed
th reat of 23 .. :ii'xb3, after which 24 'ifg5 is
pieces - then your lead in development is
now decisive .
easier to exploit. An analogy with ice
hockey can be drawn : if a player has to 23 . . .
leave the rink, an advantage of five players The q ueen returns to the defence.
against four is appreciable, but neverthe 24 'ifh5 .l:tac8
less not decisive. If a further pair is How should the offensive be conti nued?
removed , it becomes much harder to de
fend with th ree against four, and with two
against three it is probably almost impossi
ble.
Of course, Wh ite's move was also based on
more concrete considerations; in particular,
he was aiming to h i nder the development of
the bishop on ca . But a knowledge of
general rules, such as the one just men
tioned , usually makes it easier for us to take
a decision and suggests where it should be
sought.
15 . . . lLlxe4
1 6 xe4 lLld7
A dismal spectacle - the opponent's pieces
rush from place to place. He obviously 25 f4!
wants to play 1 7 . . . lLlf6, but of course, I do Usually I find it hard to decide on changes in
not allow th is. the pawn structu re - I prefer to play with the
17 e7 1 1:[e8 pieces. But here I made a pawn move - it
66 In Jazz Style

really is very strong . Wh ite is threatening tious opponents. We crossed swords in the
not only to include his rook via f3, but also to very first round and I was able to q u ickly
play f4-f5. crush him with Black. It was a double-round
25 . . . g6 event, and soon our second meeti ng took
place - with the same result.
25 . . . i.c6 is bad in view of 26 'ifg4! 'ife7 27
i.xe6+, while if 25 . . . tDc4, then 26 i.c2 is
decisive . After the move in the game Wh ite Dolmatov - Larsen
forcibly destroys the opposing defences .
Amsterdam 1 980
26 'ifh4 'ife7 Caro-Kann Defence
27 i. f6 'iff7 1 e4 c6
28 f51 tDd5 Of course, Larsen had no suspicion of how
29 fxg6 'iVxg6 dangerous it was to play this opening
30 lIf3 tDxf6 against me.
Black is forced to give up his quee n , which is 2 d4 d5
equ ivalent to resignation . 3 exd5 cxd5
31 1:.g3 g7 4 c4 tDf6
32 1:[f1 IU8 5 tDc3 e6
33 ':xg6+ hxg6 6 tDf3 i.b4
34 'ifg5 l:.c6 7 i.d3 dxc4
35 'ili'e51 l:.b6 8 i.xc4 0-0
36 g4 l:.b5 9 0-0 a6
37 d51
Black resigned.
Note that, after gaining a material advan
tage, Wh ite did not relax the pressure , but
looked for the most d i rect and energetic way
to wi n . Sometimes in such situations,
feeling that the work has already largely
been done, a player relaxes and beg ins
playing carelessly. As a result the opponent
is able to set up a defence and even gain
cou nter-chances.

The following game (played the previous


year) developed in similar fashion . The
same opening, and the same energetic
exploitation of the opponent's opening inac I didn't know anyth ing about this move.
cu racies. At the time I was an international Later I established that it makes sense to
master, and I was participati ng for the first prevent . . . b7-b5 by playing 1 0 a4 ! ? , as in
time in a strong g randmaster event, whereas the Queen's Gambit Accepted . But at the
Bent Larsen was one of the favourites . time I did not want to weaken the b4-square
Sometimes the experienced g randmaster and I devised another idea.
tended to underestimate young and ambi- 10 a3!?
In Jazz Style ttJ 67

Also prophylaxis against . . . b7-b5, only q ueens have to be exchanged : it is too risky
more refined . In the event of 1 0 . . . iLe7 1 was to play 1 4 . . . :a7 1 5 iLf4 ( 1 5 'iVh5!?)
intending to retreat my bishop to a2 in 1 5 . . . l::t d 7?! 16 'iVh5 with strong pressure on
advance and to meet the flank advance Black's kingside. After 14 . . . 'iVxd5 1 5 iLxd5
1 1 . . . b5 with the central cou nter 1 2 d5! . If lla7 1 6 f4 Black cannot play 1 6 . . . iLb7? 1 7
instead 1 0 . . . iLxc3 1 1 bxc3 b5, then after 1 2 e3, and 1 6 . . J:td7 1 7 iLxf7+ ':xf7 1 8 iLxb8
iLd3 the threat of 1 3 a4 is unpleasant. is also unfavourable. He has to agree to a
Even so, this last variation looks the most permanently inferior endgame by 1 6 . . . e6
log ical reaction to Wh ite's plan. After the 1 7 iLxe6 fxe6 . Even so, this would have
exchange on c3 the move a2-a3 is a waste been the lesser evil : objectively Black can
of time: the pawn should either be left on a2, hope for a draw. 'But why play cautiously
or moved to a4 . At the 1 982 Zonal Tourna against a boy? ' , the g randmaster probably
ment in Yerevan , Lev Psakhis prepared well thought.
for his game with me and went in for this 1 4 ltJxe7+ 'iVxe7
positio n . There followed 1 2 .. :iVd5 1 3 a4 1 5 iLg5
i. b7 1 4 'iVe2 l:tc8 1 5 axb5 axb5 1 6 l:ba8 The two bishops in an open position ensure
i.xa8 1 7 iLd2 ltJe4 1 8 iLxe4 'iVxe4 1 9 'iVxb5 Wh ite an overwhelming adva ntage. I only
i.d5 20 l:te 1 'iVg6 21 'iVe2 ltJc6, and Black have to make natu ral attacki ng moves and
had sufficient compensation for the sacri make sure that the opponent does not
ficed pawn . The game soon ended in a escape from the trap into which he has
draw. fallen.
Of course, at the board , with the clock 15 . . . ltJbd7
ticking away, it is far harder than in home
1 6 :e1 'iVc5
preparation to make a sober assessment of
a position. Therefore if you are able to think 1 6 . . . 'iVd8 was more tenacious.
up a sensible idea such as 1 0 a3, the 17 iLe3
practical chances of it succeed ing are very Of course, not 1 7 l:tc1 ? iLxf3 .
considerable, even if a solution to the 17 . . . 'iVf5
problem facing the opponent does in fact
1 7 . . . 'iVh5 was comparatively better, al
exist.
though after 1 8 ltJg5 'iVxd 1 1 9 l::t a xd 1 Black
10 . . . iLe7 has a d ifficult endgame.
1 1 iLa2 b5?!
1 1 . . . ltJc6 was better.
1 2 d5!
What should Black do now? He does not
want to al low the captu re on e6 - for the
entire game he will have to defend a clearly
inferior position.
12 . . . exd5
1 3 ltJxd5 iLb7?
Black should have exchanged knig hts:
13 ... ltJxd 5 . Larsen was afraid of the reply 1 4
'iVxd5 ( 1 4 iLxd5 i s weaker i n view of
14 . . Jla7 with the th reat of 1 5 . . J:td7). The
68 In Jazz Style

1 8 lLlh4! 24 'ii'd 4!
The q ueen is al most trapped . Of course, the Complete dominatio n ! There is no need to
routine 1 8 lLld4? was weaker because of pick the fru it - it will fal l of its own accord .
1 8 . . ...g6. Black's next move is effectively First deprive the opponent of any sensible
the decisive mistake - only the retu rn of the moves, and then finish him off. It was even a
queen to e5 promised chances of saving the pity for me to make the next few moves, as I
game. wanted simply to enjoy the ideal arrange
18 . . . 'ii'e 4? ment of the white pieces - I am no longer
able to improve it.
1 9 95 'ii'c 6
24 . . . 'ii' b 8
20 l:[c1 'ii' b6
It was not in vain that I had developed my
21 .i.e3
'prophylactic th inking' - I immed iately real
The game has tu rned out to be very ised that Black was i ntend ing 25 . . . .l:.d8 . I
amusing. My dark-square bishop moves had to calculate a wi n n i ng variation to the
backwards and forwards, each time with end (when the opponent's possibil ities are
gain of tempo. so restricted, this is very easy). I n fact it was
21 . . . 'ii'd 8 time to win the point and leave for home.
22 lLlf5 25 f4! ':d8
As you can see, since the 1 6th move only 26 f5 ..th5
Wh ite has been playing. The opponent's 27 h3 lLl b6
q ueen has wandered round the entire board 28 'ii'x b6 'ii'x b6
and finally retu rned to its i n itial square d8,
29 ..txb6 l:.xd6
but during that time I have included all my
pieces in the attack. 30 .i.e3
It is after accurate moves such as these that
22 . . . ..te4
the opponent usually capitulates (after other
23 lLld6 ..tg6
moves by the bishop Larsen would still have
The bishop has moved to the defence of the been able to consider 30 . . . :d2). Black
f7-point. Wh ite's position is won , of course, resigned .
but I suggest you try to find the way that I
found in the game. By now you will probably have gained the
impression that I can win only with Wh ite.
Therefore I will show you a game in which I
had the black pieces.

Van der Sterren - Dolmatov


Amsterdam 1 979
Reti Opening
1 lLlf3 d5
2 b3 g4
3 ..t b2 lLld7
3 . . . ..txf3 would have led to a completely
unexplored position - these I try to avoid.
Black's plan, which had already many times
In Jazz Style
llJ 69

brought me success, is simple: . . . e7-e6, playing without prejudices is typical of


... c7-c6, ... lLlgf6 , . . . d6, . . . 0-0, . . . :te8 and youth . It would seem to be a pity to leave the
at some point . . . e6-e5. opponent with a strong bishop on the long
4 c4 e6 d iagonal. A matu re player would possibly
5 e3 lLlgf6 not have risked such an exchange (which
means he would have rejected 9 . . . e5). I n
6 e2 d6
fact, Black's solid position in the centre and
6 . . . c6 is perhaps slightly more accurate, the slight vul nerability of the opponent's set
when after the exchange on d5 Black can up, which has been weakened by f2-f4 ,
captu re with the c-pawn . secure me good counterplay.
7 cxd 5 1 ? exd5 1 1 xd4
S lLlxd4 In the event of the anti-positional 1 0 exd4+?
A typical idea in the Reti Open ing - the Black gains the advantage with 1 0 . . . 'ii'e 7 1 1
knight is aiming for f5 . 'ii'x e7+ xe7 1 2 a3+ d8 1 3 0-0 l:te8 1 4
S. . . xe2 lLlc3 lLlb8! followed by 1 5 . . .lLlc6 .
9 'ii'x e2 11 . . . c5
What would you now play? 1 2 b2 0-0
1 3 0-0 J:teS
1 4 'ii'd 3?!
It would have been better to place the q ueen
at f3 .
14 . . . 'ii' b 6
Wh ite will soon have to reckon with both
. . . d5-d4, and . . . c5-c4 .
1 5 lLlc3
But what to play now?

Of cou rse, one can simply castle, but after


10 lLlf5 Wh ite will stand a little better. I did
not want to concede the in itiative to my
opponent, and I chose the conti nuation that
was the most critica l , but also slig htly risky.
9 . . . e5!
10 f4
I also had to reckon with 1 0 a3, and if
1 0 . . . c5, then 1 1 f4! c7 1 2 lLlf5. I was
planning 1 0 . . . lLle4 ! . You don't have to look immed iately for any
10 . . . xd4 brilliant idea . First see whether or not the
I must once agai n remark nostalgically, that problem (the d5-pawn is attacked ) can be
70 I n Jazz Style

solved by any normal move that is useful to


you .
15 . . . 1:[ad8!
All my pieces are now in play. After 1 6 lLlxd5
lLlxd5 1 7 'ii'x d5 lLlf6 (or 1 7 . . . lLle5 1 8 'ii'e4
lLld3) Black regains the pawn and stands
better.
1 6 l:tab1
Just in case, my opponent decided to
defend against . . . c5-c4 . But I am ready to
advance not only my c-pawn, but also my d
pawn . See how useful it is to have several
strategic th reats in reserve, without h urrying
to carry out any one of them!
By placing my q ueen on the same d iagonal
16 . . . 'ifc6
as the wh ite king , I have indirectly defended
1 7 l::tf3 d4! the d4-pawn - if 21 .i.xd4 there is the strong
Now is the time! With his last move Paul Van reply 21 . . . .:.xd4 ! 22 lLlxd4 lLle6 23 l:td3 l:td8.
der Sterren demonstrated his desire to At the same time Black avoids the exchange
beg in a flank attack, and I meet it with a of queens, which could have occu rred after
th rust in the centre, which , however, de 21 1:[c1 .
manded accu rate calcu lation. [20 . . . lLlfe4! (with the threat of 2 1 . . .lLlxd2) is
There was also another tempting possibil ity: probably even stronger, and if 2 1 lLlxd4
1 7 . . . lLle4!? Then bad is 1 8 lLlxd5? 'ifd6 with 'ii' b 6 - Dvoretsky.]
a decisive pin on the d-file, but after 1 8 21 1:[c1 lLle6
lLlxe4! dxe4 1 9 'ifc3 f6 20 :g3 Wh ite would
22 d3
have retained an acceptable positio n . I
already wa nted more . Even so, White should have tried 22 f5, after
which I was intend ing 22 . . . lLlg5 23 l:td3 (23
1 8 exd4 cxd4
:f4!?) 23 . . . a6! 24 lLlxd4 lLlg4! or 24 lLlc7
1 9 lLlb5 1:[e4. The move in the game weakens the e3-
1 9 lLle2 lLlc5 20 'ii' c4 'ife4 21 1:[e1 d3 22 square, for where the black knight immedi
'ifxc5 'ifxe2! is bad for Wh ite. [In fact, after ately aims.
23 1i.xf6 'ifxe 1 + 24 1lf1 or 23. . . gxf6 24 :!g3+ lLlg4
22 . . .
<l;hB 25 1:[e3 the outcome remains unclear;
23 f51
however, 20. . . b5! 2 1 lLlxd4 'ifxf3! wins -
Dvoretsky.] What would you play now?
19 . . . lLlc5
(see diagram)
20 'ifc4
23 . . . lLlf8 1
(see diagram)
A sober reply. After 23 . . . lLle5? Wh ite had
My pieces are excellently placed , but the prepared a queen sacrifice: 24 fxe6! lLlxc4
d4-pawn is under attack. In this sharp 25 exf7+ <l;h8 26 fxe8'if+ and wins.
position Black had foreseen a purely 23 . . . lLlg5?! was unconvi ncing: 24 1:[g3 lLle3
positional solution . 25 'ifc5 (25 'ifb4 ! ? ) 25 . . . 'ifxc5 26 l:txc5 lLl d 1
20 . . . 'ifb61 27 h4 ! . A s Dvoretsky pointed out, the
I n Jazz Style ltJ 71

combination 25 . . . lt'lf4 26 ':'xf4 'if h 6 27 ':'xg4 The opponent has just two pawns for the
'ife3+ 28 f1 'ife2+ 29 g 1 is sufficient lost piece. However, for the moment there
only for a draw. are still d ifficulties in converting the advan
But why lau nch into u n necessary compl ica tage. All my pawns are broken and the
tions, when the opponent's position is knight is out of play.
already fai rly compromised? The th reats of 30 . . . :d61
24 . . . lt'le5 and 24 . . . lt'le3 are very dangerous, 3 1 l:r.xb5 b6
and Black only needs to ascertain that the Black has g iven up a third pawn , but now he
captu re of the d4-pawn does not relieve will be able to defend his b-pawn with the
Wh ite of his serious d ifficulties. knight from d7.
24 it.xd4 'ii' h 6! 32 ':'e4 lled8
The h2-pawn is attacked ; in addition , White's 33 ':'ee5 g7
back ran k is weak, and his rook at c 1 is
If 33 . . . lt'ld7 there is the reply 34 ':'ed5. There
hang i n g . 25 ':'h3 is met not by 25 . . . 'ifd2?!
is no need to hurry with this move - for the
26 'iWc3 , but by 25 ... l:txd4 ! ! 26 lt'lxd4 'ii'd 2 27
moment it is better to bring the king towards
.l:!.f1 1:Ie 1 28 'ii'c2 (28 l:.f3? l:.xf1 + 29 ':xf1
the centre . I n the endgame any respite
'i'e3+) 28 . . .I:txf1 + 29 xf1 'ii'f4+ .
should be used to strengthen the position to
2 5 h3 a6! the maximum.
The wh ite pieces are overloaded . If 25 34 f2?1 f6?1
hxg4 , then 25 . . . axb5 26 'ii' c3 ':'xd4 , and 25
[A move earlier the capture of the d3-pawn
it.b2 It'le5 (25 . . . lt'le3 ! ? ) 26 it.xe5 axb5 27
did not have any point, since the opponent
'iff4 ':'xe5! also does not help.
would have replied 34 ':e 7, with a simulta
26 it.xg7! 'ii'x g7 neous attack on fl and b6. But now, when
27 :g3 axb5 the fl-pawn is defended by the king,
[27. . . h5! 28 lt'lc7 ':e3 29 ':xe3 lt'lxe3 or 28 34 . . Jlxd3! could have been played with
hxg4 axb5 29 'ii' f4 h4 was even stronger impunity (35 :t1xb6? It'ld7) - Dvoretsky.]
Dvoretsky.] 35 ':'e3 It'ld7
28 'ii'x g4 'ii'x g4 36 g4 .:te8
29 lbg4+ h8 37 a4 l:te5
30 ne5 38 :b4 l1d5
72 In Jazz Style

39 e2 tLle5?! vented the invasion of the enemy rook,


U p till now I had acted logically, but here , activated all his pieces, and d riven the
unfortu nately, I relaxed and began t o play opponent's king to h3. He is now threaten
carelessly. 39 . . . tLlc5 40 d4 l::tx d4 would have ing 60 . . . : a 1 .
led to an easy win . 60 a5 bxa5
40 d4 tLlc6?! 61 bxa5 11a1
40 . . . ltxd4 was stronger. 62 g3 :txa5
41 ':c4 tLlxd4+ 63 g5
42 f2 tLlc6
43 g3 tLle5
44 l:tce4 l::t c 6
45 h4 tLld7
46 h5 g7
One senses that Black has ' u nwou nd' a little
and made the win more d ifficult. In such
cases it is important to cal m down , not hurry
to force events, and try again to d iscover the
correct course.
47 h4 l:lc1
48 l:.b4 h6
49 g3 l:tcd1
I real ised that I should exchange a pair of
63 . . . h5!
rooks .
When trying to convert a material advan
50 f3 l:t5d4?!
tage, pawn exchanges should be avoided .
Again an i naccu racy! The correct 50 . . . l:t 1 d4!
The fewer the number of pawns on the
would have forced an exchange i n a more
board , the g reater the chances of a draw!
favou rable version for me - on the 4th rank.
64 f4 f8
51 l:[b5 %:td5
Again Black prolongs the play. I n view of the
52 J:[xd5 l:txd5
rule just formulated , I did not want to play
53 b4 l:td4! 64 . . .f6 , but in fact this would have led to a
54 l:te4 l:td1 q u ick win : 65 gxf6+ xf6 66 l:td6+ g7 67
Of course, one pair of rooks must be l:le6 tLlf7 68 :g6+ f8 69 f6 tLld8 ! .
retained . I n the event of the exchange on e4 6 5 l:.d6 g7
the two pawns would have been no weaker 66 l:tb6
than the knight.
In the event of 66 lIh6 I was i ntend ing to
55 h4 f6 shut in the rook by 66 . . . tLlg6+ ! 67 fxg6 fxg6,
56 f4 l:tf1 + after which th ings end in zugzwang: 68 f3
57 g3 tLle5 l:ta4 69 g3 l::t b4 70 h 3 l::t g 4.
58 l:td4 1:[g1 + 66 . . . tLlg4
59 h3 e7 67 l:tb4 l:ta3
Black has achieved much : he has pre- 68 l:tc4?! l:tb3?!
In Jazz Style ctJ 73

We both missed the possibil ity of 68 . . . :f3+! 11 . . . b5


69 xf3 (69 e4 l:lxf5 ! ) 69 . . . lLle5+ . 1 2 .id3 ..t b7
69 l:la4 lLlf2 ! 1 3 b1 ..tc6?!
Th reatening 7 0 . . . 11b4+! . Black prepares . . . b5-b4, preventing in ad
70 l:la7 l:l b4+ vance the knight from going to a4 . The
71 f3 lLlg4 immediate 1 3 . . . b4 would have been met by
72 1:[e7 1 4 lLla4 (after 14 . . . lLlxe4? 1 5 'ii'x b4 not only
the bishop is en prise, but also the g7-
72 g6 lLle5+ .
pawn ). And yet the move in the game has a
72 . . . l:lb1 ! serious drawback, which is emphasised by
Th reatening 73 . . . l:lf1 + . my reply. It would have been better simply to
73 l:1a7 l:lb3+ castle .
74 f4 l:l b4+ [Castling is indeed the main theoretical
75 g3 lLle5 continuation in this variation. But 13 . . . b4 is
76 l:la5 1:[g4+ also possible - after 14 lLla4 i.c6! the
capture of the pawn leads only to a draw: 1 5
77 f2 l:lf4+
'ii'xb4 l:lb8 1 6 'ii' c4 ..tb5 1 7 'ii' b 3 i.c6, and
78 e3 lLlc4+ White does best to repeat moves (Shmuter
Wh ite resig ned . Vydeslaver, Beer Sheva 1 996 - Dvoretsky.]
1 4 'ii'e 1 1 ?
Dolmatov - Lerner
Now i f 1 4 . . . b 4 Wh ite h a s 1 5 i.xf6 ..txf6 1 6
Tashkent 1 983 lLld5. I n addition there is the impending
Sicilian Defence th reat of e4-e5, exploiting the opposition of
1 e4 c5 the white rook and the black queen on the d
2 lLlf3 d6 file.
3 d4 cxd4 [As was pointed out by grandmaster Stefan
4 lLlxd4 lLlf6 Kindermann, after 1 4 . . . b4 1 5 i.xf6 ..txf6 1 6
5 lLlc3 lLlc6 lLld5 a 5 Black would have retained a
6 ..tg5 e6 defensible position. Therefore instead of 14
We 1!? he recommends 14 'iWe3! - then this
7 'ii'd 2 a6
defence does not work in view of 1 7 lLlxf6+
8 0-0-0 h6 Wxf6 1 8 Wb6 - Dvoretsky] .
9 ..te3 14 . . . 0-0
Now I usually choose the more cautious 1 5 i.xf6 ..txf6
9 i.f4.
1 6 e5 ..te7
9. . . i.e7
In the event of 1 6 . . . dxe5?! Black does not
1 0 f4 lLlxd4 gain sufficient compensation for the queen .
1 1 i.xd4 1 6 . . . ..th4!? deserved serious consideration.
Nowadays this variation is very popular, but There could have followed 1 7 'ii'e 3 b4 1 8
at the time it was only just coming i nto use. lLle4 dxe5 1 9 fxe5 'ii'a 5 20 lLld6 ..te7 2 1
Therefore the following part of the game is ..te4, and White's position is still preferable.
improvisation at the board , by both sides. I
should mention that today Wh ite more often
places his pawn not at f4 , but at f3 . (see diagram)
74 In Jazz Style

1 8 tDe2 !
An un pleasant surprise. Black faces the
terri ble th reat of f5-f6 ! , for example:
18 . . . 'ii'a 5? 19 f6! gxf6 20 'ii'g 3+ h8 2 1 "f4
(2 1 exd6), or 1 8 . . . i.d5? 1 9 f6! gxf6 20 'ii'g 3+
h8 21 "f4 g7 22 tDg3. If 1 8 . . . dxe5, then
1 9 fxe6 is still strong.
18 . . . exf5
1 9 tDd4
Exploiting the position of the bishop at c6!
With gain of tempo the knight approaches
the important f5-point. After 1 9 . . . i.e4 20
i.xe4 fxe4 21 tDf5 the pin on the d-file is
decisive - 21 . . . d5? is not possible because
1 7 f5!
of 22 l:lxd5. If 1 9 . . . i.d7, then 20 i.xf5 is
A standard way of conducting the attack in a strong .
situation where Black has not managed to 19 . . . 'ilVc7
exchange pawns on e5. Of course, the
20 tDxf5 dxe5
move made by me demanded accurate
21 'ilVg3 g6
calculation .
21 . . . i.g5 22 h4 i.f6 23 tDxh6+ h8 was a
How would the offensive have been contin
tougher defence.
ued in the event of the captu re of the e5-
pawn? I did not even consider the variation 22 tDxh6+ h8
1 7 . . . dxe5 1 8 f6 i.xf6 1 9 i.h7+ - Black's
rook, bishop and two pawns are stronger
than the queen . After the correct 1 8 fxe6! it
is now unfavourable to give up the queen
( 1 8 . . . fxe6 1 9 i.h7+), but otherwise Black
encou nters serious difficulties.
17 . . . b4
This is what Lerner was counting o n . What
should Wh ite do now? 1 8 tDe4 dxe5 (or
1 8 . . . exf5) is unfavourable for him. I have to
admit that I am proud of my next move.
When you are engaged in a sharp struggle it
is important to be very attentive and
resou rceful , and to exploit all your re
sources. You only need to play insufficiently The black king is vulnerable and I have
energetically at some point, for the attack to excellent attacki ng prospects. How should
come to a standstil l and the in itiative to pass the offensive be conti nued? Don't think that
to the opponent. I ndeed , Black has the two you defin itely have to find someth ing bril
bishops, and he only needs to parry the liant. Sometimes d ifficult and by no means
immed iate th reats without particular dam obvious solutions have to be fou n d , but
age . . . more ofte n , without being d iverted , one after
In Jazz Style ltJ 75

another you have to make log ica l , accu rate transpose into some safe position with an
moves. extra pawn . It is dangerous to sit between
23 l:lhf1 two stools - the dual feeling played a n
The inactive rook joins the offensive. The adverse role, l e d t o excessive expend iture
position is not yet ripe for combinations of time, and prevented me at the decisive
such as 23 ttJxf7+. moment (now in ti me-trouble) of accurately
23 . . . .itd5 choosing and calculating a way to the goa l .
T h e first possibil ity was 27 ':xf5 . After
The f7-pawn has to be defended . But how
27 . . . :g8 the king wants to ru n away to f8 ,
should Wh ite conti nue now?
and to conti nue his attack Wh ite must
A good idea has been suggested - ttJf5 ! . sacrifice a rook: 28 l:.h5+.
But i f i t is carried out, i t should b e with gain
The second way was 27 .itxf5. The only
of tempo!
reply - 27 . . . 'iVc4 - leads after 28 'ifh3+ 'i'h4
24 'ii' h 3 g7 29 'i'xh4+ .itxh4 30 l:[xd5 to an endgame
25 ttJf5+! gxf5 with a n extra pawn for Wh ite.
If 25 . . . g8 Wh ite decides matters with 26 And , final ly, it is possible to interpose the
'i'h6 .itf6 27 ttJe3 (simu ltaneously attacking check 27 'ili'h3+ g7, and only then play 28
d5 and f6) 27 . . . .itg7 28 ttJxd5. xf5 . The reply 28 . . . 'iVc4 is now pointless ;
26 'ii'g 3+ h6 apart from 29 'i'h7+ the simple 29 l:txd5 is
After 26 . . . h8 27 lIxf5, mate is unavoid a lso threatened . There is only one defence:
able. 28 . . . .itxa2+! 29 xa2 'i'c4+ and 30 .. :ifh4.
I clearly saw all these ideas, but I did not
manage to make the correct choice . Prob
ably the simplest solution (and the one most
in keeping with my style at that time) was to
transpose into an endgame by 27 .itxf5 , in
which I would only h ave had some tech nical
d ifficulties to overcome.
But I begrudged giving up the attack
immediately. At the same time I was unable
to calculate fully the conseq uences of the
rook sacrifice. This was a pity - it was a
d i rect and pretty way to win .
2 7 l:lxf5! l::t g 8 2 8 l::t h5+ ! ! xh5 2 9 'i'h3+
g5 (29 . . . .ith4 30 'iVf5+ is no better) 30
'iVf5+ h6 31 'ii' h 7+ g5 32 1:[f1 ! .ite6
Here Wh ite has th ree conti nuations, two of (32 . . . 'ii'c8 33 g3! e4 34 .itxe4 .itxe4 35 h4+
which a re win n i n g . U nfortunately, I chose g4 36 'ii'x e4+ h5 37 lIxf7) 33 h4+ g4
the th ird and sq uandered all my advantage. 34 'ii' e 4+ h5 (34 . . . g3 35 "f3+ h2 36
I can explain why this happened . On the one g4) 35 g4+! xh4 36 'ii' h 1 + with a quick
hand I was rather excited and was eager to mate .
finish off nicely a game which had gone so Thus I could not bring myself to sacrifice the
well for me. But on the other hand, I still did rook, but I did not want to excha nge the
not feel sufficiently confident, and I was queens. This is why I settled for the third
looking for a conven ient opportun ity to possibil ity.
76 In Jazz Style

27 '6'h3+ Iitt g 7
28 ..ixfS? ..ixa2+ !
29 litt x a2
29 1itt a 1 ? :hB.
29 . . . .c4+
30 Iitt b 1 .h4
3 1 '6'e3
I thought that it would be hard for Black to
defend, seeing as his king is exposed , and
in the midd legame the presence of oppo
site-colour bishops should strengthen the
attack. But this assessment is incorrect - I
missed the fact that Black, by placing a rook
on the d-file, would prevent me from using 38 . . . :c8?
my rooks in the offensive. Also, the position After 3B . . . :eB! Black would have maintained
of the wh ite king is by no means secure, the balance - the threat of exchanging
especially after the capture of the e5-pawn . queens (39 . . . e 1 + 40 Iitt b 2 'ii'e 5+) would
31 . . . Itad81 have restricted Wh ite and not allowed him
When there are opposite-colour bishops time to develop an attack.
you should not cling on to material: the My opponent's last few moves in ti me
initiative is more important. Black happily trouble were poor and they again led to a
sacrifices his e5-pawn - it is merely h i nder lost position for h i m .
ing h i m .
39 .f4 f6? !
32 .xeS+
4 0 l:te3
What else?
The correct tactics! In the opponent's time
32 . . . ..if6 trouble you should avoid forcing variations,
33 .c7 lbd 1 + and go in for them only if they are winning.
34 nxd 1 'ii'f2 ! Of course, Lerner was expecting the check
Active defence! Black not only attacks the on g3 and he would have made his
bishop, but also threatens to play 35 . . . b3! . answeri ng king move instantly. But how
This explains m y next move . should he respond now? Here it is very easy
to become flustered and make some blun
35 ..ie6 ..ixb2 !
der, for example: 40 . . . '6'd5? 4 1 '6'g4+ and
I foresaw this counter-stroke by the oppo
42 'ii'x cB. The only acceptable move was
nent, and I thought (rig htly, i n all probability)
40 . . . 'ii' d 7.
that I should al low it.
40 . . . l:tc4?
36 Iitt x b2 '6'f6+
41 '6'g3+ '6'g4
37 Iitt b 1 '6'xe6
42 '6'd6
38 l:td3
It's all over! When your flag is about to fal l ,
all that you look for are checks and
captu res, and, of cou rse, Lerner simply d id
not have time to assess the consequences
(see diagram) of my quiet move.
In J azz Style

Here the game was adjourned . It did not last 47 ':'xc3 bxc3
long on the resumptio n . 48 a2 f4
42 . . . lIc3 49 b3 e3
43 'ii'e 7+ g6 50 xc3 f5
44 'ii'e 8+ f5 51 h4!
45 'ii'e 6+ g5 Black resigned .
46 'ii'x g4+ xg4
78

PART I I I

P ractica l Exped i e n cy i n the ta ki n g


of Decis i o ns

B e n iam i n B l u menfeld

P ractical C ha n ces in a C h ess Game

Pthe game, abil ity to calculate deeply,


ure chess qualities (understanding of Kmoch - N i mzowitsch
Bad N iendorf 1 927
etc.) are not sufficient for success. One also
needs to possess those qual ities which are
the guarantee of success in life and in other
competitions: practical sharpness, the
habit of quickly finding your bearings in
a new situation, the ability to take a firm,
quickly realisable decision in a compli
cated position which does not lend itself
to calcula tion, complete purposefulness
towards your goal - to win or save
yourself from defeat, composure and
self-possession in a difficult position,
and the avoidance of dizziness from
success in a favourable position.
I n chess literatu re , it is usual for games to
be explained mainly in terms of the theoreti There followed 44 ... b4 45 .i.a4 (bad is 45
cal correctness of the contestants' play. cxb4 lLlxd4 46 bxa5+ lLlb5, when Black
However, such an explanation does not remains with two strong passed pawns).
always g ive a true reflection of the cou rse of Now 45 . . .lLlxe5 46 .i.xd7 suggests itself.
the battle. An experienced player often Wea ker is 46 dxe5 .i.xa4 47 cxb4 .i.b3 48
chooses a particular continuation, not bxa5 l:txa5, when Black reta ins an extra
because he is sure that it is the best of all pawn and winning chances, despite the
the possibilities, but exclusively on the opposite-colour bishops.
basis that it gives the best practical
[After 45. . . lLlxe5 46 i.. x d7 lLlxd7 4 7 cxb4 a4
chances.
Black wins without difficulty. But 46 cxb4!
i.. xa4 4 7 bxa5+ i.. b 3 48 i.. x e5 l:xa5 49 cJi;f4
Practical Chances in a Chess Game ltJ 79

is possible, retaining real chances of saving the preced ing moves Black did not advance
the game - Dvoretsky.] his a5-pawn , which so suggested itself.
Instead of this, N imzowitsch (after 45 a4) 51 cxb4 a4 52 b5+ Wh ite gives up a pawn to
played 45 ... b3, when there followed 46 open a path for his bishop; however, the
.i.xc6+ xc6, and the position appears to rook and bishop prove to be helpless.
be a dead draw: Black's passed a- and b 52 ... xb5 53 a3 c3 54 ::tb1 c4 55 f4
pawns are easily stopped , and on the xd4 56 f2 c4 57 e1 d4 58 e2 d5
kings ide it is impossible to break throug h . 59 f3 b7 60 l:te1 c4+ 61 f2 b2 62 f5
The game continued : 4 7 g5 :a7 4 8 l:. b 2 . I n exf5 63 e6 c6 Wh ite resigned .
blocki ng the black pawns, i t would be The combi nation carried out i n the game
dangerous to stick to purely waiting tactics. shows just how many dangers were lying in
For example: 48 f3 l:[b7 49 g3 a4 50 wa it for Wh ite in this seemingly harmless
.i.a3 b2! 51 l:txb2 l:tb3! 52 l:[xb3 cxb3 53 position . Therefore N i mzowitsch correctly
'iit f3 b5 54 e3 b2 55 xb2 c4 56 d2 decided that the conti nuation chosen by h i m
'iit b 3, and Black wins the bishop. would give t h e best practical chances .
48 ... l:tb7 49 f4 [Nimzowitsch points out
that after 49 a3! he would hardly have Kmoch - Yates
been able to break through - Dvoretsky]
San Remo 1 930
49 ... c8. Apparently with the aim of trying
to penetrate with the rook on the h-file;
therefore it was natural for Wh ite to make
the following reply, which Black provoked
with the aim of diverting the wh ite king from
the queenside and carrying out his plan ned
combi natio n .
50 g3

By excel lent play Wh ite has gai ned a


decisive advantage. Now he should have
conti nued simply 32 .l:t6xe5 dxe5 33 1:[xe5
[33 0,e7!? is simpler - Dvoretsky] , and if
33 . . . .1:.g5, then 34 0,e7 ! . In this case Wh ite's
pawns on the queenside, although isolated
and doubled , carry out their fu nction per
fectly well - they restrain the black pawns
50 .. l:tb41 With the aim of obta i n i ng th ree
. on the same wing , whereas on the kingside
passed pawns , which will advance with gain Wh ite can create two connected passed
of tempo in view of the position of the wh ite pawns. The win is achieved automatically.
rook on b2. Now it becomes clear why on I n stead of this simple conti nuation, which
80 Practical Chances in a Chess Game

does not allow Black any chances, White Wh ite could have immediately decided the
embarked on a combination. There followed : game with a simple combination : 41 CfJg6+
32 CfJe7 xe6 33 CfJxg8 xc4! (White was hxg6 (4 1 . . . lixg6 loses a piece) 42 'iWh4+
obviously hoping for 33 . . . xg8 34 f4). etc. I n stead of this, probably without any
As a result of the combination White thoug ht, Wh ite played 41 CfJh5, which is
remai ned the exchange up, but Black seemi ngly also very strong.
obtai ned defi nite cou nter-chances, since on The game conti nued 41 . . .lie5! (4 1 . . . lixg3
the q ueenside he had acquired mobile 42 i.. x d4+ e5 43 l:[xe5 etc. was bad for
pawns, supported by his two bishops. In the Black) 42 'it> h 1 [the position would still have
end Black even won . been won after 42 :e 1 ! - Dvoretsky]
42 . . . xc3 ! 43 l:[xe5 i.. x e5. Now Wh ite has
From this it can be concluded that, if you
a queen for rook and m i nor piece , but his
have a sufficient advantage, you should
attack on the kingside has evaporated ,
choose continuations where the win is
whereas Black can develop active play. I n
achieved without counterplay for the
the e n d Wh ite even lost.
opponent.
In connection with this example the fol low
The following conclusion, which is not so ing general comment can be made. When
absu rd , can also be drawn : if there is a an attack concludes with a gain of
choice between two continuations - one,
material, it is as yet too early to celebrate
giving a decisive positional advantage,
victory. Often in such cases the entire
with an equal balance of forces, and
situation changes, and pieces, which
another, giving roughly the same advan
earlier were systema tically placed for the
tage but with unequal material (as in the
conducting of the attack, after the attain
given example: rook and knight against
ing of the goal may now be misplaced, as
two bishops), it is better to choose the
play has switched to another part of the
first continuation. With an equal balance
board, where the opponent has more
of forces the methods of attack and
forces or they are better placed. There
defence have been better studied, and so
fore you should be especially careful at
here there may be fewer surprises.
critical moments, when win of material is
possible, and carefully weigh up whether
Yates - Ahues it is worth gaining a ma terial advantage if
Scarborough 1 930 this worsens your position.

I n the given example Wh ite did not notice an


im mediately winning conti nuatio n . But often
a player deliberately avoids a simple and
decisive conti nuation, since he wants to win
'brilliantly' .
There was a highly vexing occu rrence in the
following game.
Practical C h a n ces in a Chess Game ctJ 81

Sergeev - G r ig or i ev surely to the goal - to a win. It is


Master Tournament, Moscow 1 932 expediency and the choice of the most
economic methods for achieving the
goal tha t constitute the inner beauty of
chess. Doing things for effect - the result
of a false understanding of beauty in
chess - often produces poor results.

In a winning position you should aim for


simple, clear decisions, whereas in a lost
or significantly inferior position, by con
trast, you should aim to complicate the
play. In a position where you are bound
to lose after natural continuations, you
should not be afraid to make sacrifices
of material, it being important to obtain
active counter-chances.
I n particular, it should be mentioned that
In this position Black carried out an interest one of the characteristic featu res of Alexan
ing combination : der Alekh ine's play is that in inferior posi
tions he does not al low his opponent to
31 . . .tiJ g 5 3 2 tiJxb7 f3 33 tiJxf3
increase his advantage, but strives to
The only way to avoid the loss of a piece or d isru pt the natu ral development of events
mate . [ This is not so: by continuing 33 tiJxdB and take the play along d ifferent l i nes, by
fxg2 34 xg2 :r:J.xdB 35 'ila3 White would going in for complications or sacrifices . This
have retained a perfectly good game -
characteristic featu re of Alekhine's play is
Dvoretsky.] especially memorable to me from the
33 . . . tiJxf3+ 34 ..txf3 ..tc61 35 ..txc6 l:td3 36 nu merous (mainly friendly) games played
'ili'b2 with h i m , when he had not yet ach ieved the
peak of his chess fame.
Now it only remai ned for Black to obtain a
clearly winning position with the natural I n N imzowitsch's play it can also be noticed
move 36 . . . tiJxc6 , as Grigoriev would un that he does not lose heart in lost positions
doubtedly have played even in a lightning and he often saves hopeless games, by
game. But here, to the general aston ishment fi nding some practical counter-chances.
of the spectators, he stopped to think, and The following example is typica l .
after some thought he unexpectedly played
36 .. :iVxg4+ . There followed 37 ..tg2 i1ff3
(Black was planning a 'pretty' mate by
playing his knight to f4 or h4) 38 tiJc5 tiJf5
39 .l:!.e4 ( Black had overlooked this refuta (see diagram)
tion ) 39 .. :iIi'g5 40 tiJxd3, and Wh ite won .
As a result of playing for bri l l iance , Black
th rew away a deserved win. This example
should serve as a lesson to us all. The best
continuation is the one which leads most
82 <t> Practical Cha nces in a Chess Game

N imzowitsch - Euwe opponent, having gained a decisive ad


Carlsbad 1 929 vantage after a long and tiring struggle,
will to a significant degree have already
exhausted his strength and will be in a
demobilised state, assuming that all the
difficulties have been overcome. There
fore it is wrong to talk about fortune and
chance, when a lost game is saved.
Fortune favours the strong!

By no means all top-class players i nvariably


possess presence of mind, as it apparent
from this example:

Em. Lasker - Janowski


7th match game 1 909
Wh ite is doomed . His game is u ndeveloped .
Black has a strong pawn on e3, which is
th reatening to q ueen . Wh ite cannot take the
pawn , since if 23 l:txe3 there follows
23 . . .xd4 , winning a rook. 23 :txf7+ !txf7
24 . xc8 e2 is also bad . However, in this
horrible position N i mzowitsch d id not lose
his composure , but played 23 liJc3 , placing
his knight en prise.
If Euwe had reacted with care to the
opponent's desperate try, he wou ld, of
course, have found the winning continua
tio n : 23 . . . 'iWxd4 24 l:taf1 'ii'e 5+ 25 h1 l:.c7
etc. But, not suspecting any danger, he
made the obvious move 23 ... .l:1xc3? and
after 24 l:.af1 e2 25 l:.xf7+ l:txf7 26 'ii'x f7+ Janowski resigned in this position , about
h 6 27 .f8+ he was forced to resign. which Tarrasch justifiably commented : 'Ja
Some would say that N imzowitsch did not nowski , emotionally depressed , laid down
deserve to wi n . I cannot agree with this. his arms too early, instead of making use of
Even the strongest master ca nnot take his last and by no means bad chance. By
everything into account and so sometimes contin u i ng 63 . . . c5 64 liJd5? (an obvious
he ends up in an inferior or even lost move with a seemingly unstoppable mate
position , not only against an equal opponent threat) 64 . . . l:.xf3+ 65 xf3 xe4+ 66 xe4
but also against a weaker one. It is he could have achieved a pretty stalemate.
obviously good not to lose heart in a Of cou rse, the opponent could have avoided
difficult position and to find counter this, by playing 64 l:.b7, but it is possible that
chances. In this case a mistake by the in the heat of the battle he might not have
opponent is extremely probable: the noticed this stalemate . '
Practical Chances in a Chess Game ltJ 83

On the basis of my o wn experience, I opportu n ity to make it harder for Wh ite to


have seen many times that, even in the wi n .
seemingly most hopeless position, prac 3 9 . . . .l:!.d2+. No hope is offered b y 3 9 . . . c 3 40
tical chances have been found. The c8'iV c2+ 41 'iVxc2 .l:!.xc2 42 xc2 , when
following example is not without interest. Wh ite has a tech nically easy win.
40 c1 ? After this natu ral reply it is
Orloy - B l u m e nfe ld
doubtful whether Wh ite can win . He should
Moscow Championship Semi-Final 1 932 not have feared the d iscovered check and
played his king to e 1 , for example: 40 e1
lixh2+ 4 1 f1 c3 42 c8'iV c2 43 'iVh8+ h6
44 J:r.c8 c1 'iV+ 45 .l:!.xc1 .:th 1 + 46 g2 11xc1
47 'iVf6 , and the win is assured . However, in
order to decide on this continuation , it was
necessary to see the complications arising
after the natu ral reply.
40 . . . l:.e21

Black's position is hopeless, si nce the


strong c6-pawn is bound to decide the
game. Wh ite made the spectacular move 36
lLlc4 .
If Black takes the knight, he blocks the c-file
for his rook and the c6-pawn promotes
unhindered . Of course, Black could reply
36 . . . Uc3, but then White continues 37 ttJxa5
and the win becomes a matter of simple
technique. [In fact after 3 7. . . i.. b 4 38 ttJb7 it Perpetual check on the squares e1 and e2
is not easy to win, as the white knight is is th reatened . Wh ite is now forced to
badly placed - Dvoretsky.] 36 . . . .l:!.a4 37 ttJe5 sacrifice the exchange.
or 37 ttJ b6 is also hopeless. 41 l:txb4 axb4 42 c8'ii' . Wh ite has to give up
I n both of these variations it is hard for White the pawn , since if 42 e4 there follows
to go wrong ; everything is simple and clea r. 42 . . . .l:!.e 1 + , and Black gains a draw either by
Therefore I decided to allow Wh ite to obtain perpetual check, or by the pawn advance
a queen , merely to obta i n some practical . . . b4-b3 .
counter-chances. There followed: 36 . . . l:!a2+ 4 2 . . . Uxe3 4 3 f4. Now Black forces a draw.
37 d1 i.. b4 38 c7 dxc4 39 .l:!.b8 . White should have decided on 43 'iic5 Uxf3
It would now seem to be time for Black to 44 'ii'x b4 h5 followed by . . . :f5 ; in this case
resig n , since on his next move Wh ite will he would have had winning chances , al
obtain a q ueen . However, Black has a latent though d ifficult to convert.
84 P racti cal C h ances in a Chess Game

43 . . Jlc3+ 44 b2 lIb3+ 45 c2 .:tc3+ 46 I t i s well known that the endgame with rook
d2 .:td3+ 47 e2 .:tc31 (47 . . . nxd4 48 'Wc5 and knight against rook is d rawn . The
was bad ) 48 d2 nd3+ 49 c2 ':'c3+ 50 presence of the black pawn should not
d2 .:td3+ Draw. make any d ifference . Fahrn i is an experi
I ndeed , Wh ite can not achieve anything, for enced enough master to avoid doing any
example: 5 1 e2 l:tc3 52 'Wc5 .:tc2+ 53 d 1 thing really stu pid . It would appear that it is
b3 54 d5 exd5 55 'Wxd5 ':xh2, and if 56 not worth wasting time by playing on.
'Wxc4 , then 56 . . . b2, while if 56 c1 there But Rotlewi decided to play o n , since he
follows 56 . . . :c2+ 57 b 1 c3! , and it is now saw a practical chance, provided precisely
Wh ite who has to seek a draw by perpetual by the fact that Black had a pawn .
check. Later the position in the next diagram was
reached .
Whereas cool-headed conversion of an
advantage in a winning position and
presence of mind in a lost position are
typical of most experienced players,
comparatively more often one observes
a weakening of attention and will to win
in obviously drawn positions. I n his book
on the 1 927 New York tournament, Alekh ine
criticised Rudolf Spiel mann for the fact that
in certai n games he agreed a draw, al
though he had practical chances, albeit
minimal, of wi nning.
As confirmation that a tenacious striving for
victory can have a favourable outcome even
in a drawn position , I will g ive the following
example. Black made the natu ral move 79 . . . a3, after
which there followed a study-like fi nish: 80
Rotlewi - Fahrni f7 W h 6 (if 80 . . Jh 1 , then 81 d5! is
Carlsbad 1 9 1 1 decisive) 81 g8! Black resigned . It is
curious that, had it not been for Black's
pawn , he could have saved hi mself by
playing for stalemate with . . J Ig 1 .

From all that has been said it would be


incorrect to conclude that when playing you
should hope for blunders by your opponent.
The reader who has carefully thought about
the given examples will see that practical
chances can be created only as a result of a
correct evaluation of the position and an
estimation of its cha racteristic features.
CtJ 85

Vlad i m ir Vu lfson

Does it pay to s h a rpe n the P l ay?

W Of course,
e are often faced by such a question . on e6, Tal effectively condemned it to being
a ready-made solution exchanged for a white knight.
does not exist - everything depends on the 10 c3 a5
specific ci rcu mstances. We can learn to In such positions one cannot al low b2-b4 ,
understand this problem better if we see which secu res Wh ite the in itiative on the
how it was solved by other players, and q ueenside.
each time make a critical assessment of
11 ttJc4 ttJd7
their actions. It is interesting to follow how a
player's choice is influenced by his charac 1 2 ttJg5
ter and style of play, when as a conse
quence of individual preferences he is
unable to decide on the objectively best
course.
We will beg i n with an analysis of two games
by M i khail Tal . He played the first when he
was at the height of his powers, and the
second many years later (I hope you will
sense the d ifference). In the analysis of the
games we will do some training by seeking
replies to the d ifficult q uestions which
invariably arise on the way.

Vasyu koy - Tal


29th USSR Championship, Baku 1 96 1
Black is forced to part with his bishop. After
King's Indian Attack
the exchange he can recaptu re on e6 with a
1 e4 e6
piece or the f-pawn . A third possibility is to
2 d3 d5 give up the bishop for the knight on c4 .
3 ttJd2 ttJf6 Which would you prefer?
4 ttJgf3 ttJc6 Of the two knights, the one on c4 is the more
5 g3 dxe4 dangerous - it can subsequently be switched
6 dxe4 .ic5 via e3 to d5 or f5 . True, after the doubling of
the pawns on e6 these poi nts will be
7 .ig2 e5
defended , and the f-file opened . Even so,
8 0-0 0-0 after 1 2 . . . "e7 1 3 ttJxe6 fxe6 Black's posi
9 'iVe2 .ie6 tion looks dubious. Apart from playing for
I n this set-up Black usually plays . . . a7-a5, simplification with 1 4 .ie3, with the future
. . . b7-b6 and . . . .ia6. By placing his bishop hope of exploiting the weak pawns, 1 4 'iti> h 1
86 <;it Does it pay to sharpen the Play?

followed by f2-f4 also deserves serious


consideratio n . After the opening of the
position the power of the two wh ite bishops
will tel l .
12 . . . i.xc41
1 3 'ii'x c4 'iVe7
Now Evgeny Vasyukov could have retu rned
his knight to f3 and then played it to h4,
provoking the reply . . . g7-g6, after which his
bishop would have gained an excellent post
at h6. However, he begrudged moving his
knight back without a special invitation.
1 4 d2
The plan is understandable: .l:tad 1 and then This is not so. When I began considering my
c1 . 1 6th move, I automatically looked at
14 . . . .l:tad8 16 . . . f1xd2, but it very soon became clear to
1 5 .l:tad 1 lLl b6 me that with this move Black does not
achieve anything, since after 1 7 'ikxd2 I1d8
1 6 'iVe2
1 8 'ifc1 1:!xd1 1 9 f1xd1 he remains the
Of course, not 1 6 'ifb5? .l:txd2 ! . exchange down.
16 . . . l:td6 But now let's use our imagination a little. It
Now 1 7 i.c1 ( 1 7 lLlf3 ! ? ) 1 7 . . . l1fd8 1 8 ':'xd6 turns out that, if the white bishop were not at
'ifxd6 1 9 .l:t e 1 followed by 20 i.f1 is not bad , g2, but at h3, a combination would be
g radually taking control of the squares on possible. I thought over my 1 6th move for 40
the q ueenside. With this pawn structu re the minutes. Initially I wanted to play 16 . . . 1:1.d7,
two bishops would have ensured Wh ite a provoking the reply 1 7 i.h3 (with gain of
small but lasting positional advantage. A tempo!) . But I decided that this would be too
good textbook example of how to handle obvious. White can simply continue 1 7 c 1 ,
such positions is the game Petrosian-Sax, retaining a minimal positional advantage.
played i n 1 979 at the i nternational tourna In the game there followed 1 6. . . l1d6, and
ment in Tal l i n n (cf. the addendum to the now Vasyukov, to my surprise, quickly
lectu re). played 1 7 i.h3. There immediately followed
17 i.h3 1 7 . . . :txd2!.
Think about what Tal might have played 1 8 'ii'x d2 1:!d8
here . 19 'ii'c 1 l:txd 1
(see diagram) 20 ':'xd 1 'iff6!
N ow Black's idea becomes clear. He is
17 . . . l:txd2 ! ? threatening both 21 . . . h6 (the knight has no
A n u nexpected combination. But not for Tal , retreat square), and also 21 . . . 'ifxf2+ 22
who anticipated i t before his previous move. 'iti> h 1 e3.
Here is his commentary: Tal 's clever combi nation is certainly tempt
A glance at the position suggests that play ing. But is it correct? After a l l , even after the
will continue on quiet positional lines, and captu re on f2 he will only have a pawn for
that here there is no place for combinations. the exchange.
Does it pay to sharpen the Play? ltJ 87

The g randmaster g ives the variation 2 1 In the game 22 . . . axb4 23 cxb4 i.e7 was
i.g2 'ii'xf2+ 2 2 h 1 iLe3 23 ltJh3 i.xc1 24 played . But why not place the bishop on d4?
ltJxf2 i.xb2 with the better endgame for It turns out that in the variation 23 . . . i.d4 24
Black. Another try, 2 1 i.d7 'ii'xf2+ 22 h 1 b5 ltJa5 25 .1i.g4! 'ii'xf2+ 26 h 1 i.e3 there
i.e3 23 i.xc6, is refuted by 23 . . . i.xg5! with is the defence 27 ltJh3! . Wh ite's position
the terrible threat of 24 .. .'iVf3+. relies on this tactical nuance .
However, it is possible to defend more In which version is it better to retreat the
strongly. Both the players and the commen bishop to e7, immed iately or after the pawn
tators overlooked the simple move 21 1:[d3 ! , exchange on b4? What is the d ifference? I n
taking control o f the important f3- and e3- each case, i f he wishes, Black obtains two
squares. After 27 . . . 'ii'xf2+ 28 h 1 ltJc4?! pieces for a rook, but it is important that the
(28 . . . h6? 29 1:[f3 e2 30 i.f1 , and the opponent should not be able to activate his
queen is trapped ) Wh ite has a pleasant forces. After 22 . . . iLe7!? 23 i.d7 for the
choice between 29 1:[f3 'ii'x b2 30 'ii'x b2 moment the c-file is closed and there is the
liJxb2 31 .l:!.xf7 ( 3 1 ltJxf7 ! is simpler) 31 . . . h6 excellent resou rce 23 . . . ltJb8 ! . However, the
32 i.e6 h8 (32 . . . hxg5 33 1:[f2+) 33 J::t xc7 conseq uences are far from clear: 24 i.g4
hxg5 34 l:txb7 and 29 'ii'f 1 'ii'x f1 + 30 i.xf1 'iWxg5 25 'ii'x g5 i.xg5 26 bxa5 ltJc4 27 i.c8!
liJxb2 (30 . . . ltJd6!?) 31 .u.d7. I n both cases it ltJxa5 28 i.xb7! (28 . . . ltJxb7 29 l:t b 1 ).
is doubtful whether Tal would have been [By playing 2B... c6! 29 iLcB (with the threat
able to save the game. As Dvoretsky of 30 :r:Lb 1) 29. . . iLe 7!, Black retains the
pointed out, Black's play can be improved better chances, since he prevents the
by 28 . . . iLe7! 29 ltJc3 ltJc4 , but in the ending invasion of the rook and securely blockades
arising in the variation 30 ltJd2 ! ? i.g5 3 1 the passed a-pawn. On the other, instead of
'iNf1 'iVxf1 + 3 2 ltJxf1 ltJxb2 3 3 J::t d 7 Wh ite's 24 iLg4 White can try 24 i.b5!? (24. . . c6 25
chances are better. bxa5 ltJ 6d7 26 a6!? bxa6 27 i.xa6)
Vasyukov also devised a reasonable idea, 24 . . . 'ii'xg5 25 "fixg5 iLxg5 26 bxa5 ltJcB 27
but even so it was much inferior to 21 lid 3 ! . f4! with a complicated and double-edged
21 iLf5?! 96 ending Dvoretsky.]
-

22 b4! 22 . . . axb4
The play has become much sharper. Tal 23 cxb4 i.e7
now has a choice of three or four possibili [I think that the simple 23. . . i.xb4 deserves
ties. Which of them is the strongest? serious consideration, with good compen
sation for the sacrificed exchange Dvo
-

retsky]
24 i.d7 ltJd4
Tal writes: 'Black does not want to simplify
the position and he avoids 24 . . . 'ii'xg5 25
i.xc6 'ii'xc 1 26 1:!xc 1 bxc6 27 xc6 iL d6,
continuing to devote his main attention to
the kingside. '
There is no point in going in for the variation
24 . . . ltJxd7 25 1:!xd7 'ii'x g5 26 'ii'x g5 iLxg5
27 :xc7 (stronger than 27 b5 ltJa5 or 27 a3
i.c1 ) 27 . . . ltJxb4 28 a4 b6 29 1:!b7, when
Wh ite obtains a dangerous passed a-pawn .
88 Does it pay to sharpen the Play?

25 'ii'x c7 d8 'ifxe5). Stronger, apparently, was the simple


26 'ifxb7 'it'xg5 30 g2 'ifxb4? (as pointed out by Dvoretsky,
30 . . . g7! 3 1 a5 'ifxb4 32 e8 'ifb3 leads to
Objectively, Tal's decision was risky - he
a draw) 3 1 'iWb8 'ife7 32 a5.
has allowed his opponent not one, but two
passed pawns on the queenside. However, 30 . . . lDxa4!
he has kept the queens o n . I n an endgame 31 g2
with rook and pawn against two minor 31 xa4 lDe2+.
pieces, a very important factor is the lD b6
31 . . .
presence of an outside passed pawn , but in
32 l:[c5 'iff6?
the midd legame there are chances of
creating an attack on the king - after all, A mistake in reply. Here is Tal 's explanation:
Black has one piece more . ( However, it is 'Here White unexpectedly offered a draw.
not at all easy to incl ude the knight at b6 and Somewhat confused, I forgot about the
bishop at d8 in the attack . ) I n add ition, Tal intended 32. . . g7, which would have given
always handled his strongest piece with an easy win, and instantly replied 32. . . "ikf6?. '
great skill - it is sufficient to remember his A conclusion about the importance of
famous game against Oscar Pan no from the ' resistance to interference' suggests itself.
1 958 I nterzonal Tournament in Portoroz. I ncidentally, look at the position after
27 e8 'ii'f6 32 . . . g7! . How well the black knights are
28 a4 f8 ! placed - together with the remaining pieces
they control all the i nvasion squares!
It is important to drive away the bishop, i n
order t o free the black queen for active play. 33 'iWb8t
29 b5 'ifd6 Wh ite gains saving cou nterplay.
Attacki ng b4 and th reatening 29 . . . lDf3+. 33 . . . 'ii'f3+?
[ 'Mistakes never come singly! It was not yet
too late to play 33 . . . g 7!, for example: 34
'ifxe5 (34 ':'xe5 e 7 with the threats of
35. . . d6 and 35. . . xb4) 34 . . . 'it'xe5 35
':'xe5 c 7 36 :c5 d6, retaining winning
chances - Dvoretsky.]
34 g1 'ifd 1 +
35 g2 'iff3+
36 g1 lDe6
37 ':'c6 'ifd 1 +
38 g2 'ifd4
39 ':'d6 'ifxe4+
40 g1 'ifb1 +
41 g2
30 lIc1 ? Draw.
I n the time scramble White blunders his Throug hout the game Tal constantly took
main pawn on a4. After 30 .l:.b1 lDxb5 31 risks. First he d isru pted the balance , by
axb5 'ifd3 32 l:tc1 Black would not have conced ing the advantage of the two bishops
stood worse (32 . . . 'ifxb5?! 33 'it'b8 'iWd3 34 for the sake of rapid development. Then he
Does it pay to sharpen the Play? ttJ 89

decided on a d ubious combination , and knight, conced ing to his opponent the
finally, in search of attacking chances he advantage of the two bishops. However,
allowed the opponent to obtain two con here th is does not play a particular role.
nected passed pawns . Such was his style of 11 h3
play at that time!
11 'ife2 followed by ltJd2-c4-e3 was prefer
able.
Ribli - Tal 11 . . . i.. xf3
Candidates Tou rnament, Montpell ier 1 985 1 2 'ifxf3 'iVe7
Reti Opening
1 3 nad 1 ? !
1 ltJf3 d5
Another routine move, after wh ich Wh ite is
2 g3 i.. g 4
already in some d ifficulties . He should have
3 g2 c6 placed his pawn on a4, preventing not only
4 b3 ltJd7 the exchange of the dark-square bishops by
5 i.. b2 ltJgf6 1 3 . . . i.. a 3, but also 1 3 . . . b5, which deprives
6 0-0 e6 his knight of its lawful c4-square.
7 d3 i.. c 5 13 . . . b5!
8 ltJ bd2 0-0 14 h4?!
g e4 dxe4 One mistake often leads to another. Appar
1 0 dxe4 e5 ently Zoltan Ribli remembered about his
light-sq uare bishop and decided to bring it
out to h3. But in so doing he weakens the
g4-square .
Wh ite's primary objective is t o rearrange h i s
badly placed knight on d2. There is only one
route available to it: via f1 to e3. This means
that the correct move was 14 :fe 1 ! .
14 . . . as
The g4-square can not be occupied immedi
ately: if 14 . . .'iVe6 there is the reply 1 5 'ii'f5! .
Therefore for the moment Tal harasses his
opponent on the queenside, by preparing
1 5 . . . a4.
1 5 c3
In order to answer 1 5 . . . a4 with 1 6 b4. But
The structu re of the position is roughly the
allowing the opening of the a-file would have
same as in the previous game. The only
been the lesser evil , since now another
difference is in the placing of Black's
important square is weakened - d3.
queen's knight (there it stood at c6 ,whereas
here this square is occupied by a pawn ) and 15 . . . ltJb6
White's dark-sq uare bishop. These changes 1 6 1:[fe1 ?
are rather to Black's advantage. 1 6 "'e2 or 1 6 i.. h3 was better.
It is probable that on this occasion too Tal
will have to exchange his g4-bishop for the (see diagram)
90 Does it pay to sharpen the Play?

Black's last few moves have entirely fo 21 'ii'x e6? fxe6 22 xf2 l:tfxf3+ 23 li:)xf3
cused the opponent's attention on the l:txd 1 is completely bad for Wh ite. After 2 1
q ueenside - he has forgotten about possi xf2? 'ii'd 6 h e ends u p in a mortal p i n on
ble d iversions on the opposite side of the the d-file. The best chance of a defence was
board and incautiously weakened his f2- offered by 21 .te2 ! :xd2 22 :'xd2 .te3
point. This is immediately exploited by Tal , (22 . . . .tc5 23 'iVxe6 fxe6 24 .tg4) 23 .l:.d3
who, i t would appear, always remembers .tc5 (23 . . . 'ii'xf5!? 24 exf5 .tc5 25 .tf3 %lc8)
about the enemy king. 24 'ii'x e6 fxe6 25 .tf3 (or 25 l:tf3). Black is a
16 . . . 'ii'e 61 pawn up, but the win is stil l a long way off. It
is amazing how g reat the safety marg i n is in
1 7 'ii'f5 li:)g4
chess - despite Wh ite's nu merous errors,
1 8 .:te2 .l:.ad8 his position can still be held !
White's position is already d ifficult. If 1 9 [ The position can no longer be held! In the
.th3 Tal was intending 1 9 . . J:td3! (with the event of 2 1 .te2 the pretty stroke 2 1 . . . li:)d5!!
threat of . . Jbg3+) 20 g2 'ii'xf5 21 exf5 is decisive: 22 xf2 (22 'ii'xf2 li:)e3+ 23 g 1
li:)xf2 22 ::'xf2 .txf2 23 xf2 l::t fd8 24 e2 li:)xd1; 22 'ii'xe6 li:)e3+) 22. . . li:)e3 23 .txd3
e4 . [In the event of 19 .th3? there is a li:)xd1 +24 e2 li:)xb2 - Dvoretsky.]
simpler win by 1 9 . . . li:)xf2! 20 l1xf2 g6! - 22 xf2 'ii'd 6
Dvoretsky.] 23 .tc1 g6
1 9 .ttl lld3! 23 . . . 'ii'c 5+ followed by 24 . . ....xc3 was also
20 g2 li:)xf2 1 strong.
It should be said that Tal liked sacrificing two 24 'ii' g 5 f6!
pieces for a rook. So that the previous Before the f-file is opened , the wh ite queen
game, i n which completely the opposite must be d riven away. 24 .. .f5 is unconvinc
balance of force a rose, is rather an excep ing: 25 g2 l:txf3 26 xf3 "'d3+ 27 f2
tio n . fxe4+ 28 g 1 (Tal) 28 . . . e3 29 'ii'g 4! (weaker
.txf2 is 29 'ii'x e5 li:)d7! 30 'ii'e 6+ g7) 29 . . . li:)d7!?
21 lbf2
(29 . . . e3 30 'ii'e 6+ with a d raw) 30 li:)e4! e2!
Can White somehow set up a defence? 31 :e1 l:tf1 + 32 g2 l:txe 1 33 'ii'e 6+ , and
the battle ends in perpetual check (Dvo
retsky).
Does it pay to sharpen the Play? ttJ 91

25 'ifh6 f5 prophylactic move 1 0 . . . h 6 ! , so that if the


26 'it;Jg2 knight moves he has the reply 11 . . . g S ,
If 26 'it;Je1 (e2), then 26 .. .f4 27 gxf4 ':xf4 is preventing t h e opening o f t h e position by
strong , while if 26 1:1e 1 - 26 . . . fxe4 27 ':xe4 f2-f4 . The Sth game of Spassky-Fischer
l:lxc3 . world championship match (Reykjavik 1 972)
went 1 0 lLlh4 h6 11 f4 (hoping for 1 1 . . .exf4
26 . . . nxf3 1
1 2 i.xf4 gS 1 3 eS! with complications
27 lLlxf3 favourable to Wh ite) 1 1 . . . lLlg6! 1 2 lLlxg6
Now after 27 'it;Jxf3 'ifd3+ 28 'it;Jf2 fxe4+ 29 fxg6. Here Boris Spassky made a serious
'it;Jg 1 Black has 29 . . . 'ifxg3+ 30 'it;Jh 1 ':f2 . strateg ic mistake by exchanging pawns on
27 . . . 'ifxd 1 eS, after which the position became static
28 lLlg5 and the white bishops had no scope. Robert
And Wh ite resigned, since he is the ex Fischer outplayed his opponent and went on
change and a pawn down , and the mate to win .
threat is easily parried by 28 . . . 'ifhS or 10 . . . 0-0
28 . . . 'ifd7. Black also has another possibilities. The
As you see , although the matu re Tal had not move made clearly shows his intention - to
lost his former resourceful ness, and he still prepare . . . f7-fS .
liked to attack and make combi nations, he 1 1 lLlf1
did this on a strict positional basis, endeav
ouring not to take the 'liberties' typical of his
youth .

The fol lowing game is q u ite d ifferent in


character to those exami ned earl ier - it
bears directly on the question of prophy
laxis. This is not surprising - playing Wh ite
was Tigran Petrosian.

Petros ian - Ivkov


Olympiad , N ice 1 974
Nimzo-Indian Defence
1 d4 lLlf6
2 c4 e6
Petros ian takes the opponent's plan into
3 lLlc3 Jt. b4 consideration and takes measures before
4 e3 c5 hand , by transferring his knight to g3. Black
5 Jt.d3 lLlc6 should possibly have changed plan by
6 lLlf3 i.xc3+ playing 1 1 . . . lLlg6, i n order to have the option
of jumping with his knight to f4 . Wh ite in turn
7 bxc3 d6
can react flexibly to this move , by placing his
8 e4 e5
knight on e3 and his pawns on g3 and f3 ,
9 d5 lLle7 and then advancing his h-pawn , taking
10 lLld2 advantage of his delay in castling. The
In the event of 1 0 0-0 Black makes the move order chosen by Petrosian is q uite
92 Does it pay to sharpen the Play?

venomous, but in recent times for some 1 9 . . . tbf6 20 .i.xg6!? hxg6 2 1 'iVxg6+ h8 2 2
reason it has not been employed . fS with a powerfu l attack). B u t t h e compen
11 . . . 'iVa5 sation for the pawn is hardly sufficient. Black
Black has decided to play on the kingside, probably does better to reject the pawn
and it is not clear why he moves his q ueen sacrifice in favour of 1 8 . . . tbg7!? 1 9 fxeS
to the queenside. If he was going to develop dxeS.
his queen at as, he should have done this a Petrosian wants to preserve his knight from
move earl ier, when , firstly, there was not the exchange and so he does not hu rry to take
reply .i.d2, and second ly, he would have decisive action. However, the opponent
retained the option of castling on the gains time to strengthen his position .
q ueenside. 17 . . . tbf6
1 2 .i.d2 tbe8 1 8 tbg5
1 3 tbg3 f5 The knight is very strongly placed here,
14 exf5 tbxf5 since the attempt to d rive it away by . . . h7-
Black has a d ifficult position after 1 4 . . . .i.xfS h6 leads to a weakening of the kingside.
1 S tbxfS tbxfS 1 6 'if c2 . 18 . . . ':ae8
1 5 'ii'c2 ! g6 19 f3 !
1 6 0-0 .i.d7 A typical Petrosian move . Having taken
It is important to note that, in contrast to the control of the e4- and g4-squares, he is
King's I ndian Defence, Black's knight can ready at a convenient moment to play g2-
not go to d4 - the square is defended by the g4, depriving the enemy pieces of the fS
white pawn. point.
19 . . . tbg7
20 g4!

17 tbe4
1 7 f4! ? suggested itself, in order to open up
the position and exploit the power of the two Of course, the ex-world champion prevents
bishops. Possibly Wh ite was concerned the exchange of bishops by 20 . . . .i.fS plan ned
about 1 7 . . . tbxg3 1 8 hxg3 e4!? ( 1 8 . . . exf4 1 9 by the opponent. All the black minor pieces
.i.xf4, i ntend ing ':ae1 and at some point are now shut out of play, and yet for the
.i.xg6) 1 9 .i.xe4 tbg7! ( 1 9 . . . .i.fS 20 g4 ! ? ; moment the situation remains unclear. For
Does it pay to sharpen the Play? ttJ 93

complete happiness White still needs also The answer is clear: in Black's favour, of
to cramp Black on the queenside with a2- course. In cramped positions you should
a4. exchange pieces! I think that after 26 . . . 'ii'x b6!
20 . . . 'ifa4 (27 'ifa2 Wb2 ) he would have retai ned
Borislav Ivkov misses an excellent chance to excel lent drawi ng chances.
complicate the play, pointed out by Petrosian: But now remember the situation before
20 . . . b5! 2 1 cxb5 c4 22 xc4 xb5. Wh ite's 1 7th move. I should like to ask: have
21 'ifb31 : b8 Petrosian's subtle manoeuvres been justi
22 .i.e2 1 'ifaS fied? Wouldn't it have been simpler, by
playing 1 7 f4! ? , to immediately 'cut the
23 a 4
Gordian knot'?
Thus, Wh ite has also succeeded in restrict
ing the opponent's possibil ities on the [At any event he should have struck in the
queenside. But even now the battle is not centre, without waiting for the opening of
yet over. lines on the queenside: 25 f4! (instead of 25
a5 ?!) 25. . . exf4 26 :'xf4 or 25. . . b5 26 axb5
23 . . . 'ife7
axb5 27 fxe5 dxe5 28 .i.e3 with advantage
24 h3 to White - Dvoretsky.]
White has to support the g4-pawn , to
26 . . . lIxb6?
prepare f3-f4.
27 'ifa3
24 . . . a6
Wh ite switches his queen to the kingside for
an attack, whereas the black queen lacks
any prospects .
27 . . . Wd8
28 'ife1 We7
The rook on its own cannot do anyth i n g , and
it is q u ickly driven off the second rank.
30 d3 e8
3 1 e1 lIb3
32 .i. e2 :b6
The prophylactic work has been success
fu lly accompl ished . There now follows what
is effectively the fi rst active move in the
game, and Black's position immed iately
collapses.
2S aS
33 f4! h6
Otherwise White would have had to reckon
34 fxeS WxeS
with 25 . . . b5, and after the capture on b5 with
the c-pawn - . . . c5-c4. 3S 'ii' x eS dxeS
2S . . . bS 36 ttJe4 hS
26 axb6 37 a3
What do you th ink, with wh ich piece should It is time to gather the harvest.
Black captu re on b6? In whose favou r is the 37 . . . ttJxe4
exchange of queens? 38 .:txf8+ 'iit xf8
94 Does it pay to sharpen the Play?

39 i.xe4 b3
40 i.xe5+ e8
41 'uf1
Black resigned .

In conclusion I will take the liberty of offering


one of my own games.

Tsariov - Vu lfson
Moscow 1 989
Sicilian Defence
1 e4 e5
2 tbe3 tbe6
In view of my retarded development, it
3 f4 e6
would be good to strike a blow in the centre,
4 tbf3 d5 by advancing the c-pawn . The pawn on b4
5 d3 slightly h inders the realisation of this idea.
Apparently my opponent was satisfied with Even so, 11 c4 bxc3 1 2 bxc3 , i ntending 1 3
the endgame after 5 . . . dxe4 6 dxe4 . I was c4 , was q u ite possible. My opponent found
aiming for more complicated play. a more cun n i ng way of carrying out this
5. . . tbf6 plan.
1 1 a3!? bxa3
6 e5 tbd7
1 1 . . . a5 1 2 axb4 cxb4 came into considera
7 g3 b5!?
tion . However, after 1 3 i.e3 (weaker is 1 3 c4
Usually this advance has to be prepared ,
bxc3 1 4 bxc3 i.a6) White would have stood
but here there is an opportun ity to carry it
better.
out i m med iately.
1 2 bxa3!
8 i.g2 b4
I had only reckoned on 12 l:txa3 'it'b6 1 3 c4
9 tbe2 g6?!
d4 with u nclear consequences.
It would have been better to continue i n the
12 . . . i.a6
same spirit: 9 . . . a5 and then . . . tbb6. But it
1 3 tbg5!
seemed important to me to halt the wh ite
pawns on the kingside. Another strong move . It transpires that if
1 3 . . . i.e7 there follows 14 c4! i.xg5 1 5
1 0 0-0 h5
cxd 5 ! . Therefore Black defends his knight on
The standard plan for Wh ite i n such posi
c6 .
tions involves the preparation of an offen
13 . . . 'fII e 7
sive on the kingside: h2-h3, g 3-g4 and at
some point f4-f5 . However, i n the g iven 1 4 e4!
instance he also has another very promising Th i n k what happens in the event of the
pla n . Try to fi nd it. pawn sacrifice being accepted.
If 14 ... dxc4, then 1 5 'iVa4 cxd3 16 tbc3. For
example: 1 6 . . . d2 1 7 'fIIx a6 , or 1 6 . . . i.b7 1 7
tbb5 followed by 1 8 tbe4.
All this looks extremely dangerous, but the
Does it pay to sharpen the Play?
CD 95

defence can be improved . Black should not 'ii'f3! with irresistible threats.
take the second pawn - it is better to play To be honest, I did not see the pawn
immediately 1 5 . . . i.b7! 1 6 dxc4 tLlb6 1 7 'iVc2 sacrifice, whereas my opponent saw it and
tLld4, reta i n i ng a defensible position. conscientiously tried to calculate it. But he
I preferred to keep the position closed , got bogged down in the mass of variations
which , alas, did not get Black out of serious and in the end he decided not to risk it. 'I felt
difficu lties. that I should play this, but I couldn't
14 . . . d4? calculate it fully', he explained after the
game. 'But why calculate it fully?', I asked in
surprise. ' If such an idea had occu rred to
me, I would defin itely have sacrificed . '
Having decided not to risk the sacrifice, my
opponent easily persuaded h imself that he
would win after 1 5 'it'a4 .
[And he was right - in this way White does
indeed achieve a significant advantage, by
simple means, without resotting to risk.
From the practical point of view the decision
taken by White is the most advisable -
Dolmatov.]
1 5 'it'a4 i.b7
1 6 :r.b1 tLlb6
17 'iWb5 nb8
My hopes were based on the lack of active
Black has to cover his gaping wound - the
possibilities for two of the white pieces - the
b-fi le.
knight on e2 and the bishop on c 1 . But such
18 tLle4 tLld7
possibilities appear after the positional
pawn sacrifice 1 5 f5. Another way of Otherwise the c5-pawn cannot be de
developing Wh ite's in itiative is 1 5 'it'a4 i.b7 fended .
1 6 J:.b 1 . 1 9 tLlf6+1
Thus, there is choice of two conti nuations. Wh ite's calculations were based on th is. I n
Which of them would you prefer? It is rather the event o f 1 9 . . . tLlxf6 h e has the decisive
difficult to calculate the variations fu lly 20 i.xc6+. But he clearly underestimated
(especially in the f4-f5 variation). At some my reply.
point you have to trust your intuition . 19 . . . cJ;d8!
1 5 f5!? gxf5 1 6 tLlf4 . The threat is 1 7 tLlxe6, 20 . . . i.a8 is threatened , and the knight at e2
after which the king can no longer be saved . is still out of play. And no forced win is
The best defence is 1 6 . . .l: h6. Then 1 7 apparent: 20 tLlxd7 xd7 ( i ntend ing 21 . . .
tLlxf7 ! ! cJ;xf7 1 8 tLlxe6! ':xe6 1 9 'iVxh5+ cJ;g7 i.e7) 2 1 i.xc6+? 'it'xc6 2 2 'iWxc6+ cJ;xc6.
( 1 9 . . . cJ;g8 20 i.d5 with the threat of 2 1
20 i.d2 i.a8
'i'g6+; 1 9 . . . :g6 2 0 i.d5+ cJ;g7 2 1 i.h6+!
20 . . . tLlxf6? is bad because of 21 i.xc6 ! .
:txh6 22 'ili'f7+) 20 i.d5! 'iVxe5 2 1 i.h6+!
with a mati ng attack. And if 18 . . . cJ;xe6 21 'iWa4 nxb1
(instead of 1 8 .. JIxe6), then 1 9 i.d5+ cJ;e7 22 l:txb1 tLlxf6
20 i.g5+ e8 21 i.xh6 is strong, as is 1 9 23 exf6 i.d6
96 Does it pay to sharpen the Play?

24 h4! Again we face a dilemma. After 25 . . . 'iVc8


It is not easy to increase the pressure, and White will strengthen his position by lIb5
my opponent sets a cu nning psychological and ttJe2-c 1 -b3. The alternative is the
trap. He defends against a possible . . . h5- queen sacrifice 25 .. J::t b 8. From the practical
h4, as though demonstrating that his i n itia point of view it is usually better to choose
tive on the queenside has evaporated . And I the more active conti nuatio n . Let us see:
swallowed this bait, by incautiously making 25 .. J:tb8!? 26 i... xc7 l:txb 1 + 27 'itt h2! (pre
a natu ral move . pari ng ttJg 1 -f3) 27 . . . 'itt x c7 . Now it would be
absu rd , of course, to win the a7-pawn by 28
24 . . . 'itt d 7?! xc6? After 28 ttJ g 1 [first 28 'ifc2 is more
25 a5! methodical - Dolmatov] 28 . . J:lb2 the ad
[Of course, it was unpleasant to allow such vantage is probably with Wh ite, but in a time
a stroke. But how otherwise could Black scramble that is what Black should have
have brought his rook into play? And after played .
all, the opponent was threatening to streng 25 . . . 'ife8?!
then his position by 'itt h 2 and ttJe2-g 1-f3 or 26 'ifb5?
i:!b5 and ttJe2-c1-b3. The pawn on f6 A pictu resque position would have arisen
seriously cramps Black, and his king feels after 26 ltb5! 'iVa6 (in playing 25 . . . 'ifc8 , I
vel}' uncomfortable in the centre. It has to be was relying on this pin) 27 ttJc1 l:Ib8 28 ttJb3
admitted that White has a serious advantage 'itt c8 . Now 29 ttJxc5? does not work because
- Dolmatov.] of 29 . . . l:Ixb5 30 ttJxa6 l:txa5. The opponent
plays 29 'itt h 2 ! , and what move can Black
make now? It transpires that in the middle
game he has ended up in zugzwang! For
example, 29 . . . l:Ib6 30 l:[xb6 axb6 31 xc6,
or 29 . . . f8 30 l:Ixb8+ 'itt x b8 3 1 c7+.
26 . . . 'ifb8 !
27 e1 a6
28 'ifxb8 l:txb8
29 l:Ixb8 xb8
By forcing the exchange of queens, Black
has equalised . My opponent's winning
chances had already evaporated , but sub
sequently he tried too hard to recover them
and he even went on to lose.
Does it pay to sharpen the Play? ttJ 97

Adden d u m

Petros ian - Sax to an ending wh ich is better for White . The


Tal l i n n 1 979 game Andersson-Hazai (Pula 1 975) contin
Pirc-Ufimtsev Defence ued 1 8 . . . ltJd4 1 9 f3 ltJe8? ( 1 9 . . . a6 20 ltJa4
1 ltJf3 g6 2 e4 .ig7 3 d4 d6 4 ltJe3 ltJf6 ltJd7) 20 ltJd5 c6 2 1 ltJe7+ 'itf8 22 ltJc8 a6
S .i.e2 0-0 6 0-0 .ig4 7 .ie3 ltJe6 S 'iVd2 eS 23 c3 ltJb5 24 a4 ltJbc7 25 .i.b3 i.. f6 26 ltJb6
9 dxeS dxeS 1 0 l:tad1 WeS 1 1 We1 lidS ltJe6 (26 . . . 'ite7 27 .ic5+) 27 .i.xe6 fxe6 28
1 2 1bdS+ 'ili'xdS 13 .l:td1 WfS 14 h3 .ixf3 .ic5+ 'itf7 29 ltJd7 'itg7 30 'itf2 .i h4+ 3 1 g3
1 S i.. xf3 .if6 32 'it e2 .id8 33 .i.f2 .i.f6 34 ltJc5 ltJd6
35 ltJxe6+ f7 36 ltJc5 i.. d 8 37 b3 .ia5 38
b4 .i.c7 39 a5 .i.b8 40 .i.e3 Black resigned .
I n the opinion o f Petrosia n , Black should
aim for the exchange of the dark-sq uare
bishops with 1 5 . . . h5!? followed by . . . 'ith7
and . . . .ih6.
1 6 ltJb1 ! :IdS 17 lIxdS 'iVxdS 1S e3 Wd3 1 9
ltJd2 .i.fS 20 'ili'b1 'iVbS
The exchange of queens would lead to an
unpleasant endgame for Black, roughly
similar to the Andersson-Hazai game.
21 'i!i'e2 ltJdS
If 21 . . . .i.c5, then 22 .i.g5 followed by the
advance of the queenside pawns with a2-a4
1 S . . . a6?! and b2-b4 .
The move which was recommended by 22 'iVb3 'ii'd 3?! 23 'iVe4 'iVd6 24 We2 We6
theory at that time, on the basis of the game 2S 'ii' d 3 ltJe6 (25 . . . 'iVe7 and 26 . . . ltJe6 was
Timman-Matulovic (Wij k aan Zee 1 974), better) 26 a3 'iVe7 27 b4 ltJdS 2S ltJe4 ltJd7
which went 1 6 ltJd5? ltJxd5 1 7 ':'xd5 ltJd4 29 .ig4! ltJe6 30 ltJaS! bS 31 ltJe6 'iVeS 32
1 8 .ixd4 exd4 1 9 'iVf4 'iVe7 20 e5 :e8 2 1 e4! ltJf6 33 exbS axbS 34 WxbS ltJxe4 3S
'i'xd4 .ixe5 with a q u ick draw. 'ii'e4 ltJd6 36 'ili'dS hS 37 .ixe6 fxe6 3S WeS
The direct 1 5 . . J::t d 8 leads after 1 6 ':xd8+ ltJfS 39 'iVe2 .i.g7 40 bS ltJd4 41 We4 Wd7
'i'xd8 1 7 Wd 1 Wxd 1 + ( 1 7 . . . 'iVe7!?) 1 8 .ixd 1 42 a4 ltJfS 43 'iVe2 Black resigned .
98
Mark Dvoretsky

Thoug hts about a Book

There are n o hopeless positions, there are only inferior


ones, which may be saved. There are no drawn positions,
there are only equal ones, in which one may play for a
win. But at the same time don 't forget that there is no such
thing as a winning position, where it is impossible to lose.
Grigory Sanakoev

I correspondence and it is u n l i kely that I will


n my life I have played just two games by instructive , but they are not terribly interest
ing. I n the given instance , fortunately, we
play any more. Nevertheless, on learn i ng meet not only some fascinating duels, but
that a games collection of Grigory Sanakoev, also the l ive person who played them - his
the 1 2th world correspondence champion, experiences , thoughts, assessments and
had been publ ished , I promptly acquired the advice. I am especially impressed by the
book. There were several reasons for this. enl isti ng (invariably at an appropriate mo
Fi rstly, I remember how grandmaster Si ma ment) of opinions by famous thinkers from
g i n , who in the mid 1 960s won the USSR the past. Chess is one of the fields of human
Correspondence Championship, spoke with culture, and we should not impoverish
great respect about the creative style of one ou rselves by sticki ng only to its na rrow
of his opponents - Grigory Sanakoev. After professional aspects.
read ing the book, I can confi rm with pleas Contrary to the author's convictions, I am
ure that Vlad i m i r Pavlovich was right. sceptical about the prospects for corre
Secondly, as a trainer I am always in need of spondence play. The emergence of comput
fresh and high-quality material. Many inter ers analysing at g randmaster level inevita
esting games are played in the world, but, bly creates the temptation to use their
since they are publ ished i n chess maga services to ach ieve good competitive re
zines or Informator, they become known not sults. Nowadays practically all players em
only to a trainer, but also his pupils. And yet ploy powerful computers for open ing analy
over-the-board players hardly deign to pay sis (the detri mental effect of this process on
any attention to the world of correspond the popula rity of chess is obvious - chess
ence play. This is a pity - from here one can fa ns are interested in a competition be
derive nu merous ideas, deep and vivid , tween individuals, not mach i nes). But in
which have been carefully developed in correspondence play a computer can be
home analysis. used throughout the game.
Books of games wh ich are l i mited to the However, there is no doubt that Sanakoev
moves and illustrative variations may be always has played and will play independ-
Thoug hts about a Book lZJ 99

ently. What mainly attracts h i m in chess is A propens ity for pretty moves
the creative search, the intellectual strug
If a player fi nds a spectacular and seem
gle, and only then the resu lt. A read ing of the
book pa ints the author i n a very attractive ingly tempting possibility, he is often hypno
light (I am sure that it is authentic, even tised by it and can no longer resist the
though I do not know Grigory Konstantinovich temptation. Probably all of us have lost
- an experienced reader cannot be de poi nts for this reason . I can not refrain from
ceived ! ) - a vivid , u ncompromising, self showi ng you a memorable example from
confident chess player, and an erud ite and one of my own games.
at the same time non-trad itional thinker.
I couldn't help beg i n n i ng with my overall Dvoretsky - Peev
very favourable impression of Sanakoev's European Champions Cup, Plovd iv 1 975
book, but it wasn't for th is that I 'picked up
my pen' (an obsolete phrase in the compu
ter age ! ) . The topic of my a rticle is certai n
critical aspects o f chess mastery, thoug hts
about which were in itiated by my read ing of
the book.
After choosing the games which to me
seemed the most noteworthy, I invited
grandmaster Zviag intsev to study them .
Vad im's task was an independent search for
difficu lt decisions at critical moments (of
course, without movi ng the pieces on the
board ), and someti mes also the playing of
that most tense episode of a game, when its
outcome was being decided . In many
instances the conclusions of the experi Wh ite's position looks to be won . True, 29
enced correspondence player and the young l:r.b6? 'ii'c 5+ is pointless, and 29 c4 'iVa6 ! or
over-the-board g randmaster did not coin 29 "a8+ c7 30 "a5+ b6 is unconvincing.
cide - such situations were additionally However, 29 :f4! is very strong , with the
analysed , discussed and interpreted. terrible threat of 30 :tc4 . If 29 . . . lbd6, then
I should mention that a deep exami nation of 30 .l:[ b6 "c5+ 31 :d4 and there is noth ing
even the most conscientious analysis is that Black can move . And in the event of
bound to reveal questionable aspects or 29 . . . .l:.d2 the following pretty variation is
even mistakes - chess is just too compli possible: 30 "a8+ c7 31 l::t x b7+ ! d6! 32
cated . Therefore the followi ng critical analy lId7+!? (32 lbe4+! is simpler) 32 . . . 'iVxd7 33
sis of episodes from Sanakoev's book of 'iWa3+ e5 34 lbxd7+ xf4 35 'iVc1 e3 36
games is not at all an attempt to cast doubts lbe5 and wins.
on it. I n my time I have written in similar U nfortunately, I was tempted by a showy
fashion about excellent books by Jan move plan ned in advance, which proved on
Timman and John Nunn, which beforehand verification to be not very effective.
I had used for training pu rposes with Sergey
29 lbd7?!
Dolmatov. Books with less interesting con
tent simply wouldn't have come with in our I nto a th ree-fold attack!
field of view. 29 . . . l::t xd7 1
1 00 Thoughts about a Book

The only defence. 29 . . ....xd7 30 "'a8+ is undoubtedly have to be taken seriously, but
bad , if 29 . . .'.txd7 there is the decisive 30 even so I th ink that the commentator
lIxb7+ 'iii> e 8 31 lIxf5 ! , while if 29 . . . lLld6, sign ificantly exaggerates the danger threat
then (if there is noth ing better) 30 lLle5 'iWd5 ening h i m .
3 1 lLlxf7. F o r example, after t h e natural 2 4 . . .llc8!?
After the move in the game I real ised that Sanakoev g ives 25 hxg6 fxg6 26 'iWh3 i.d5
the plan ned 30 "'a8+ 'iii> c 7 31 "'xh8 leads to 27 'iii> b 1 ! , preparing 28 lLlxd5 . However,
an i mmed iate draw: 31 . . ....c5+ ! (but not Black gains the advantage if instead of
31 . . . l:.d2? 32 'iWe5+ and 33 11f2 ) 32 'iii> h 2 26 . . . i.d5?! he chooses the sharp 26 . . . b4!
"'d6+ 33 'iii> g 1 "'c5+ 34 11f2 l:td2 35 ]1f1 27 'iWxe6+ 'iii> f8 28 l:tc4 bxc3 ! 29 ':xc7 cxd2+
lLlg3 36 "'f6 lLlxf1 37 "'xf7+ 'iii> b 8 38 'iii> x f1 30 'iii> x d2 ':'cxc7 31 'iii> c 1 i.c5 32 'ii'x e5 'iii> g 8.
l:.xf2+ 39 "'xf2 "'xc3 . But the attempt to Generally speaking , the knight is well
play on with 30 l:.xfS?1 gxfS 31 "'a8+ 'iii> c 7 placed at c3 - from here it prevents Black
32 "'xh8 proved even worse in view of 32 .. . from conven iently supporting his e6-point
'iWe4 1 . After 33 1:.f1 %:td2 34 'iWg7 peace was by . . . i.d5, and in some cases it can go to
nevertheless concluded , although Black's e4 . Therefore 24 . . . b4 !? suggests itself.
position is now somewhat better. Sanakoev thinks that after 25 lLld 1 l:tc8 26
lLle3 Wh ite has a clear adva ntage (i ndeed ,
any min ute now the knight will jump to g4).
An exami nation of Sanakoev's games
showed that he is characterised by this But why let the knight out from d 1 ? I n stead
tendency to choose pretty moves , even if of 25 . . . .:tc8 Black has the Significantly
this is sometimes at the expense of their stronger 25 . . . i.c5 ! ? 26 'iVh3 i.d5. Now the
qual ity. rook is intending to go to c8 , in the event of
27 hxg6 fxg6 Black has everyth ing safely
Engel - Sanakoev defended , in reply to lLle3 there always
follows . . . i.xe3, while the consequences of
An niversary Tou rnament of the
27 i.xb4 i.xb3 (27 . . . i.xb4 28 l:txb4 l::t c8
Romanian Chess Federatio n , 1 976-79
also comes into consideration) are uncer
tai n . It is clear that Black has the right to go
i n for this.
25 . . . i.e 7! would appear to be even stronger.
Wh ite can not play 26 lLle3? i.xg5, and 26
hxg6 ':xd2! is also unfavourable for h i m . But
after 26 i.e3 the knight can no longer go to
e3, and Black calmly plays 26 . . J::tfd8 ,
i ntending 27 . . . l:txd 1 + or 27 . . . 'iWa5 .
T h e above considerations are prosaic. By
contrast, the solution found by Sanakoev
was highly spectacu lar.
24 . . . i.a3 ! ?
Now White loses immed iately after 2 5
.:. 1 h 2 ? ':xd2! 26 l:txd2 'iWxc3 . If 25 bxa3
'Only an immediate counterattack can save Black was intending 25 .. J:txd2 ! 26 'iii> x d2
Black', writes Sanakoev. Wh ite is i ntending l:td8+ 27 'iii> c 1 'ii'x c3 28 'iii> b 1 lId2 29 'ii'c 1
25 'iWh3 followed by 26 hxg6. His th reats i.xf3 30 : 1 h3 11e2 with advantage. How-
Thoughts about a Book ttJ 101

ever, Wh ite's play can b e improved b y 30 would have hardly allowed his opponent 'off
hxg6! (instead of 30 l::t 1 h3?) 30 . . . iLxh 1 (in the ropes' so soo n . But what told here,
the event of 30 . . .fxg6 3 1 l:t 1 h3 Black no apparently, was the magic of a pretty move ,
longer has 31 . . . .:e2?? because of 32 l:txh7) forci ng h i m to convi nce h imself that after
3 1 gxf7+ 'itt xf7 32 l:txh7+, for example: other continuations Wh ite would gain the
32 . . . 'itt g 6 33 l:txh 1 e4 34 J:th6+ 'itt x g5 35 advantage.
I:txe6 e3 36 'ifg 1 + 'itt f4 37 J:te4+! 'itt xe4 38
'iVg4+ 'itt d 5 39 'ifd7+ with perpetual check. A. Zaitsev - Sanakoev
In the game there followed 25 lDb1 1! . Here 6th USSR Championsh i p , 1 963-65
Sanakoev resisted the temptation to again
play 'for brilliancy' : 25 . . . J:txd2? ! 26 lDxd2
'iVc3. In the event of 27 bxa3?! l:tc8 28 'ifd 1
l:td8! the game ends in a draw after both 29
'itt b 1 I:txd2 30 'ifc1 iLxf3 31 hxg6! (we have
already seen this position in our analysis of
the 25 bxa3 variation), and 29 hxg6!? l:txd2
30 gxf7 + 'itt f8 31 .l:txh 7 .l:txd 1 + 32 l:txd 1 .
Stronger is 27 'ifd3! iLxb2+ 28 'itt b 1 iL a 1 29
'iVxc3 iLxc3 30 lDe4 iLxe4 31 fxe4 - here it
is Black who would have to fight for a d raw.
25 ... iLc51 26 'ifh3 'ifc61 27 hxg6 'ifxf3 1 28
gxh7+ 'itt h8 29 'ifxf3 iLxf3 , and in the
endgame Black had an obvious advantage ,
which he successfully converted .
Zviagi ntsev also hit on the move 24 . . . iLa3. 25 . . . iLxh4
But he was not sure about his choice , since 26 .:t. h 1 iLxg3!
he calculated that Wh ite could force a draw, The exclamation mark is mine. Sanakoev
and he wondered whether instead he hi mself considers the move made by h i m to
should play the complicated position after be dubious. This is what he writes:
24 . . . b4 !? 'The temptation prove too great . . . I recal/ed
25 hxg6! ':xd2 ! that ''the wise man understands that it is
26 ':'xh7 'ii'x c3 ! simpler to deny himself a passion than to
27 :h8+ 'itt g 7 struggle against it afterwards" (Franc;ois La
28 I:t1 h7+ Rochefoucauld), but with the chance of a
sacrificial attack against Zaitsev, I thought
28 l:t8h7+ is just the same.
"No, I can 't chicken out!"
28 . . . 'itt x g6
'A s for the purely objective assessment of
29 ':'h6+ <3;g7 the manoeuvre . . . iLf6xh4xg3, 26. . . iLg5
The king can not move forward (29 . . . 'itt f5? was undoubtedly stronger. After the modest
30 'ifh3+ 'itt f4 3 1 I:t h4+ 'itt e 3 32 ':'e4+! iLxe4 reply 27 11xh5, by 27 . . . iLxd2 28 iLxd2 bxc4
33 fxe4+ and 34 'ifxc3), and so th ings end in Black would have gained the initiative on the
perpetual check. queenside in the absence of any serious
Of course, Sanakoev saw this variation and counterplay for the opponent, which would
gave it i n his book. He is a very combative have promised long months of very pleas
player and under other ci rcumstances he ant analysis in the range from 'better' to
1 02 Thoug hts about a Book

'much better'. The more critical 27 cxb5 30 :h2 h4 31 liJf1 bxc4 32 bxc4 J.xc4 33
would have allowed the pawn sacrifice J.xc4 l:r.xc4 34 'ii' b 3 :ec8 35 J.xf4 exf4 is
27 . . . h4! 28 bxa6 'ike 7 followed by . . . hxg3, bad for Wh ite.
obtaining an attack on the dark squares,
,
which would be not at all easy to parry.
A player's impression of a game he has
played usually depends strongly on its
result. If Sanakoev had won (as we see, he
had every basis for doing so), the piece
sacrifice would probably have been awarded
two exclamation marks. But he lost, and
hence the dou bts about the qual ity of the
decision take n .
I n fact, after Sanakoev's recommendation
26 . . . J.g5 27 l:r.xh5 J.xd2 28 J.xd2 bxc4 29
bxc4 Black has a good game, but noth ing
more. And yet the piece sacrifice was not
only tempti ng, but also very strong . You only
have to look at the position arising with in 2- 30 . . . h4
3 moves, and the sure feeling is that Black's Sanakoev makes no comment on this
attack is fu lly correct. move, although it is not self-evident. After
In such situations, 'correspondents' aim to 3 1 :g 1 the queen will temporarily have to
analyse variations as deeply and accu rately retreat - there is no longer a check at h4.
as possible. But over-the-board players, However, then there follows . . . h4-h3 and
who have neither a sufficient reserve of the h4-sq uare again becomes accessible to
time, nor the right to move the pieces on the the q ueen .
board , are forced , by contrast, to cut short Black had another tempting attacking possi
their calculation at the first conven ient bil ity, suggested by Zviagi ntsev: 30 . . . bxc4 !
moment and evaluate the position reached . 3 1 bxc4 J.xc4 32 liJxc4 :xc4 33 'ii' b 3 :ec8 ,
This is why correspondence players are after which , in my view, neither 34 J.xf4
bound to be less good at making correct l:r.c2+ 35 J.e2 exf4 , nor 34 J.xc4 'ifg2+ 35
assessments than over-the-board experts - e3 'ifxh 1 36 J.b2 'ifh2 (and if 37 ':c1
simply, here they have less experience, ':xc4 ! ) leaves Wh ite any real hopes of
since they solve most of their problems saving the game.
analytically. 31 1:1g1 'ii' h 6
Of course, any observation of this sort, even 32 'ifb61
if in general it is correct, ca nnot be extended Wh ite has to prepare the king move to e 1 ,
to every eventual ity in l ife . For example, I which did not work immed iately because of
am familiar with the games of M i khail 32 . . . liJg2 + .
Umansky, another world correspondence 32 . . . h3
champion , and they are impressive pre
33 e1 1:1c51
cisely for their depth of strategy.
An excellent move, cutti ng off the q ueen
28 'ii' e 3 'ii'e 7 from the important e3- and f2-squares. If 34
29 f2 'ii'g 5 'ii'x d6 there follows 34 . . . 'ii' h 4+ 35 d 1 'ii'f2 ,
30 J.f1 and then . . . 1:1c5-c8-d8 .
Thoug hts about a Book ltJ 1 03

34 .i.a3 th i n ks that after 34 . . . 'ii' h 4+ 35 ltd 1 'ii'f2 36


l:[h 1 'White has no particular problems ', but
in fact there a re problems, and very serious
ones: 36 . . . lLlg2! 37 .i.xc5 dxc5 (or even
37 . . . 'i'e 1 + 38 ltc2 'i'xa 1 39 .i.f2 'ii'x a2+ ) 38
ltc2 lLle3+ 39 ltc3 b4+ 40 ltd3 lLlf5 ! .
Why d id Sanakoev underesti mate these
possibilities, and why did he reject a
favourable variation that he had calculated?
Wel l , firstly, he was not altogether sure
about the assessment of its concluding
position . But it was mainly because he was
tempted by a possibil ity of 'playing for
brilliancy' .
34 . . . lLld5?
I n such a tense situation it is rarely possible 35 exd51
to conduct an attack, simply by making 'White has no reason to plunge into the
common sense moves. At some point you maze of variations such as 35 "fixd6 'iVe3+
have to exert you rself, i n order to find and 36 ltd1 b4 37 cxd5 "fixg 1 38 i.. xb4 .:tcc8, or
accu rately calculate a concrete way to the make a dubious attempt to clarify the
goa l . In correspondence play this is much position with 35 .i.xc5 lLlxb6 36 i.. xb6 bxc4
simpler, of course, than in a normal game. 3 7 bxc4 llb8 38 .i. f2 r1b2, when in either
Sanakoev saw a convincing solution, which , case it is again not altogether clear how to
however, was also found by Zviagi ntsev. combat the h3-pawn. ' (Sanakoev)
34 . . . lLlg2+1 35 ltd1 (Wh ite loses qu ickly 35 . . . 'ii'e 3+
after 35 .i.xg2? 'ii'e 3+ 36 ltf1 hxg2+ 37 36 ltd1 'ii'x g1
l:txg2 .i.h3, while 35 ]:txg2 hxg2 36 .i.xg2
37 dxe6
'i'e3+ 37 lt d 1 bxc4 leads to a transposition
Black has no time for the captu re on e6: a
of moves) 35 ... 'ii'e 3 36 lixg2 hxg2 37
characteristic variation goes 37 . . . fxe6 38
i.. x g2 bxc4 38 .i.xc5 dxc5 39 bxc4 .i.xc4!
.i.xc5 dxc5 39 ltc2 (39 'i'c6 ! ? followed by
(39 . . . 'ii'd 4? 40 l:tc1 l:td8 41 'ii' a 5) 40 lLlxc4
40 ltc2 is even more accu rate) 39 . . . h2 40
'i'd4+ 1 41 lLld2 (4 1 lte2 or 41 ltc2 is
.:tb1 ! h 1 'iV 41 .i.d3. But otherwise Wh ite
completely bad because of 41 . . . 'ii'x c4+)
captu res on f7, severely weakening the
41 .. JWxa 1 + 42 lte2 'ii'x a2, and Black is
position of the opposing king. On reach ing
clearly close to a win.
this point in his calcu lations, it is probable
I should mention that there are also alterna
that an over-the-board grandmaster would
tive ways of conducting the attack, which
have intuitively rejected 34 . . . lLld5 and looked
are no worse than the above variation . For
for someth ing else.
example, the captu re on c4 can be made
37 . . . h2
not only with the pawn , but also the bishop:
37 . . . .i.xc4!? 38 bxc4 (38 .i.xc5 .i.e2+; 38 38 exf7+ ltxf7
It)xc4 'i'g 1 + ) 38 . . . 'ii'x a3 with a winning 39 'iVxd6 'ii'd 4?1
positio n . I n stead of 35 . . . 'iIi'e3 , very strong is A detailed analysis convinced Black that
35 . . . lLle3+ ! ? 36 lte2 lLlc2 or 36 ltc1 h2 37 after 39 . . J:tcc8 40 'ii'd 5+ his king would
l:[h 1 lLlxf1 38 .:txf1 h 1 'i'. Finally, Sanakoev come under a decisive attack. I n my view,
1 04 <t> Thoug hts about a Book

there things are not altogether clear. For Razuvaev - Bel iavsky
example, after 40 . . . f6! Sanakoev gives 4 1 47th USSR Championsh ip, Minsk 1 979
ttJe4+ g6 4 2 c2 bxc4 4 3 bxc4 h 1 'iV 44
.i.d3 ! , and now for some reason 44 . . Jbc4+?
45 'iVxc4 'iVxa 1 46 ttJg3+ etc. But I don't see
how Wh ite can checkmate his opponent
after the immed iate 44 . . . 'iVxa 1 .
The move in the game led to a hopeless
ending.
40 i.xc5 'iWxd6
41 i.xd6 h 1 'iW
42 c2 'iWh6
43 c5 'iWe3
44 a4!
Wh ite has both a material advantage (three
minor pieces for a queen) and a positional
advantage, which he successfully con Yu ri Razuvaev restricted hi mself to the
verted. immediate regaining of the pawn 1 8 ttJxe4,
which allowed Black to equal ise by 1 8 . . .
Calculation horizons i.xe4 1 9 'ii'xe4 'iWd5 ! . There followed : 20
'iVxd5 exd5 21 l:lfd 1 e6 22 f1 i.d6 23
i.xd6 Draw.
As has already been mentioned , it is natu ral The interposition of a rook move to d 1
for over-the-board players to aim to cut suggested itself. However, in the event of 1 8
short their calculation of variations as early l:r.fd 1 ? ! Black h a s a n excellent reply:
as possible. In this way they save time and 1 8 . . . 'ii'e 8! 1 9 lId7+?! g6, when 20 ttJxe4?
energy, but sometimes they delve i nsuffi is bad because of 20 . . . i.xf2+. The q ueen
ciently deeply into the position , overlook should be attacked with the other rook.
latent tactical or strategic resources, and as 1 8 .ucd1 ! 'iWe8
a result miss the strongest continuations.
After 1 8 . . . 'ii' b 6 1 9 ':'d7+ and 20 ttJxe4 the
What can be done: 'real life is, to most men,
in itiative remains with Wh ite.
a long second-best, a perpetual compro
mise between the ideal and the possible. ' 1 9 l:td7+ g6?
(Bertrand Russell). The natu ra l , but incorrect move . 1 9 . . . f8 ! is
stronger.
[After 1 9. . . f8! it is not clear that White has
much compensation for the pawn, e.g. 20
l:tfd1 e5. It seems dubious to assert that 1 8
l:r.cd1 is better than 1 8 ttJxe4 Translator.]
-

20 ttJxe4 e5
After calculating this far, Razuvaev rejected
1 8 l:r.cd 1 . But he was wrong!
21 ttJxc5 l:.xc5
22 l:[xg7+! xg7
Thoug hts about a Book ttJ 1 05

23 'ii'x a7+ I have to admit that i n itially I did not agree


White regains the pawn and emerges two with Sanakoev's assessment, thinking that
pawns up. after the prophylactic move 29 :e3 ! ? Wh ite
Many correspondence games provide us would retai n a g reat advantage by simple
with excellent training material for overcom means. A more detailed verification did not
ing this psychological barrier, for expanding confi rm th is conclusion . The opponent re
our calculation horizons. After a l l , a corre pl ies 29 . . . e4! 30 ttJd4 d5. After this I
spondence player usually conti nues his considered 3 1 h4 l:tad8? 32 g3 ttJd3 33
analysis at a point where the over-the-board ttJac6 l:td7 34 l:.dxd3 and 31 g3 ttJxh3+ 32
player would probably stop. It is important g2 ttJg5 33 'ii'x d5. However, in the first
only to select examples in wh ich the variation Black has the excellent move
problems facing a player are not analytica l , 31 . . . 'ii'f7 ! , not al lowing g2-g3; and in the
but are o f a thematic or psychological second - instead of 32 . . . ttJg5? he can play
natu re. 32 . . . 'ii'd 7! followed by . . . 'iVf5 or . . . 'ii'g 4.
But why not drive the knight away immedi
'The longest and most complicated trap in ately? The point is that the opponent has a
my career was probably the one I carried clever tactical resou rce , enabling him to
out in the following game. (Sanakoev)
'
remain a pawn ahead .
29 g3! ! ttJxh3+
Sanakoev - Shevechek Wh ite gains an overwhelming advantage
VI World Championship 1 968-70 after 29 . . . d5?! 30 ':'e3 ttJxh3+ 31 g2 e4
(31 . . . ttJg5? 32 ttJxg5 hxg5 33 l:th 1 + ) 32
xh3 exf3 33 l:.xf3 .
30 g2 ttJxf2 !
On discovering this stroke , the over-the
board player would almost certainly cut
short his calculation and look for a safer way
- there you have the first psychological
barrier. But Sanakoev continued studying
the position and ca me to the conclusion that
here Wh ite wins by force. This means that
what resu lts is an excellent trap: the
complete ill usion of a blunder is created ,
and the opponent will probably decide that
Wh ite simply overlooked the captu re on f2 .
31 xf2 'ii' b 6+
How should this position be assessed? The 32 g2 :txa5
author of the book writes:
33 'ii'f7 11aa8
'Black appears to have achieved his aim.
The only defence.
The knight at a5 is still out of play, on his
next move he will make the long-awaited 34 ttJ h4
. . . d6--d5 advance, and although in the
forthcoming play White, with his sound extra (see diagram)
pawn, certainly has the better chances, the
outcome seems completely unclear. '
1 06 Thoug hts about a Book

look for other defensive possibil ities .


Sanakoev examines the variation 34 .. :ii'a 7!?
35 It)g6+ 'ifth7 36 It)xf8+ :txf8 37 'ili'xa7
':'xa7 38 ':'xd6 l:ta2+ 39 'ifth3 and thinks
that Wh ite wins easily, since the counterat
tack along the 2nd ra nk 39 . . . .l:.ff2 is point
less - after 40 1:Ixe5 the king escapes from
the checks via g4. I n fact it is also possible
to counterattack along the 3rd rank: 39 ...
nf3 ! ?

'The end of the trap ', writes Sanakoev. This


is inaccu rately stated - in fact the calcula
tion continues. What does Wh ite want? 35
It)g6+ 'ifth7 36 'iWf5 is not dangerous i n view
of 36 . . . 'iftg8, while the real threat of 35 1:Ig4
followed by 36 lt)g6+ 'ifth7 37 lt)xf8+ can be
parried by pinning the rook.
34 . . . 'ii'c 6
Here we face a second psycholog ical
barrier - incidentally, Zviag i ntsev stu mbled
at it, after successfully overcoming the first.
35 It)g6+ 'ifth7 For example, 4 0 nc6 (Wh ite also h a s other
36 'ilYf5 tries : 40 c4!? ':aa3 41 :g4 g5 42 b5 ':a1
It turns out that with his q ueen on c6 Black is with the th reat of 43 . . . :f2 , or 40 ':d7 ! ? .uxc3
no longer able to defend, since now 36 . . . 'iftg8 41 11g4 'ifth8, but here too the outcome
encou nters the dagger-blow 37 'ilYxf8+! remains unclear) 40 . . . .l:.c2 (40 . . . 11a3!? 4 1
lixf8 38 lt)e7+. An excellent idea ! llec4 .l:. b 3 4 2 'iftg4 e 4 ) 4 1 c 4 ':cc3 4 2 ':'g4
36 . . . 'ilYxc3 g5.
37 ':'c4! I s there a w i n here? If there is, t h e fact can
Black resigned in view of 37 . . . 'iVxc4 38 probably be establ ished only in a corre
It)xe5+ or 37 . . . 'ilYb2+ 38 l:tc2 'ii'x b4 39 spondence game. The problem has become
It)xf8+ 'iftg8 40 'iWh7+ 'iftf8 41 ':f1 + 'ifte7 42 purely analytical and everyth ing hangs by a
'ilYxg7+. thread .
On purely aesthetic g rounds one would like After 43 b5 ':'f2 44 :e4! :ff3 (44 . . . h5 45
the deep and pretty study, created by White , l:txe5) 45 l::t x e5 or 45 'iftg4!? it would appear
t o have no refutatio n . Alas, in practical that Wh ite wins. However, there is also
games this does not often happen - the 43 . . . .l:tf5!? 44 c5 h5 45 l:ta4 l:tf7 ! , and if 46
defensive resou rces in chess are just too b6? ! , then 46 . . . ':'c2 ! (th reatening 47 . . . g4+
g reat. Sanakoev's opponent was by no 48 'ifth4 :g7) 47 :a7 l:txa7 48 bxa7 :a2.
means doomed . Let's take back the natural Another interesting possibil ity is 43 .. J:tfe3!?
but objectively weak move 34 .. :iVc6? and 44 c5 e4 (44 . . . .:.e2? 45 l:ta4 ! ) 45 b6 .uc1 !
Thoug hts about a Book ttJ 1 07

(but not 45 . . .1le2? 46 b7 .:I.cc2 47 .:I.xg5) 46 displays itself mainly in a situation where
b7 (a pretty draw results from 46 ':xg5 hxg5 there is a choice between roughly equ iva
47 b7 .:I.b3 48 l:[b6 ':xb6 49 cxb6 e3 50 b8'if lent possibil ities (in particular, in the choice
e2) 46 .. J:th 1 + ! 47 'itt g 2 ':b1 48 .:I.xg5 (of of a particular opening strategy). Of course,
course, not 48 l:[b6?? .:I.e2+ 49 'itt h 3 .:I. h 1 this is merely a scheme - in fact th ings are
mate) 4 8 . . . .:I.xb7 4 9 %1 h 5 ':g7 a n d the far more compl icated . There are many
position is most probably drawn . borderl ine, problematic situations, and also
The actions of the two players can probably decisions are sometimes taken (and qu ite
be improved , but this is all rather compli rightly) on psychological grounds. 'An expe
cated and unclear, and in practice Black rienced player often chooses a certain
retains real chances of saving the game. continuation, not because he is sure that it is
Later I found another way of defending , one the best of all those possible, but exclu
which is perhaps more reliable. sively on the basis that it gives the best
34 . . . 'itt h 71 35 lDg6. Now 35 . . . 'ifc6? and practical chances' (Beniamin Blumenfeld).
35 . . . 'ifa7!? lead to variations wh ich have You can deliberately embark on a path ,
already been considered . There is also the known to be not the strongest, merely to
clever attempt 35 . . . .i.e7?! , hoping for 36 give the play a character which is desirable
t'iJxe7? ':xe7 37 'ifxe7 ':'a2+ 38 'itt h 3 'iff2 for you and undesirable for your opponent.
with an attack. Wh ite retains the advantage, The only q uestion here is the acceptable
by contin u i ng 36 'ii'f5! :a2+ (bad is 36. Ag8 measure of such psycholog ical play, and the
37 'ife6+ 'itt h 7 38 lDxe5! 'ifb7 39 'ifg6+ 'itt g 8 lim its which should not be overstepped.
40 lDf7 ) 37 'itt h 3 ':f2 38 lDf4+ 'itt h 8 39 'ifd 7 . It would be very interesting and usefu l ,
But Black c a n p l a y 35 . . . d51 36 ':g4 'iff6 3 7 u s i n g an analysis o f concrete examples, to
'i'xf6 gxf6 38 ':xd5 h5 3 9 lDxf8+ :xf8. follow how a player's style influences the
With material eq ual, White's position is decisions he takes. U nfortunately, as far as I
preferable, thanks to his two connected know, as yet no one has carried out such a
passed pawns, but even so a draw is the study - everything has merely been re
most probable outcome. stricted to speculative attempts to construct
various style classifications.
It is hardly right to call the idea carried out by
Sanakoev a trap. After all, as we have Sanakoev - Lungdal
establ ished , 'falling into the trap' has not
6th World Championsh ip, 1 968-7 1
been refuted and it was objectively Black's
best chance. No, essentially this is a
complicated combination with the sacrifice
of two pawns on h3 and f2 .
By embarking on the combination , Sanakoev
played in full accordance with his style - he
usually prefers a tactical way of solving the
problems facing h i m . The q uestion of chess
styles is very important and deserves to be
dwelt on for at least a short time.
It is logically clear that continuations which
are obviously the strongest, whether posi
tional or tactica l , should be chosen by a
player irrespective of his style of play. Style
1 08 <;t> Thoug hts about a Book

Wh ite stands better, of course, and the only main d ifficulty here is not in fi nding Wh ite's
question is how to extract the maximum move , but in assessing its consequences.
possible from the position. 18 . . . 'ii'xd4+
After 1 8 cxb4 ?! "ilxd4+ 1 9 1:!xd4 l:!c2 20 1 8 . . . bxc3 1 9 bxc3 'ii' x d4+ is less accu rate,
iL.d3 1:!xb2 21 :Lc1 d7 22 lIc2 l:!xc2 23 since Wh ite can choose between 20 l:.xd4
iL.xc2 White would have lost the greater part (as in the game) and 20 cxd4. Black can not
of his advantage. ' (Sanakoev). avoid the opening of the b-file: 1 8 . . . a5? 1 9
Let us try refining this variation with 1 8 'ii'x c5 ':'xc5 2 0 cxb4 axb4 2 1 l:.d4 o r 21
.xc5 l:txc5 1 9 cxb4 1:!c2 20 f2 (20 iL.d3 %:tbc1 is bad for him.
l:[xb2 2 1 1:!db1 lId2 22 b5 a5!) 20 .. J:txb2 2 1
19 l:lxd4 bxc3
l:[db1 l:txb 1 22 1:!xb 1 d7 2 3 lIc1 - here
Wh ite, who has seized the c-file and brought 20 bxc3 ':'c7
his king towards the centre, has a very 21 l:ldb4 iL.c8
sign ificant advantage. But Black can im
prove his defence by sacrificing a pawn with
20 . . . e7! (instead of 20 .. J:txb2) for the
sake of retaining control of the open file and
the 2nd rank. For example, 21 lIdc1 ? ! l:lhc8
22 l:txc2 :xc2 23 b3 d4! 24 l:ld 1 %:'xa2 (now
it is clear why Black did not place his king on
d7) 25 l:[xd4 iL.xg2.
A dangerous plan was suggested by grand
master Stefan Kinderman n : 1 8 .xc5 l:[xc5
1 9 l:[ac1 !? bxc3 20 b4! l:tc7 21 l:td3 . In the
bishop endgame arising after 21 . . . d7 22
l:[dxc3 l:[hc8 23 1:!xc7+ l:[xc7 24 l:[xc7+
xc7 Black faces a d ifficult defence. His
only hope: 25 . . . b6 (followed by 26 . . . iL.c6
or 26 . . . a5) is not hard to d ispel , by playing On reach ing this position , the over-the
25 a4! followed by a4-a5 and f2-e3-d4- board player would most probably terminate
c5. The pawn ending arising after 25 . . . ii.c6 his calculations and reject the plan begin
26 a5 iL.b5 27 iL.xb5 axb5 28 f2 is lost (the ning with 1 8 l:.ab 1 (as Zviag intsev did). In
reader can check this for h imself). fact, what has Wh ite ach ieved? Well , he has
seized control of the b-file, but on it there are
Black does better to avoid the exchange of
no targets to attack. On the other hand , his
rooks, by choosing 21 . . . d4! 22 ':xd4 e7
q ueenside pawns have become weak,
with an acceptable positio n .
wh ich ensures that the opponent has real
1 8 1:!ab1 1 1 counter-chances, even if (as is very prob
'A mysterious rook move' - a s Aaron able) Wh ite succeeds in winning the a6-
N imzowitsch expressed it. By defending his pawn. No, Sanakoev's decision does not
b2-pawn , Wh ite strengthens the th reat of 1 9 look convincing, it is somehow unstrategic!
cxb4 . This assessment can be corrected only by
'Such a continuation can be more difficult to continuing to study the position and fi nding
find than a forcing combination involving the a fu rther plan for White. In fact there is
sacrifice of several pieces', writes Sanakoev. noth ing u nexpected here , since , as N imzo
He is rig ht, although it seems to me that the witsch emphasised long ago, 'the entry into
Thoug hts about a Book ltJ 1 09

enemy territory, in other words into the 7th domi nation of the 7th rank, it would appear
and 8th ranks, forms the logical conse that Black can hope for a draw.
quence of play in a file. ' But it is q u ite Wh ite should not hu rry with the captu re of
impossible to establish in advance how the a6-pawn. It is far more dangerous to
dangerous for the opponent is the doubling interpose the check 24 ':'a7+ ! . For example,
of rooks on the 8th rank. Here a detailed 24 . . . d8 25 :tb8! (weaker is 25 ':'xf7? :tc7 ;
analysis is needed , which is not easy to a not altogether clear rook endgame arises
carry out, even playing by correspondence. after 25 i.xa6 i.xa6 26 ':'xa6 d7 27 l:tb7+
But at the board , with l i m ited time for l:tc7 28 l:txc7+ xc7 29 lla7+ c6 - the
thought, it is not worth even trying to passed d-pawn and the activity of Black's
calculate the variations accu rately - one king ensure him counterplay) 25 . . . l:tc7 26
has to rely on i ntuition. It would be interest :taa8 (th reatening an eternal pin on the 8th
ing to know - what does it suggest to you ran k after 27 i.xa6) 26 . . . a5 27 ltxa5 llb7 28
here? l:tba8, or 27 .. Jk1 + 28 f2 1:c2 29 e1 (but
22 l:tb8 ! not 29 .:taa8? c7 30 e1 1:xe2+ 31 xe2
Why doesn't Wh ite defend his c3-pawn? i.a6+) 29 . . . c7 (the th reat was 30 1:1a7 or
Probably, so as not to allow the opponent 30 11aa8 followed by 31 i.a6) 30 llb3 and
time for the following arrangement of his 3 1 l:ta7+ . I n this variation Black is appar
forces: 22 II 1 b3 e 7 23 llb8 (23 f2 is ently unable to disentangle h imself.
better) 23 . . . 1:d8 24 1:a8 d4 ! . In the event of 24 . . . i.d7 (instead of 24 . . . d8)
25 :tbb7 l:td8 Wh ite does best to play 26 a4 !
The tempting move 2 2 c4 would b e justified
after 22 . . . dxc4 23 l:txc4 l:txc4 24 i.xc4 d7 with an overwhelming advantage. 26 i.xa6? !
25 .ub8 l:.d8 26 lIa8 i.b7 27 :a7 and 28 is weaker: 26 . . . ':'a3! (26 . . . e8? 27 1:xd7
i.xa6. But Black is not obl iged to exchange l:txd7 28 i.b5; 26 . . . 1:1c1 +?! 27 f2 llc2+ 28
on c4 - 22 . . . e7! 23 cxd5 exd5 is stronger. i.e2) 27 i.e2 l:txa7 28 Iha7 e8 29 a4
l:tb8 30 a5 l:t b 1 + 31 f2 l:[b2 with a
22 . . . e7 probable draw.
23 l:ta8! 24 :b3 f6
Sanakoev consistently pursues his course. If 24 . . . i.d7, then 25 1:xa6 1:ec8 26 1:1aa3 .
In the event of 23 i.xa6?! ltd8(e8) he would Weaker is 25 l:txe8+ i.xe8 ( 2 5 . . . xe8) 2 6
have either had to agree to the exchange of i.xa6 11 a 7 27 ll b 6 llc7 , when 28 l:[ b 7 i s
bishops, which favou rs the opponent, or unfavourable in view o f 28 . . . l:txb7 29 i.xb7
give up his c3-pawn . i.b5! with the threat of . . . d7-c7.
23 . . . lle8 25 i.d3!
Let's consider 23 . . Jbc3 . An interesting Not immed iately 25 f2? fxe5 26 fxe5 l:tf8+
variation goes 24 i.xa6 %:te8 25 :a7+ (25 27 e3 :f5.
l:[bb8 i.d7) 25 . . .f8 26 i.xc8 (26 i.b5 lIe7 25 . . . fxe5
27 l:ta8 :tc7) 26 . . ..uexc8 27 l:[bb7 %:tc1 +
26 fxe5 h6
(27 . . . g8 28 h4 l:tf8 is also possible) 28 f2
27 f2
l:t1 c2+ 29 g3 .:t8c3+ 30 h4 g8! 3 1 f5!
(31 ':'xf7? ':'xg2) 3 1 . . . exf5 32 e6 (32 l:tb8+ Wh ite's advantage has become obvious
l:[c8 33 :taa8 g5+ 34 xg5 l:txb8 35 l::t x b8+ and su bsequently he convincingly con
..t?g7 with equal ity) 32 . . .fxe6 33 llxg7+ h8 verted it into a win .
(33 . . . f8 is worse because of 34 g5! ) 34 27 ... 11c5 28 e3 ltd8 29 d4 lla5 (29 . . . llc7
l:1xh7+ g8 - despite the enemy rooks' 30 ':'bb8) 30 1:bb8 d7 31 lla7+ c6 32
1 10 Thoughts about a Book

i.xa6 l:ta4+ 33 'Ot>e3 (33 'Ot>d3? i.xa6+ 34 29 l1xa6 l:!.fc8 30 1:tbb6 Itxc3 31 b5 i.xb5
1:txa6+ l:txa6 35 lhd8 l:txa2 ) 33 . . . d4+ 34 32 l::1 x e6+ 'Ot>f7 33 axb5 is qu ite probably not
cxd4 l::t a 3+ 35 'Ot>e4 i.xa6 36 lba6 ! l:txa6 lost.
37 l::t x d8 :xa2 38 l::t d 6+ 'Ot>b5 39 l:txe6 Zviagintsev suggested playing 23 .. .f6!? (in
tlxg2 40 'Ot>d5 Black resig ned . stead of 23 . . . l:te8). I will show some of the
Black lost without a fig ht. So what about our variations that we found together.
considerations regard i n g his hopes of
cou nterplay, and the 'unstrategic' nature of
Wh ite's decision - were these merely empty
words?
No, we based these on objective factors in
the position and therefore we have the right
to assume that Black could have defended
much more tenaciously. Here are some
considerations which will ease the search
for a plan of defence. Firstly, the loss of the
a6-pawn should not be fea red , especially if
at the same time the bishops are ex
changed . Secondly, it is important to pre
vent the wh ite king from making its way to
the centre.
A) 24 l:tbb8 l::t e 8 25 i.xa6 i.d7 26 l1xe8+
i.xe8, and if 27 l:tc8, then 27 . . . l:ta7.
B) 24 l:tb3 fxe5 25 fxe5 l::1 f8 ! , and the rook
restricts the mobil ity of the king , while also
creating the threat of 26 . . JH5 .
C) 24 i.xa6 l:te8 25 i.xc8 l:texc8 26 .uxc8
.ll x c8 27 l:tb7+ 'Ot>f8 28 exf6 gxf6 29 J:txh7
l:txc3 - in the rook endgame Black retains
real hopes of savi ng the game. The same
assessment appl ies to the position arising
after 25 l:tb3 fxe5 26 fxe5 i.xa6 27 l::t xa6
:ec8 (27 . . . <t>f7 ! ? ) 28 l:tbb6 l:txc3 29 J::1 x e6+
<t>f7 .
D) 24 i.xa6 l:te8 25 d3!? fxe5 26 fxe5
I n stead of the insipid 26 ... h6? Black should i.d7 (26 . . . l:txc3 27 ':a7+ is dangerous for
have tried 26 .. J:tf8 ! , intend ing 27 xh7 d4! Black) 27 l:txe8+ (27 l:ta3 l:tec8 28 J:tbb3
28 c4 ! (28 cxd4?? l::t c 1 mate) 28 . . . l:txc4 29 also comes into consideration) 27 . . . i.xe8
l:ta7+ <t>d8 30 h3 l:tc1 + ! ? 31 'Ot>h2 l:tc7 , and 28 l:tb3. Here Wh ite's advantage is sign ifi
of Wh ite's advantage only memories re cant, although the outcome still remains
mai n . If 27 h3 there is the satisfactory reply unclear.
27 . . . l:tf4 ! ? , and also the rook endgame Another possible approach to the defence
arising after 27 . . . h6 28 a4!? (28 xa6? (with which, to tell the truth, the analysis
i.xa6 29 l:txa6 l:tf5; 28 .ll b b8 l:te8) 28 . . . d7 should have beg u n ) i nvolves the captu re of
Thoughts about a Book ttJ 111

the c3-pawn in one version or another. Let Let us check 22 .. J:txc3 ! ? Now 23 i.. x a6
us retu rn to the position after Wh ite's 22nd suggests itself, consideri ng that after 23 . . .
move. <ltd7 24 1:[a8 :d8 2 5 l:Ibb8 the eternal pin
I n reply to 22 . . . 0-0 !? Sanakoev g ives the along the 8th rank ensures White a decisive
variation 23 1:[a8 i.. d 7 24 1:[bb8 1:[xb8 advantage (he brings his king up to the
(24 .. Jlcc8 25 l:txc8 i.. x c8 26 <ltf2 ! ) 25 centre and advances his passed a-pawn ).
.l:txb8+ 1:[c8 , and now not 26 l:t b7? i.. b5! 27 And the attempt by Black to disentangle
i.. x b5 axb5 28 1:[xb5 h5 (28 . . . g5!?) 29 1:[b3 h imself by 23 . . . <lte7 (with the idea of
l:tc4 30 g3 1:[a4 with cou nterplay, but simply 24 . . . .l:te8 and 25 . . . i.. d 7) ru ns into the tactical
26 l:txc8+! i.. xc8 27 <ltf2 , and the i nvasion of stroke poi nted out by Artur Yusupov: 24
the wh ite king decides the outcome. How l::t 1 b7+! ! , leading after 24 . . . i.. x b7 25 lixb7+
ever, Black can play 23 . . . .l:ixc3 !? 24 l:tbb8 and 26 l:tb8+ to the win of a piece. However,
nc6. Black is rescued by 23 . . . 0-0 ! , and if 24
i.. b5, then 24 . . . 1:a3, attacki ng the a2-pawn
and preparing to bring out the bishop to a6.
23 1:[a81 is stronger. The situations arising
after 23 . . . <ltd7 24 i.. x a6 .l:id8 25 .l:ibb8 and
23 . . . <lte7 24 .l:la7+! have already been
d iscussed above - they are defi nitely in
Wh ite's favour. The best defence is 23 ... 0-0 !
24 1:bb8 1:[c6. We have again reached the
position in the last diagra m . Evidently its
assessment also determines the objective
assessment of Wh ite's entire plan beg in
ning with 1 8 .l:.ab1 .
Let us sum up. The complicated (and,
probably, not faultless) analysis that we
have carried out once again illustrates the
How should this position be assessed? viabil ity of even the seemingly most d ifficult
Black has retained his extra pawn and no positions, but even so it does not cast
immed iate danger is apparent. But his doubts on the brill iant decision taken by
forces are tied down : it is not possible to Sanakoev on the 1 8th move . After a l l , the
disentangle hi mself by . . . lie8 (with the idea defence is very d ifficult, Wh ite everywhere
of . . . <ltf8 and . . . i.. d 7) because of the reply retains chances of success , and all the
i.xa6. He is forced to play . . . g7-g6 and same we have not found anyth ing more
. . . <ltg7, su bsequently restricting h imself to convincing for him .
waiting tactics. The q uestion (the reply to
which seems unclea r to me) is whether or
not Wh ite has sufficient resou rces to breach Conversion of an advantage
the opponent's defences. When examining the last two examples, we
In princi ple, after castl ing Wh ite is not have already begun discussing this topic,
obliged to sacrifice the c3-pawn - with 23 one that is very important for every player.
l1b3!? he retains the advantage. Therefore Just like another one, which is closely linked
it makes sense for Black to captu re the to it - the search for defensive resou rces in
pawn sl ig htly earlier. difficult positions.
1 12 Thoughts about a Book

Sanakoev - Engel 44 xd5 lIg4


1 0th World Championship 1 978-84 45 f3 ! lIxh4
46 g1 !
The point of Wh ite's plan becomes clear -
the rook has been trapped .
46 . . . ..i f5
After 46 . . . ':g4 47 ..i xg4 hxg4 48 b4 (48
lId4 ! ? ) 48 . . .f3 49 b5 ..i e4 50 b6 Wh ite must
wi n .
47 11d4
It is important to prevent 47 . . . ..ig4.
47 . . . e6?!
I n Sanakoev's opinion , the more accu rate
47 . . . ..id7! 48 b4 e6 also does not provide
any saving hopes in view of 49 l:td 1 ! (of
course, not 49 ..ie2? ..ic6 50 b5? ..ixg2 ! )
The author of the book writes: 4 9 . . . b 5 50 l:!. d 8 a n d 5 1 lI b 8 . However,
'The critical position, the play in which, after the waiting move 49 . . . e7! (instead of
strictly speaking, makes this game notewor 49 . . . b5) it would be far more d ifficult for
thy. By a clever regrouping Black has set his Wh ite to demonstrate the strength of his
opponent a specific problem - what to do position .
with his g- and h-pawns? For the moment 48 b4 e5
the black bishop may be deeply shut in, but 49 l:!.d5+ e6
does White 's passed pawn on the queenside
50 b5 iLg4
outweigh his material losses on the kingside ?
51 b6! ..i xf3
After al/, in the end Black's pieces may again
become active after the . . . f5-f4 advance. ' 52 gxf3
Fi rst let us see how the game concluded Black resigned i n view of 52 . . . xd5 53 b7
(relying on the assessments and certain l:[h3 54 b8'iV l:[xf3 55 'iVb7+, or 52 . . . lIh3 53
variations of Sanakoev). l:Id8! (but not 53 b7? ];[g3+ 54 h2(f2 ) lIg8
followed by . . . lIb8 and . . . d6-c6).
42 d5! f4+ 1
Zviagi ntsev rejected 42 d5 because of the
T h e exclamation marks are Sanakoev's. I
reply 42 . . . c5! . He had doubts about the
th ink that both should be replaced with
assessment of the bishop ending arising
question marks, or, at least, the '?!' symbol
after 43 d6+!? (we will trust the author of the
(dubious move).
book, who claims that 'after other conti nua
43 f2 tions Black is out of danger' , although 43
After 43 d4 l:tg4 ! 44 c5 cxd5 45 ..i xd5 l:Ia2 nevertheless deserves exami nation)
l:txh4 46 b4 l:t h 1 Black would have retained 43 . . .l:txd6 44 11xd6 xd6 45 ..ixf7.
saving chances.
43 . . . cxd5
(see diagram)
Hardly any better was 43 . . . c5 44 ..i d3 l:th6
45 d6+! d8 46 ..ixh7 with a big advantage
in the rook endgame.
Thoughts a bout a Book lb 1 13

involving a pawn sacrifice: 45 .. .f4+!? (in-


stead of 45 . . . c;i;>e5) 46 c;i;>xf4 c2 ! (much
worse is 46 . . . c;i;>e7? 47 xh5 c2 48 c;i;>e5!
xb3 49 i.g6) 47 xh5 (47 c;i;>g5 c;i;>e7)
47 . . . xb3 48 g6 d 1 . Wh ite is not able to
advance his knight's pawn : 49 g4 xg4 50
c;i;>xg4 c;i;>e7 leads to an i m med iate draw. He
should probably play 49 h5 c;i;>e7 50 c;i;>e5,
but I do not see how he can win after 50 . . . c4
5 1 h6 c;i;>f8 52 c;i;>d4 e2 53 e4 c;i;>f7 ! .
After rejecting 4 2 d 5 ? ! (as w e see, with
some justification), Zviagi ntsev chose an
other plan for converting the advantage,
based , however, on the same idea of
Sanakoev analysed the variations arising trapping the black rook as was carried out in
after 45 .. .'e5 46 g3 c;i;>f6 4 7 xh5 g8 48 the game by Sanakoev.
i.d 1 c;i;>e5 49 h5 and showed that Wh ite 42 c;i;>f2 1 .l:tg4 43 e2 ':xh4 44 f3 c;i;>d6 (it
wins. is a pity to give up the c6-pawn) 45 b4!
And yet Vadim's i ntuition d id not deceive (intend ing 46 lld 1 and 47 c;i;>g3 ; the immedi
him: Black can save h imself by playing ate 45 ':d1 is inaccu rate because of
45 . . . c;i;>e5 46 g3 f4+ ! ! 47 gxf4+ c;i;>f6 48 d5 45 . . . c5! ) 45 ... f4 (45 . . . .:h1 46 .l:tc2 l:[b1 47
(48 xh5 c2 ) 48 . . . c2 . Wh ite is not able l:txc6+ c;i;>d7 48 l:tb6).
to convert his material advantage. He is tied
down by the fact that his b-pawn is on a
square of the colour of his bishop. And if his
king goes to c3 , the opponent gains the
opportu n ity for a cou nterattack on the
opposite wing.
In a new edition of his book Sanakoev
disagreed with my opin ion , and suggested
the variation 49 c;i;>d2 g6 50 c;i;>c3 c;i;>f5 5 1
'it;>c4 ! . The resulting position is i ndeed won :
5 1 . . . c;i;>xf4 5 2 c;i;>xc5 c;i;>g3 53 b 4 d 3
(53 . . . e 8 5 4 c6) 5 4 c4 e 4 55 b 5
'it;>xh4 5 6 b 6 i.b7 (otherwise 57 d5) 5 7
i.d5 i. a 6 5 8 c;i;>c6 h 4 59 c4 .
Black's play c a n b e improved b y 49 . . . f5!?
50 c;i;>c3 g4 5 1 e4 (51 c;i;>c4 d 1 52 c;i;>xc5 I n itially the central breakth rough seemed
i.xb3 ! 53 xb3 c;i;>f5) 51 . . . e2 ( 5 1 . . . e6) to be correct: 46 d5?! cxd5 47 llxd5+ c;i;>c7
52 i. d3 d 1 (52 . . . g4) 53 c;i;>c4 c;i;>e6. It is 48 .l:tb5 (a drawn bishop ending arises after
even simpler to play 49 . . . b 1 ! 50 c;i;>c3 c;i;>f5 48 ':xh5 ':xh5 49 xh5 e4! 50 xf7
51 c;i;>c4 a2(c2 ) 52 c;i;>xc5 xb3 with a c;i;>b6) 48 . . . g6 49 c;i;>g 1 ! with good chances
draw. of success. Alas, a more carefu l verification
In the position from the last d iagram there is reveals that by playing 48 . . . d3! (instead
another interesting way of defending , also of 48 . . . g6) 49 l::t b 7+ c;i;>d6 50 %:txf7 c;i;>e6!
1 14 Thoughts about a Book

51 l:tf8 <:J;; e 7 52 lIc8 : h 1 (th reatening mate without it, Black cannot combat the passed
on f1 ), Black forces the exchange of bishops pawns.
and gains a d raw. I hope you have seen that Sanakoev's
Therefore White should shut in the rook interesting book offers us a mass of food for
immed iately: 46 <:J;; g 1 ! , and only after thought. I have dwelled only on a few
46 J.f5 (with the idea of . . . J.g4 or . . . J.d7)
... episodes (another example of the author's
reply 47 d5! c5 (47 . . . cxd5 48 l::t x d5+ <:J;; e 6 49 play is exami ned i n the chapter 'Vi rtuoso
b5 is hopeless) 48 b5! (but not 48 bxc5+ defence'), but, of course, there are many
<:J;; x c5 49 d6 J.d7 50 l:td5+ <:J;; b 6 with a more games in the book, and in each of
probable draw). It is here that the tragi them the reader will defin itely find some
comic position of the black rook is felt - thing interesting and usefu l .
ltJ 1 15

PART IV

Attac k

Artur Yusu pov

M issed B ri l l ia n cy P rizes

M write
ark Dvoretsky's suggestion that I should sol idly and boringly (here, unfortunately,
about some spoiled 'master they are more correct about the latter). I
pieces' came at just the right time. should like to try and change this image for
Firstly, I have long been wanting to make a the better: 'he plays badly, but interestingly' .
more careful analysis of certain old games. And , final ly, perhaps my dismal experience
With the passage of time, the vexation will prove useful to others , although , I have
caused by missed wins has now subsided , to admit, I have learned l ittle even from my
and perhaps I will be able to look at them own mistakes.
more objectively and critically.
Yusupov - Rebel 8
Secondly, I am i ndeed a leading expert in
th is field ( I have in mind not the critical 1 3th match game, I schia 1 997
examination of my own games, but the 'active chess' (30 mins. for the game)
spoiling of masterpieces). Although d u ring Queen 's Pawn Opening
my career I have managed to create several 1 tt'lf3 tt'lf6 2 d4 e6 3 e3 c5 4 .1i.d3 b6 5 b3
games of which even now, after the strict .1i.e7 6 i. b2 0-0 7 0-0 d5 8 tt'le5 tt'lfd7 9 f4
test of time and chess analysis, I can be tt'lxe5 1 0 dxe5 .1i.a6 1 1 c4 tt'lc6 1 2 a3 dxc4
proud, nevertheless for each such game 1 3 bxc4 f5 1 4 exf6 .1i.xf6 1 5 tt'lc3 tt'la5 1 6
there are a dozen others , which up to a point 'iVc2
were excellently played , but then hopelessly
spoiled .
Th irdly, I can imagine what a 'pleasure' I
have afforded my trainer and co-author (of
course, book co-author, not co-author in the
spoiling of masterpieces) in observing my
numerous lapses. Now I can at least explain
that I was col lecti ng material for a book.
In addition , there is a mercenary aim. I fear
that some tournament organisers have
developed (alas, not without certain grounds)
an unfavourable impression of my chess
style. They possibly th ink that I play too
1 16 M i ssed Brilliancy Prizes

My last (for the moment) example of a 21 .i.g6+ g8


spoiled attack occurred in an exh i bition
match against a computer program. There
follows a typical computer decision :
16 . . . lLlxc4
In the event of the 'human' 1 6 . . . h6 1 7 l:.ad 1
'ike7 1 8 lLlb5 ..txb2 1 9 'ikxb2 Wh ite would
have gai ned an obvious advantage , si nce
Black's castled position is weakened and his
knight at a5 is out of play. Now, however,
variations have to be calculated , which even
with a normal time control would have been
a d ifficult task for me.
17 ..txh7+ h8
1 8 ':ad1
This natu ral move cost me more than ten
I was intending to i nclude my queen in the
precious minutes and proved to be a
attack with 22 'ii'xd 1 ! and I was somewhat
sign ificant mistake. The immediate 1 8 ':f3!
d iscouraged on fi nding the defence 22 . . .
was better. I was wrong to be afraid of
.i.g5 ! . I saw that Wh ite's attacki ng re
1 8 . . . 'ikd2 , since then the simple 1 9 'ii'x d2
sou rces were not exhausted and that he
lLlxd2 20 l:[h3 is sufficient. In the event of
had the move 23 lLl d 5 ! , but after 23 . . . 'ikd8
1 8 . . . lLlxb2 1 9 l:th3 ..t h4 (or 1 9 . . . lLlc4 20
I overlooked the intermediate check 24
'ike2) 20 ..te4 ..td3 2 1 'ikxb2 ..txe4 22 lLlxe4
.i.h7+ ! , and if 24 . . . h8 (24 . . . f7 25 'ikh5+)
g8 23 'ii'e 5 Wh ite has a powerfu l attack.
- 25 fxg5 :
18 . . . 'ii'e 7?
1 ) 25 . . . 'ikxg5 26 ..t g 6 + g 8 (26 . . . 'ik h 6 27
1 8 . . . lLlxe3 ! was correct. After 1 9 l:[xd8 Black xh6+ gxh6 28 'ika 1 + ! ) 27 lLle7+! 'ikxe7 28
has the satisfactory reply 1 9 . . Jlaxd8 20 l:th8+ xh8 29 'ikh5+ g8 30 'ikh7 mate;
'ii'g 6 ..td3 21 'ii'x d3 ':xd3 22 ..txd3 ..td4 ! ? ,
2 ) 25 . . . 1:tf1 + 26 'ii'x f1 ..txf1 27 ..tf5+ g8 28
but 1 9 . . . lLlxc2! 20 :Xa8 l:txa8 2 1 .i.xc2 .i.d4+
..txe6+ f8 29 l:[h8 mate.
22 ':'f2 ':'d8! is even stronger. I overlooked
I don't know which of these mates is the
the last move of this variation , of course.
more pretty, but in any case it was a pity that
19 ':f3 lLlxb2
I was unable to find this worthy conclusion
20 l:[h3 lLlxd1 to the game. I nstead of this there followed :
If the computer sees someth ing it can 22 ..th7+ h8
captu re , it invariably does this. A human
22 ..tg6+ g8
player knows that it isn't good to be greedy,
Draw.
but the concrete confi rmation of this truth
l ies beyond the computer's calculating
horizo n . However, being in time-trouble, Yusu pov - Xie J u n
Wh ite was unable to fi nd a win and he Linares 1 997
satisfied h imself with a repetition of moves. Ruy Lopez
The correct way was found the following day 1 e4 e5 2 lLlf3 lLlc6 3 ..t b5 a6 4 ..txc6 dxc6
by some interested chess fans, who ana 5 0-0 'ikd6 6 lLla3 b5 7 c3 c5 8 lLlc2 f6?!
lysed Wh ite's attack. (8 . . . lLle7; 8 . . . ..tb7) 9 a4 ..tb7 1 0 axb5 ..txe4
M i ssed Brilliancy Prizes CD 117

1 1 d4 cxd4 the other hand, the black pawn centre


should also not be underestimated ! If Black
is al lowed to complete her development, it is
Wh ite who will be in trouble. I sensed that a
critical moment had been reached and I
sank i nto thought, looking for the best
solution.

The opponent is beh ind in development and


Wh ite has gained an opportun ity to attack. A
trifle such as the lack of a pawn should not
concern h i m , of cou rse.
12 ':e1 ! ..t b7
1 3 cxd4 axb5 Here I made an amusing mistake in my
1 4 1:txa8+ ..txa8 calcu lations. I very much wanted to estab
1 5 'ii'e 2 e4 lish my knight at eS, and so I comparatively
1 6 'ii'x b5+ ..tc6 qu ickly hit on the correct solution 22 liJxfS ! !
'ii'xfS 23 liJeS, which with great regret and
1 7 'ii'a 5
roughly the same speed I rejected in view of
White has already regai ned his pawn and is the simple 23 . . . liJxeS 24 'ii'x bS+ liJf3+. The
continuing to develop his i n itiative . 1 7 fact that i n this remarkable variation the
l1xe4+?! is weaker in view of 1 7 . . . 'iti>f7 . black king is already in check and hence
17 . . . liJe7 Black's last move is forbidden by the strict
1 8 liJd2 f5 rules of the game was someth ing that, of
1 9 b31 cou rse, I overlooked . The reason for such a
curious mistake was probably an inaccu rate
A good move, creati ng the threat of ..ta3
picture of the chess board during the
and preparing to transfer the knight via c4 to
calculation of variations, when a significant
the central square eS.
detail of the position simply did not register
19 . . . 'ii'd 7 in my mental visio n .
20 liJc41 liJg6 I think that t h e knight sacrifice would have
21 liJ2e3 ..t b5 led to a win. Thus if 23 . . . c6 there is a pretty
Against the optimistic 2 1 . . .f4 Wh ite had mate by 24 'ii' a 8+ 'iti>e7 2S ..tgS+! 'ii'x gS 26
prepared the strong reply 22 liJeS. The 'ii' b 7+ 'iti>e6 27 Wd7+. 23 . . . ..td7 is more
natu ral and forced move i n the game has tenacious, although in this case too after 24
led to a situation in which Black's lack of Was+ ..tc8 2S nxe4 liJe7 26 ..ta3 Wh ite has
development has become da ngerous. On a decisive attack.
1 18 M issed Brilliancy Prizes

22 h4! ? was promising , in order to provoke Wh ite's broken pawn structu re does not
the reply 22 . . . h5 (22 . . . xc4 23 lLlxc4 d6 leave h i m any real hopes of more than a
is more tenacious), and now the same sharing of the point, which withi n a short
sacrifice is very strong: 23 lLlxf5! 'ii'xf5 24 time d id in fact occu r.
lLle5 d7 25 'ii'a 8+, although here Black
may have an opportun ity to bring the rook Yus upov - Ivanehuk
into play via h6.
Tal Memorial Tou rnament, Riga 1 995
An alternative knight sacrifice was sug Queen 's Gambit Accepted
gested by Thomas Wed berg : 22 lLle5 ! . After 1 d4 dS 2 e4 dxe4 3 e3 lLlf6 4 xe4 e6 5
22 . . . lLlxe5 23 dxe5 c6 (23 . . . d3? 24 lLld5! ; lLlf3 eS 6 0-0 a6 7 b3 lLle6 8 We2 exd4 9
23 . . . c5 24 'ii' a 8+ 'ii'd 8 25 'ii' b 7 with a .l:r.d1 d3 (9 . . . e7) 1 0 .l:r.xd3 We7 1 1 lLle3
decisive advantage) 24 l:td 1 d3 25 a3 d6?1 ( 1 1 . . . c5) 12 e4 lLleS 1 3 lLlxe5
xa3 (25 .. .f4 26 lLlc4 'ii'g 4 27 ':'a 1 xc4 28 xeS
xf8 :txf8 29 'ii' c7 a6 30 e6! is bad for
Black, but 25 . . . 'ifi>f7!? deserves considera
tion ) 26 'ii'a 8+ 'ii'd 8 27 'ii' x a3 .%:tf8 28 'ii'c 5
Wh ite, according to his analysis, retains the
advantage.
I deviated from the correction conti nuation
of the attack, for the reason that I was
tempted by the strateg ically tempting under
m i n i ng of the centre.
22 g4?
This move looks stronger than it really is.
22 . . . fxg4
23 'it'a8+ 'ifi>f7
24 'it'xe4 b4!
Black uses the respite g ranted to complete In trying to gain control of the b8-h2
her development as q u ickly as possible. d iagona l , the opponent has rather fallen
White wins a pawn , it is true, but his i n itiative behind in development. Of course, Wh ite
completely evaporates. must immed iately try to seize the in itiative.
2S .l:r.d1 .l:r.e8 1 4 f4!
26 'it'xg4 'it'xg4+ The right way! By sacrificing a pawn , I
The correct assessment of the position . I n further i ncrease my lead in development.
the endgame Black's king will b e safe and The slow 14 g3 would have allowed Black to
the two strong bishops fully compensate for obta i n a n acceptable position after 1 4 . . . d7
the small material deficit. 1 5 f4 .i.xc3 1 6 l:1xc3 c6. For example, 1 7
27 lLlxg4 .l:r.e4 e5 lLld5 1 8 .i.xd5 exd5 1 9 e3 d4 20 .i.xd4
28 h3 hS 'it'd7 with cou nterplay.
29 lLlgeS+ lLlxes 14 . . . xf4
30 lLlxeS+ 'ifi>e6 1 S .i.xf4 'ii'xf4
Here we can take stock. Black has gained 1 6 eS! lLld7
sufficient compensation for the pawn , and The point of the pawn sacrifice is that the
M i ssed Brilliancy Prizes ltJ 1 19

active 1 6 . . . tl)g4? is met by the simple 1 7 g3, strongest players in the world, qu ickly
and if 1 7 . . . 'it'f5 , then either 18 i.. c2 or 1 8 making his repl ies with an imperturbable
1:[f1 , and Wh ite's attack develops u n h i n appearance!
dered . 21 . . . tl)e5
1 7 1:[f1 1 After 21 . . . .l:.d8 the suggestion by Ljubomir
Again Wh ite finds the most energetic solu Ftacn i k is possible: 22 tl)f6+! gxf6 (if
tio n . Of cou rse, he could have retained an 22 . . . h8, then 23 :h3 h6 24 'it'd2 most
attack without any additional sacrifices: 1 7 simply decides matters) 23 'it'g4+ f8
1:[e3 ! ? 0-0 1 8 1:[f1 'it'd4 1 9 h 1 , but in this (23 . . . h8 24 :g3) 24 'it'b4+, and if 24 . . . 'it'c5
case the active queen i n the centre of the (24 . . . e8 25 :g3 is bad for Black), then
board would have seriously h i ndered his simply 25 'it'xc5+ tl)xc5 26 ':'xd8+ e7 27
offensive. After the move in the game White l:th8, with a big advantage for Wh ite i n the
evicts the queen from the centre , and the endgame.
loss of the e5-pawn is compensated by h i m 22 l:th3 tl)g6
opening li nes and g a i n i n g t i m e for the
22 . . . h6 was worse in view of 23 ':'xh6 gxh6
attack.
24 tl)f6+, destroying the castled position .
17 1:[ad 1 would have been a fundamental
Black brings his knight closer to his king , but
mistake, allowi ng Black to pa rry the attack
Wh ite has already concentrated nearly all
at the cost of a small sacrifice: 1 7 . . . 0-0 ! 1 8
his forces for the attack.
l:txd7 i.. x d7 1 9 1:[xd7 l::t a d8.
17 . . . 'it'xe5
1 8 1:[e3 'it'd4
Of course, 1 8 . . . 'it'c5 was weaker because of
1 9 tl)e4 . Pinn ing the rook is Black's best
chance. He would have lost q u ickly after
1 8 . . . 'it'd6? (the reply to 1 8 . . . 'it'c7 would
have been the same) 1 9 l::t xf7 xf7 20 ':'xe6
'it'd4+ 2 1 h 1 tl)c5 22 'it'f3+.
19 :td1 'it'a7
1 9 . . . 'it'b6 was bad because of 20 i.. x e6 ! !
fxe6 2 1 tl) d 5 'it'c5 22 b4.
20 tl)e4 0-0
21 h1
White does everyth ing correctly, but he
23 'ii' h 5
spends too much effort and time. The only
reason I did not manage to bring the game Short of time for the calculation of varia
to a logical concl usion was that I d id not tions, White tries to play rational ly. It was
trust my assessment and I tried to calculate already possible to launch a decisive attack
the variations al most to the end. The result with 23 l:txh7!? The immed iate acceptance
was that at the critical moment I simply did of the sacrifice loses, accord ing to analysis
not have enough time for thought. I should by Sergey Dol matov:
have had more faith in my powers, but try 23 . . . xh7 24 'ii' h 5+ g8 25 tl)g5 1:[e8 26
retaining your confidence and composure , 1:[f1
when opposite you is sitting o n e o f the A} 26 . . . tl)e5 27 :txf7 tl)xf7 28 'it'xf7+ h8 29
1 20 M i ssed Bri l l i ancy Prizes

'ife8 mate; 2 9 . . . xg8 3 0 'ili'h7+ f7 3 1 .i.b3+ f6 32


B) 26 . . . ':e7 27 'ii' h 7+ f8 28 'ili'xg6 b5 29 .:tf1 + ;
'ii' h 7 and wins (if 29. Ae8 there is the reply C 2 ) 26 . . . e 4 27 xe4 ( 2 7 ':xg6? fxg6 28
30 ':d 1 ); 'ifxg6 f5 ! ) 27 . . . lDf4 28 'irxf7! g4 29
C) 26 . . . b5 27 ':xf7 (27 'ili'h7+ f8 28 'ili'xg6 'ii'e 7;
is also strong) 27 . . . 'ii'xf7 28 lDxf7 xf7 29 C3) 26 . . . lDf4 27 'ili'xe5 (27 l:.xg7! xg7 28
'ii'f3+ lDf4 30 'ii'x a8. 'ili'xe5 is more qu ickly decisive) 27 . . . lDg6 28
During the game I was concerned that Black xg6 fxg6 29 l:txg6 b5 30 l::t x h6+ gxh6 31
had another defensive possibil ity. But here 1:[d7;
too a way to win can be fou n d : C4) 26 . . . 'iff2 27 l::t x g6 'iVf4 (27 .. .fxg6 28
23 . . . lDf4 2 4 'ii'g 4 'it>xh7 25 lDg5+ 'ii'x g6 'ii'x c2 29 'ii'xc2 gxf6 30 'ii'g 6) 28 lDd5
A) 25 . . . h6 26 'ili'xf4 f6 (26 . . . g6 27 'ili'g3) fxg6 29 lDxf4 gxh5 30 lDg6+ g8 31 .i.b3+.
27 'ii' h 4+ g6 28 'ii' h 7+ xg5 29 'ili'xg7+ In all the variations Black is unable to avoid
h5 30 l1d5+! with mate; a qu ick mate or heavy loss of material ,
B) 25 . . . g8 26 'ili'h4 l:[e8 27 'ifh7+ f8 28 whereas Wh ite's prematu re combination
'ii' h 8+ e7 29 'ii'x g7 J:r.f8 30 lDh7 .i.d7 3 1 allowed Vasily Ivanchuk to retain the bal
'ii'f6+ e8 3 2 'ii'xf4 with a winning position. ance.
However, the natural attacki ng move in the 24 . . . gxf6
game is an equally good alternative. 24 . . . h8? would have lost to 25 'iVg5! e5 26
23 . . . h6 l:txh6+ gxh6 27 'ii'x h6 mate .
At this point I had only five min utes left on 25 'ii'x h6 ':e8
my clock. I saw that my plan ned combi na 26 1:[g3
tion would guarantee Wh ite perpetual check, Now if 26 .i.c2 Black has the reply 26 .. .f5,
and I was hoping for someth ing else to tu rn lead ing to perpetual check after 27 'iVh7+
up. f8 28 'ii' h 6+.
24 lDf6+? 26 . . . 'ii'f2 !
To Wh ite's great an noya nce, immed iately By the irony of fate, it is the inclusion of the
after the game he easily d iscovered a queen that saves the game. (Remember
decisive strengthening of the attack. He how much effort Wh ite made to shut this
should have included the bishop i n the important piece out of the game . ) Attempts
offensive by 24 c2 ! , when the opponent by Black to play for a win, taking account of
would have had no defence against the the opponent's time-trouble, would most
threat of 25 lDf6+. He would have faced a probably have boomeranged . 26 . . . b6 27
d ismal choice fro m : l:td4! 'irc7 28 :g4 is completely bad for him.
A) 2 4 . . . b 6 25 lDf6+ gxf6 ( 2 5 . . . h8 26 'iVg5) If 26 . . . b5, then after 27 l::t f 1 ! 'iVd4 28 l::t x g6+
26 'ili'xh6 1:[e8 27 xg6; fxg6 29 'ii'x g6+ f8 30 'ifh6+ if he wishes
B ) 24 . . . lDf4 25 lDf6+ h8 26 'iVxf7! d7 27 Wh ite can force perpetual check, since
l:txh6+; 30 . . . e7 is unfavourable because of 3 1
C ) 24 . . . e5 25 lDf6+ h8 (25 . . . gxf6 26 l::t g 3 'ii'g 7+ d 6 3 2 l::t d 1 .
f5 27 lbg6+ fxg6 28 'iVxg6+ h8 29 'ii'x h6+ Now, however, Black has parried the th reat
g8 30 .i.b3+) 26 l::t g 3, and now: of h2-h4, and Wh ite has noth ing better than
C 1 ) 26 . . . e6 27 ':xg6 fxg6?! (27 . . . 'ii'e 3 28 to reconcile himself to a draw.
lDg4) 28 'irxg6 g8 29 lDxg8 (29 :d7! ) 27 l::t x g6+ fxg6
M i ssed Brilliancy Prizes ctJ 121

28 "xg6+ f8 The accepta nce of the sacrifice would have


29 "h6+ f7 lost: 20 .. .fxg6 21 "xe6+ f8 22 f5. How
30 "h7+ f8 ever, 20 . . . O-O !? 2 1 d3 a6 was more
circu mspect, since the obvious 22 f5? exf5
31 "h8+
(but not 22 . . . axb5? 23 f6 or 23 fxe6) 23 .l:txf5
Draw. (hoping for 23 . . Jld5? 24 .l:r.xf7 ! ! ) is refuted
Yusu pov - H u bner by 23 . . . liJxb2 ! .

lilburg 1 987 21 xf7


Slav Defence Wh ite can no longer stop halfway (2 1 d3
a6 22 f5 axb5 23 fxe6 f6 24 Wg3?! Wxd2 will
1 d4 dS 2 liJf3 liJf6 3 c4 dxc4 4 liJc3 c6 S a4
fS 6 e3 e6 7 xc4 b4 8 0-0 liJ bd7 9 not do). Both players become carried away
liJh4 g6 1 0 liJxg6 hxg6 1 1 f4 ( 1 1 h3) by the wild compl ications, with not the
1 1 . . . liJdS ( 1 1 . . ...a5; 11 ... 0-0) 12 d2 ( 1 2 slig htest impression of where they will be
liJe4 ! ? "e7 1 3 liJg5) 1 2 . . ...a S 1 3 "e1 able to escape from them.
liJSb6 14 b3 cS 1 S I1d1 cxd4 16 liJ bS 21 . . . xf7
xd2 1 7 ':'xd2 liJcs 1 8 c2 l:.d8 1 9 The consequences of 2 1 . . . Wxb5 22 Wxe6+
exd4? ! ( 1 9 1bd4 "xe 1 2 0 ':'xe 1 i s sounder, f8 23 g6 Wd7 24 lle2 "xe6 25 ':'xe6
and if 20 . . J:1xd4 2 1 exd4 liJcxa4 , then 22 liJd5 26 ':'fe 1 liJf6 were unclear. I think that
d5). after 27 g3 or 27 :e7 Wh ite has sufficient
compensation for the sacrificed piece.
22 fS eS
Black would have lost after both 22 . . . llhe8
23 fxe6+ g8 24 e7, and 22 . . ...xb5 23
"xe6+ f8 24 f6 Wd7 (or 24 . . ...xf1 + 25
xf1 ':'e8 26 "d6+ f7 27 llf2 ) in view of
25 fxg7+ xg7 26 Wf6+ g8 27 _g6+.

Wh ite has somewhat overestimated his


chances, and now the cool-headed 1 9 . . .
O-O ! could have set h i m serious problems.
Instead of this Robert H u bner falls in with
his opponent's idea .
19 . . . liJcxa4
20 xg6
A practically forced move , leading to an 23 f6
abrupt sharpening of the play. The immed iate 23 Wxe5 was a serious
20 . . . e7 alternative. The following variations do not
1 22 M issed Brilliancy Prizes

exhaust all the possibil ities in the position , less finds new ways to strengthen the
o f course, but they show how strong Wh ite's attack. The threat is 27 1:1e7+ , for example:
attack is: 26 . . . 'ii'x bS 27 1:1e7+ 'iti>f8 28 'ii'x g6 'ii'x f1 + 29
A) 23 .. Jlhe8 24 tDd6+ ; 'iti>xf1 tDxe7 30 'ii'g 7+ with mate.
B) 23 . . . 'ii'x d2 24 f6! (24 'ii'e 6+ 'iti>f8 2S f6 26 . . . lId7
l:r.d7! ) 27 1:1e7+ tDxe7
B 1 ) 2 4 . . . gS 2S tDd6+ ! .l:.xd6 (2S . . . 'iti>g6 26 27 . . . ':xe7 28 fxe7+ (28 'ii'x dS+ ! ) 28 . . . tDf6 is
'iVe4+ 'iti>h6 27 tDfS+ 'iti>g6 28 f7) 26 'ii'e 7+ hopeless in view of 29 'ii'e S with the decisive
'iti>g6 27 'iVg7+ 'iti>hS 28 'ii'x h8+; threat 30 tDd6+.
B2) 24 . . . 'iti>g8 2S fxg7 ':h6 (2S .. Jbh2 26 28 fxe7+ 'iti>e8
'ii'e 6+ 'iti>h7 27 g8'ii' + ':'xg8 28 l:tf7+ :g7 29 29 'ii'e 5
'ii'fS+ 'iti>h8 30 .:tf8+ ':'g8 31 'ii'f6+ 'iti>h7 32 29 l:tf6 was probably even stronger.
l:tf7+ ) 26 'ii'e 7, and Wh ite wins;
29 . . . 1:1xe7
C ) 23 . . . tDdS 24 'ii'e 6+ 'iti>f8 2S f6 g6 26 :e2
30 'ii'x h8+
C 1 ) 26 . . . 'ii' b 6 27 'ii'g 4 'iti>f7 28 'ii' g S!?
(intending 29 1:[e7+ ), or immed iately 28
1:1e7+! tDxe7 29 fxe7+ 'iti>xe7 30 'ii'g S+ 'iti>d7
31 l:If7+ 'iti>c8 (31 . . . 'iti>c6 32 .l:.f6+) 32 'ii'e S!
with the decisive th reats 33 tDxa7+ and 33
1:1c7+;
C2) 26 . . . 'ii' b4 27 f7 'iVe7 28 'ii'g 4 'ii' h 4
(28 . . . tDe3 29 'iVxg6 or 29 'it'f4) 29 l1e8+!
1:1xe8 30 fxe8'if+ 'iti>xe8 31 tDd6+ 'iti>d8 32
'it'c8+ 'iti>e7 33 ':'f7+ 'iti>xd6 34 'ii'd 7 mate ;
C3) 26 . . . 'ii'x bS 27 f7 'ii'x e2 (27 . . . 'ii'd 7 28
'ii'e S! 1:1h7 29 'ii'e 8+) 28 'ii' x e2 tDab6 29
'ii'e S with advantage to Wh ite .
The conti nuation in the game is probably
just as good and in many cases it leads to a
simple transposition of moves. After making this move , Wh ite, who was in
23 . . . 96 moderate time-trouble, timidly offered a
If 23 . . . gS there follows 24 'ii'x eS with a draw, wh ich my opponent sensibly ac
strong attack. cepted . G reat was my astonishment, when
24 'ii'x e5 i n subsequent analysis I discovered that in
the concluding position I was a pawn up! I
24 dxeS!? 'ii'x d2 2S e6+ 'iti>f8 ! 26 e7+ 'iti>f7 27
had been material down for so many moves
exd8tD+ 'ii' x d8 28 'ii' b4! was also interest
and was so happy to regain it, that I did not
i n g , with the th reats of 29 tDd6+ or 29 b3.
even notice that I was now ahead ! Of
24 . . . tDd5
course, the sou nd extra pawn determines
25 1:1e2 'ii' b 6 the eval uation of the position, and after the
If 2S . . . 'ii'x bS 26 'ii'e 6+ 'iti>f8 27 f7 'ii'x e2 28 natural 30 . . . 'iti>d7 31 'ii' h 3+ 'ii'e 6 32 'ii'x e6+
'ii'x e2 with advantage to White. l:t.xe6 33 b3 even my tech nique should have
26 'ii' 9 51 sufficed for a w i n .
Although in time-trouble, White neverthe-
M i ssed Brilliancy Prizes ttJ 1 23

Yusu pov - Anand strai ned 1 6 . . . .:tcB ! ? was less in keeping with
Linares 1 99 1 the temperament of my opponent, who very
Queen 's Pawn Opening rarely avoids compl ications.
1 d4 ttJf6 2 ttJf3 e6 3 e3 b6 4 ..i d3 ..i b7 17 ttJxh71 ttJxh7
5 0-0 d5 6 ttJe5 ttJbd7 7 f4 g6 8 b3 ..ig7 9 1 8 ttJxg6
ttJd2 e5 1 0 ..i b2 0-0 1 1 'Wf3 1 ttJe8 1 2 'Wh3 1 B ..i xg6 was weaker because of the simple
ttJd6 13 ttJdf3 .:te8 1 B . . . ttJfB .
18 . . . 'We7
1 B . . . .:tcB 1 9 .:tf3 c4 was risky in view of 20
ttJe7+ 'Wxe7 2 1 'iVxh7+ 'it?fB 22 ..ig6 :edB
23 .:tg3 or 23 ..ia3 with a strong attack.
However, possibly Black should have de
cided on 1 B .. .fS!? 1 9 ..ixg7 'it?xg7. Then 20
I1f3 ttJf6 21 I1g3 leads to a repetition of
moves: 21 . . . ttJfe4 (2 1 . . . ttJg4? 22 ttJeS! and
wins) 22 ..ixe4 ttJxe4 23 ttJeS+ ttJxg3 24
'Wxg3+ 'it?fB 2S ttJg6+, while 20 'Wg3 'Wf6 2 1
ttJeS+ 'it? h B 2 2 ttJd7 'Wc3 2 3 ttJeS leads to an
unclear position .
1 9 .1:.f3 ttJe4
1 9 . . . c4 was dangerous in view of 20 ttJe7+
Both players have practically completed the .:txe7 21 ..ixh7+ (2 1 'Wxh7+ 'it?fB 22 ..ig6 is
mobil isation of their forces and Wh ite weaker) 21 . . . 'it?fB 22 :g3 c3 23 ..i a3 fS 24
switches to determined actio n . However, :g6 with an attack.
Black too has prepared well for the oppo 20 ..i xe4 dxe4
nent's attack, by erecting powerful defen 21 :g3
sive l i nes. Possibly I should have preferred The rook joins the offensive against the
the restra ined 1 4 l1ad 1 ! ? , but I was already weakened position of the black king. Natu
seized by a creative mood . rally, the opponent tries to create counterplay
1 4 ttJg5 ttJf8 along the now open d-file.
1 5 dxe5 21 . . . l::t a d8
A standard exchange, opening the long 22 ':xd8 'Wxd8
diagonal for the bishop. 23 'Wg4
15 . . . bxe5 23 'WhS? is wrong , since after 23 . . . 'WdS
1 6 :ad1 ! (23 . . . 'Wd2 !? 24 h3 'ii'xc2) White has to reply
By including the rook in the game and 24 ttJeS, and 24 . . . l:te7 Ieaves Black with too
offering a piece sacrifice , Wh ite g reatly many defensive resources.
sharpens the position. It was not possible to 23 . . . 'Wd5
calculate all the variations, but it seemed to Here Black also had other possibil ities. I n
me that a couple of pawns and the in itiative the event o f 2 3 . . . 'Wd2 White would have
would provide sufficient compensation . played 24 h4 and if 24 .. :ii'xc2 , then 2S ttJfB
16 . . . f6 %:te7 26 ttJxh7 'ii'x b2 27 'Wg6 1:[f7 2B hS ..i dS
Anand accepts the challenge. The re- 29 h6 with a decisive advantage, while after
1 24 M issed Brilliancy Prizes

24 . . .f5 there would have followed 25 'it'h5 to demonstrate the correctness of his
'ife 1 + 26 'ifi>h2 'ifxg3+ 27 'ifi>xg3 i.xb2 28 attack.
tDe5 11e7 29 'it'd 1 with the better game. However, he has available another, stronger
23 .. .f5 came into consideration. After 24 conti nuatio n , which occu rred to me only
'ifh5 tDf6 25 tDe7+ l:be7 (25 . . . 'it'xe7 is after the game. Wh ite should pursue the
weaker in view of 26 i.xf6) 26 i.xf6 'it'd2 27 knig ht: 25 tD h 7 ! . As shown by the variations
h3 (27 ':'xg7+ ':'xg7 28 'it'e8+ leads to given below, Black now has to solve some
perpetual check) 27 . . . 'it'e 1 + 28 'ifi>h2 'it'xg3+ d ifficult problems:
29 'ifi>xg3 i.xf6 30 'ifi>f2 Wh ite, in my view, A) 25 . . . ':'d8 26 tDxf6+ i.xf6 27 i.xf6 'it'd 1 +
has somewhat the better chances. 28 'it'xd 1 ':'xd 1 + 29 'ifi>f2 - the piece is
regained and White should win;
B) 25 ... 'iVd2 26 h4 'iVe 1 + 27 'ifi>h2 'ii'x g3+ 28
'iVxg3 tDxh7 29 h5 with a winning position;
C) 25 . . .f5 26 tDf6+! (26 'it'e2 i.xb2 27
':'xg5+ 'ifi>f7 28 c4 'it'd3 29 c4 'it'd3 30 'it'xb2
'it'xe3+ is unfavourable for Wh ite, but he
ca n consider 26 'iVh5 tDf3+ 27 'iVxf3 ! exf3
28 ':'xg7+ 'ifi>h8 29 :d7+ e5 30 ':xd5 i.xd5
31 tDf6 ':'d8 32 i.xe5 with a favourable
endgame) 26 . . . i.xf6 27 'it'h5
C 1 ) 27 . . . i.xb2 28 'it'xe8+ 'ifi>h 7 29 h4!
with a big advantage (less is promised by 29
'iVh5+ 'ifi>g8 30 l::t x g5+ 'ifi>f8 31 h3);
C2) 27 .. J:td8 28 i.xf6 (weaker is 28
24 h4 'iVg6+ 'ifi>f8 29 'iVxf6+ 'ifi>e8 30 ':'xg5 'iVd 1 +
3 1 'ifi>f2 l::t d 2+! 32 'ifi>g3 ':'xg2+! 33 'ifi>xg2
The most natu ral development of the game.
'it'f3+ with perpetual check) 28 . . . 'iVd 1 + 29
Wh ite makes an escape square for his king
'it'xd 1 ':'xd 1 + 30 'ifi>f2 ':'d2+ 3 1 'ifi>e 1 l:txc2 32
and includes his rook's pawn in the offen
fxg5, and the endgame is most probably
sive. But at the same time he had a more
won .
camouflaged way of conducti ng the attack. I
rejected 24 tDf8 ! ? in view of 24 . . . tDg5 24 . . . 'iVf5
(24 . . . .:.e7 is weaker because of 25 tDxh7 If 24 . . .f5 , then 25 'it'h5.
'ifi>xh7 26 i.xf6). I ndeed , now noth ing is 25 'it'd1 'iVd5
promised by 25 fxg5 f5 26 'it'h5, since Black 26 'it'94 'it'f5
repl ies not 26 . . . ':'xf8 (in view of 27 g6 i.h6
It appears that things will end i n a repetition
28 i.d4! cxd4 29 'it'xh6 'ifd7 30 g7 and
of moves, especially since I was already in
wins), but either 26 . . . 'ifi>xf8 , or 26 . . . i.xb2. I n
my customary time-trouble.
the fi rst case 2 6 . . . 'ifi>xf8 2 7 i.xg7+ 'ifi>xg7 2 8
'ifh6+ 'ifi>f7 2 9 'iff6+ leads to perpetual 27 'ifd1 'it'd5
check, but it is possible to play for a win by 28 'ife2 1
29 g6+ 'ifi>e7 30 h3. More i nteresting is After plucking up cou rage , Wh ite decides to
26 . . . i.xb2! 27 tDh7 'ifi>g7 28 'ifh6+ (or 28 c3 play on. Now it is not easy for the opponent
i.xc3 29 tDf6 .:th8 30 tDxd5 l:Ixh5 31 tDxc3 to find a useful move. Thus if 28 . . . 'it'd6 there
i.c6) 28 . . . 'ifi>f7, when it is not easy for Wh ite follows 29 'it'g4 ':'d8 30 'ifi>h2 ':'d7 3 1 tDe5
M issed Brilliancy Prizes It:J 1 25

(31 h5 f5) 3 1 . . . :e7 (bad is 3 1 . . . fxe5 32 32 'ile2 ..txb2


..txe5 'ile7 33 ..txg7) 32 h5 'ili'a6 33 h6 with 32 . . . 'iid 6 33 ..txg7 Wxg3 34 ..tf6 was no
a powerful attack. wh ile if 28 . . . ..tc6. then 29 better.
h5 f7 30 'ilg4 :g8 31 h6. The continuation
33 e4 Wd6
in the game hardly makes a sign ificant
change to the position. 34 'ilxb2 e51

28 . . . ..te8 underestimated this move . Of cou rse.


34 . . . 'ii'x g3? would have lost immed iately to
29 h5
35 'ii'f6+ 'itt g 8 36 'ii' h 8+ 'itt f7 37 'ii' h 7 mate.
This pawn acts as a battering-ra m . breaking
But now my pieces lose coord i nation . wh ich
up the nearly coordi nated black pieces.
not only makes it more d ifficult to convert
29 . . . f7 the material advantage. but also hands the
30 'ilg4 in itiative to my opponent. Discouraged by
Another way of concluding the attack was this tu rn of events. Wh ite loses the thread of
suggested by Dvoretsky: 30 h6! ..txh6. and the game.
now either 3 1 liJe5+ fxe5 32 'ilh5+. or 3 1 c4 35 11h3?1
'ild3 32 Wh5.
An unsuccessful manoeuvre . 35 h2 was
30 . . . liJg5! better. in order to answer 35 .. .f4 with 36
exf4 exf4 37 'iif2 e3 38 Wxf4+ 'ilxf4 39
liJxf4 e2 40 liJxe2 ':xe2 41 g6+ g7 42
:g5. retaining good winning chances. It is
possible that 35 'ii c 1 I ? preventing the
. . . f5-f4 breakth roug h . is even stronger.
35 . . . f4
36 l:t h4
The consequences of the exchange sacri
fice were unclear: 36 exf4 ..txh3 37 liJxe5+
g8 38 gxh3 Wd 1 + 39 f2 :d8 40 g3
%:td2 4 1 'iia 3.
36 . . . fxe3

Vishwanathan Anand defends very resou rce


fully. The counter-sacrifice of a piece is his
best practical chance . Unfortu nately. fatigue
and shortage of time were already begin
ning to affect Wh ite's play: instead of
looking for the strongest conti nuation he
satisfied h imself with 'the bird in the hand'.
3 1 liJe5+! fxe5 32 fxg5 was correct. The
strong connected passed pawns qu ickly
decide the outcome. for example: 32 .. J:td8
33 h6 'iid 1 + 34 'iix d 1 ':xd 1 + 35 h2.
31 fxg5?1 f5
1 26 <t> M issed Brilliancy Prizes

37 'it>h2? 9 ttJe5 .l::t e 8


In time-trouble White conclusively loses his
bearings and makes a blu nder. 37 'iVe2 !
was correct. I n this case he would at least
not have been in danger of losi ng, as the
following variations show:
37 . . . 'iVd2 38 'it'f1 + ;
3 7 . . . 'iVd4 38 h6! i.f5 3 9 h7 i.xg6 40 h8'iV
l:Ixh8 4 1 .l::t x h8;
37 . . . 'iVd3! 38 'iVf1 + ! 'it>g8 39 'iVf6 (39 h6
'iVxf1 + 40 'it>xf1 'it>h7 41 ttJxe5! l:txe5 42 g6+
'it>h8! leads to an unclear endgame) 39 . . . e2
40 'iVh8+ .
37 . . . i.f5!
38 'iVe2
In ti me-trouble it was completely impossible This game was played in the last round of a
to find the last chance , later pointed out by tou rnament i n which I performed very badly
Dvoretsky: 38 b4! ! cxb4 (38 . . . 'iVd2?! 39 'iVa3 and was a complete outsider. Therefore
with the dangerous th reat of 40 'ii'x a7+) 39 even if only at the finish I wanted to score a
c5! 'ii'x c5 (39 .. .'iVd4 40 'ii b 3+) 40 .l::t f4! ! exf4 first win . I cannot say that I was prepared for
4 1 'iVf6+ with perpetual check. the variation chosen by my opponent, but
38 . . . 'ii'd 2 the character of the play was more or less
39 'iif1 'it>e6 familiar to me: after all, the N i mzo-I ndian
40 .:txe4 i.xe4 Defence is firmly established in my open ing
repertoire. Black's somewhat mysterious
Wh ite resigned .
9th move is q uite simply explained : he
Although in this game Wh ite did not gain wants to retain the option of playing . . . i.c8-
any reward for his boldness, I was not a6 without loss of time. (After the i m med iate
seriously upset to have lost half a point by 9 . . . i.a6 Wh ite has the un pleasant 1 0 ttJc6!).
avoiding a d raw. My annoyance would have But now, without particular effort, I was able
been far g reater if, after agreeing to a to find an idea which, although not original,
repetition of moves, I had then d iscovered a was q uite sensible, and, as it later tran
win. From my experience I can assu re the spired , was also a novelty.
reader that playing for a win in such 1 0 ttJe2 !
situations more often brings success than
After 1 0 i.d2 Black is able to carry out his
d isappoi ntment, and, in any case , more
idea : 1 0 . . . i.a6, when 1 1 ttJc6? ttJxc6 1 2
creative satisfaction that the prematu re
i.xa6 i s now bad in view of 1 2 . . . cxd4 1 3
termination of the fight.
i.b 7 dxc3 1 4 bxc3 ttJa5! 1 5 i.xa8 i.c5. This
variation , l i ke a nu mber of others , is taken
Yusupov - Gulko from Boris Gulko's comments in Informator
Novgorod 1 995 NO.63. But 1 1 i.xa6 ttJxa6 1 2 'it'a4 'iVc8 1 3
Nimzo-Indian Defence .l::t a c1 'iVb7 1 4 'iVc6 .l:tab8 leads to equality
1 d4 ttJf6 2 c4 e6 3 ttJc3 i.. b4 4 e3 0-0 (Portisch-Spassky, Candidates match , Ge
5 i.d3 d5 6 ttJf3 c5 7 0-0 b6 8 cxd5 exd5 neva 1 977).
M i ssed Brillia ncy Prizes ttJ 1 27

The point of the move in the game is clear:


the knight is switched to g3, where it not
only controls e4 , an especially important
square in this variation, but is also ready to
join the attack on the opponent's king via f5 .
The slight loss of time is fully compensated
by the fact that Black, in view of the
dangerous position of his dark-square
bishop, can not maintain the tension in the
centre and is forced either to exchange on
d4 , thereby releasing the opponent's dark
sq uare bishop, or rel ieve the pressure in the
centre by advancing his c-pawn .
10 . . . c4
In the event of 1 0 . . . cxd4 1 1 exd4 .i.a6 1 2 1 6 a41?
.i.xa6 liJxa6 Black did not l i ke 1 3 .i.g5. An i nteresting and somewhat unexpected
decision . Wh ite is looking for more than the
11 .i.c2 .i.d6?1
standard development of the attack by 1 6
A loss of time, wh ich aggravates Black's
liJf5 g6 1 7 'ii' h 3 gxf5 1 8 .i.xf5 . By weakening
difficulties . After the natu ral 1 1 . . . .i.b7 G u l ko
the opponent's q ueenside he introduces
was afraid of 1 2 b3!? Even so, this was the
new motifs into the play.
lesser evil and Black could have defended
16 . . . b41 ?
with 1 2 . . . cxb3 1 3 .i.xb3 .i.d6.
The critical conti nuatio n . After 1 6 . . . a 6 1 7
12 f41
liJf5 g6 1 8 Wh3 gxf5 1 9 .i.xf5 liJb6 20 ':f3
A standard idea . Such a set-up is good if, as Wh ite has a strong attack.
in the game, Wh ite is able to control the e4-
17 a5
square . It was illog ical to begin play on the
queenside: 1 2 b3 b5 1 3 a4 cxb3 1 4 .i.xb3 b4 Of course, not 1 7 liJxc4? ':c8 1 8 b3 .i.a8,
with chances for both sides. and White loses a piece.
17 . . . ':c8
12 . . . b5
Possibly 1 7 . . . c3 should have been tried .
1 3 liJg3 liJbd7?1
Apparently Black d id not like 1 8 a6 .i.xa6 1 9
Black allows the opponent additional possi ':xa6 cxd2 20 'ii'e 2 Wc8 2 1 .i.f5 'ii' b 7 22
bil ities. 1 3 . . . .i.b7 was more accu rate . ':fa 1 , but 1 8 . . . cxb2 1 9 ':a2 .i.c8 came into
1 4 Wf3 consideration .
I decided not to deviate from the basic pla n . 1 8 a6 .i.a8
T h e alternative 1 4 e4 !? would have led to 1 9 liJf5 c3 1 ?
unclear play after 1 4 . . . liJb6 1 5 liJc6!? ( 1 5
1 9 . . . liJe4? is incorrect: 20 .i.xe4 dxe4 2 1
liJxf7 'it>xf7 1 6 e5 .i.g4 1 7 Wd2 'it>g8 is
'ii'g 4, a n d 2 1 . . . g6? is not possible because
unconvincing) 1 5 . . .Wc7 1 6 e5 .i.g4 (Gulko
of 22 liJxd7. And i n the event of 1 9 . . . g6,
gives 1 6 . . . liJfd7 1 7 exd6 Wxc6 1 8 liJf5 'it>h8
thanks to his provocation on the queenside,
19 Wg4 g6 20 liJh6) 1 7 Wd2 'ii' xc6 1 8 exf6.
Wh ite has acquired a new motif: 20 .i.xb4!?
14 . . . .i.b7 gxf5 (20 . . . .i.xb4 2 1 liJh6+ 'it>g7 22 liJhxf7
1 5 .i.d2 .i.f8?1 We7 23 'ii' h 3) 2 1 Wg3+ .i.g7 22 .i.xf5 .
1 5 . . . liJf8!? was more logica l . 20 bxc3 g6
1 28 M issed Brillia ncy Prizes

21 'ifh31 b3! g reat. Another weakness of the author


Accepting the knight sacrifice looks terribly made itself felt: as soon as I make a couple
dangerous. I n the event of 21 . . . gxf5 22 xf5 of attractive moves, I feel the desire to
lLlb6 23 l:tf3 .i.g7 24 :g3 c;i;>f8 25 cxb4 (or 25 create a 'masterpiece ' . Alas, excessive
c4!?) Wh ite already has three pawns for the emotions during play have damaged me on
piece with a dangerous in itiative. Gulko tries many occasions.
to gain at least some counterplay. 25 'ii'g 8+?!
22 .i.xb3 lLle4 I n a joint analysis after the game we
At the cost of two pawns Black has establ ished that Wh ite should have played
managed to establish his knight in the 25 'ifxg6+ 'iff6 26 lLlg7+! (26 'ifg8+ 'iff7 27
centre. Although objectively his cou nterplay 'ii'xf7+ c;i;>xf7 28 .i.e1 lLlxc3 is unclear). If
is insufficient, he has set his opponent some now 26 . . . xg7, then 27 'ifxe4+ c;i;>d6
serious practical problems, by sharply chang (27 . . . c;i;>f7 28 xd5+) 28 'ife5+! and wins.
ing the situation on the board . After 22 . . . gxf5 26 . . . c;i;>e7 27 'ii'xf6+ lLldxf6 28 lLlxe8 lLlxd2
23 'ii'xf5 it would have been much easier for 29 lLlxf6 c;i;>xf6 30 .i.a4 lLlxf1 31 c;i;>xf1 l:txc3
Wh ite to conduct the attack. is more critica l , but after 32 c;i;>e2 (32 .:tb1 !?
.i.c6 33 .i.xc6 l:txc6 34 c;i;>f2 is also strong)
23 lLlxf7 !
32 .. .l:tc4 (or 32 . . . a3 33 :b1 .i.c1 34 .l:.b3)
O f course, t h i s is t h e correct conti nuation, 33 g4 Wh ite has a clear advantage.
since now the black king is forced to go for a
Now, however, he simply has insufficient
walk. Not 23 .i. e 1 lLlxe5 24 fxe5 gxf5 25
reserves to gain more than perpetual check.
'ii'xf5 l:tc7, and Black can defend .
25 . . . c;i;>xf5
23 . . . c;i;>xf7
26 g4+
24 'ii'x h7+ c;i;>e6
Noth ing was given by 26 'iff7+ in view of
26 . . . 'ii'f6 27 'it'xd7+ 'ife6.
26 . . . f6
27 f5
I was pinning g reat hopes on this move . 27
e 1 was bad because of 27 . . . c;i;>e7 28
.i. h4+ lLldf6 29 'iVxg6 l:txc3 .
27 . . . .i.g7!
A sober assessment of the position : Black
forces a draw. 27 . . . gxf5?! 28 ':'xf5+ c;i;>e7
was too risky in view of 29 .i.xd5! (not 29
'iVf7+ c;i;>d6 30 :xd5+ c;i;>c7) 29 . . . .i.xd5 30
'ii'x d5. Now both 30 . . . lLlef6 3 1 'iff3 d6 32
e4! lLlxe4 33 .i.f4+ c;i;>c6 34 ':e1 ! or 34 'ii'e 2!
(but not 34 .i.e5 lLlxe5 35 dxe5 because of
The critical moment of the game. Wh ite saw 35 . . . .i.c5+ 36 c;i;>f1 'ifd5), and 30 . . . lLld6 3 1
the correct continuation , but he was unable l:taf1 ! 'ifb6 (31 . . . lLlxf5 3 2 l:txf5 'ifc7 33 e4
to evaluate correctly the endgame arising in c;i;>d8 34 lIxf8 lIxf8 35 .i.g5+ leads to mate)
the main variation . I n add ition , he was 32 l:[f7+ c;i;>d8 33 c4 'ifxa6 34 .i.a5+ .:tc7 35
g ripped by creative feelings: the temptation l:t 1 f6! g ive Wh ite a prom ising position .
to drag the king out even fu rther was too There was also the i nteresti ng move
Missed Brilliancy Prizes ttJ 1 29

27 . . . Wb6 !? - it is impossible to predict how, opponent, who in the event of a win would
with both players short of time, it would all obta i n chances of fi rst place .
have then ended . 1 d4 d6 2 e4 liJf6 3 liJ c 3 g6 4 f4 g 7 5 liJf3
28 fxg6+ c5 6 b5+ d7 7 e5 liJg4 8 xd7+ Wxd7
28 Wh7 :g8 29 Wxg6+ e7 30 'ii' e 6+ f8 9 d5 dxe5 1 0 h3 e4 1 1 liJxe4 liJf6 1 2 liJxf6+
3 1 c1 does not work because of 3 1 . . . liJg5 xf6 1 3 0-0 0-0 1 4 e3 ! ? liJa6 1 5 liJe5
(or 31 . . . f6 32 a3+ g7) 32 a3+ liJc5 Wd6
33 Wg6 xd4 .
28 . . . xg6
Now White repeats moves several times in
order, after reach ing the time control, to
check the variations accu rately and once
again convince h imself that the win , alas,
has already been missed .
29 Wf7+ h7
30 Wf5+ g8
31 Wf7+ h8
32 Wh5+ g8
33 Wf7+ h8
34 Wh5+ g8
35 xd5+ xd5 Wh ite played the open ing confidently, which
36 Wxd5+ h8 was mainly explai ned by the fact that the
37 Wh5+ g8 entire variation and this particular position
had already occu rred in my game with
38 Wf7+ h8
Vlastimil Hort from the German Team
There is no point in giving White chances by
Championsh ip. It is probable that M ichael
38 . . . h7?! 39 :f5 liJdf6 40 :aa5 Wxa5 4 1
Adams simply did not know about this
:xa5 liJxd2 4 2 g5.
game, which was played a few months
39 Wh5+ g8 before Dortmund.
40 Wf7+ h8 1 6 liJg4
Draw. A standard idea in this variation . From g4
the knight su pports the attack well and at a
In concl usion I will give yet another example conven ient moment it is threatening to go to
of a spoiled attack, in which , however, it all h6.
ended happily for the author. 16 . . . xb2
1 7 llb1 g7
Yusu pov - Adams 18 f5
Dortmund 1 994 An important move in Wh ite's plan. The
Pirc-Ufimtsev Defence regaining of the pawn can wait (if 1 8 l:txb7
This game was also played in the last round, there would have followed 1 8 .. .f5), it being
and I wanted to win it without fai l , in order to far more important to develop his in itiative
improve my tournament result somewhat. on the kingside and secure the h6-square
The game was even more important for my for the knight.
1 30 M issed Brilliancy Prizes

18 . . . lDc7?! 20 f6!
In the afore-mentioned game Hort played The point of Wh ite's idea is to exploit the
more strongly: 1 8 . . . lDb4 ! ? 1 9 c4 lDxa2 20 opposition of the queens.
':xb7 lDc3, trying to create counterplay. The 20 . . . exf6
move in the game is rather passive, and 20 . . . i.xf6 was weaker in view of 2 1 ':xf6
Wh ite gains the opportun ity to dictate the lDxf6 22 'iix d6 exd6 23 lDxf6+ g7 24 lDg4!
fu rther cou rse of events. He now has with a winning position .
perhaps too wide a choice : h5
21 c4
A) 1 9 c4 bS; 22 lDh6+
B) 1 9 f6 exf6 20 i.f4 'it'd8 ; But not 22 lDf2? in view of 22 . . . 'iig 3 23
C ) 1 9 lDh6+!? i.xh6 20 i.xh6 l:.fd8 2 1 c4 ; i.xcS lDf4 .
D) 1 9 i.f4!? i.d4+! (as shown by Adams, 22 . . . h7
dangerous is 19 . . . 'it'd8 20 i.xc7 'it'xc7 21 d6 23 lDf5
'ii'd 7 22 dxe7 i.d4+ 23 h 1 'it'xe7 24 f6 with Wh ite has not managed to win a piece , but
an attack) 20 'iix d4 cxd4 21 i.xd6 exd6 22 he completely destroys the opponent's
l:txb7 l:tac8 (22 . . . lDxdS 23 f6 followed by pawn cha i n . 23 'ii'x dS 'iix dS 24 cxdS i.xh6
lDh6+ - Black's f7 -point is weak), and if 23 2S ..txcS would also have led to a better
f6? hS - however, after 23 lDf6+ g7 24 endgame for h i m .
fxg6 fxg6 (24 . . . hxg6 2S a4 ! ? , and if 2S . . . a6,
23 . . . gxf5
then 26 lDd7) 2S lDe4 l:txf1 + 26 xf1 the
advantage remains with White . Although in the variation 23 . . . 'ii' c6 24 'iix dS
'ii'x dS 2S cxdS gxfS 26 i.xcs Black is
During the game my choice was mainly
nominally a pawn up, the endgame with
between this last continuation , which seemed
tripled pawns will hardly afford him any
to me to be not too clear, and the text move .
pleasure.
1 9 l:.xb7! 24 cxd5 g8
The start of a forcing operatio n . Black's 25 1:.xf5
reply is compulsory, since 20 i.f4 is
2S i.f4 ! ? 'ii'a 6 26 'ii' b 1 was interesting, not
threatened .
paying any attention to the f-pawns and
19 . . . lDxd5 concentrating all efforts on the advance of
the passed pawn .
25 . . . 'ii'a 6
26 'ii' b 1 1:.fe8
I was expecting 26 . . . 'ii'e 2 , after which 27
i.f2 is the simplest way to retain the
advantage. 27 i.xcs l:[fc8 28 1:.b2 is also
possible, only not 28 d6? 1:.xcS 29 ':xcS
'ii'e 3+ 30 h 1 'ii'x cs 3 1 l::t b 8+ l::t x b8 32
'ii'x b8+ h 7 33 d7 in view of 33 . . . 'ii' c 1 +
(transposing moves does not work: if
33 . . . i.h6? Wh ite has 34 'ii' b 1 + , winning) 34
h2 i.h6 3S d8'ii' i.f4+.
The continuation in the game also parries
the obvious threat of 27 ':xhS.
M i ssed Brillia ncy Prizes CZJ 1 31

32 'iVb1 g8
33 l:txe5 fxe5
34 'it'b8+ h7
35 'iVc7
It was on this move that I was pinning my
hopes. 35 d6 'iVd3 (35 . . . f6 ) 36 'ii c7 g6
(36 . . . c4 37 'iVxf7 'ii'x d6 38 'iix h5+ 'iVh6 is
also possible) 37 d7 f6 is not dangerous
for Black. Since now, apart from the ad
vance of his passed pawn , Wh ite is also
threatening the f7-pawn , I was feeling
optimistic, u ntil I noticed a defence. Of
cou rse, my opponent also found it - Adams
does not miss such chances!
I had no doubts about the assessment of the
position , but in the calculation of variations I 35 . . . 'iix a2 1
began to get confused . Everywhere I i mag 36 'it'xf7 'ifb1 +
ined some kind of counterplay for the This is the point! The queen switches to the
opponent. As a result I decided to play as kingside with gain of tempo.
simply as possible, by analogy with the 37 h2 'iVg61
26 . . . 'iVe2 27 f2 variation . Of course, Wh ite 38 'iVxa7
should have exerted h imself a little and
U nder the impression of his poor play in the
ascertai ned that after the simple 27 xc5!
tech n ical stage, Wh ite takes a sensible
Black's minimal activity does not cause any
practical decision - he wants to red uce to
great problems: after 27 . . . 'iVa5 there follows
the minimum the probabil ity of losing the
28 b4 . a6 29 :Xxh5, while if 27 . . . liac8 -
game. The bolder 38 'iic 7 'iVf5 would have
28 d6. The sharpest continuation 27 . . .1le2
left the opponent's passed pawn alive.
leads after 28 .l:1b8+ (28 l:xh5? l:txg2+)
28 .. Jtxb8 29 'iVxb8+ h7 30 l:txh5+ g6 38 . . . c4
3 1 lih4! to an easy win . 39 'iVc7
2 7 f2?
Wh ite reckoned that after the practically
forced exchange of rooks his passed pawn
would decide the outcome, but he over
looked a strong defensive manoeuvre by
the opponent.
27 . . . :e5
28 l:.b8+ lixb8
29 'iVxb8+ h7
30 'iVb1
A usefu l device. To avoid time-trouble,
Wh ite repeats moves.
30 . . . g8
31 'it'b8+ h7 39 . . . 'it'd3?
1 32 M issed Brillia ncy Prizes

Upset by the cou rse of the game, which did overlooked Wh ite's 42nd move.
not leave h i m any chances of first place i n 41 d7 e2
the tournament, Adams was unable to
42 .ie3 !
concentrate fully on the fight for a draw and
he made this natu ral but losing move almost Black's downfall is caused by the fact that
without th inking . Meanwhile Black had a his king is on the same rank as the wh ite
way to save the game. After 39 . . . 'ii'f5! ! 40 queen, and if he moves his bishop there is a
.ig3 'ii'e4 41 d6 (4 1 'ii'f7 also leads to a decisive discovered check.
draw) 4 1 . . . h4 or 40 .ie3 'iVe4 4 1 .ig5 'iVxd5 42 . . . 'ii'x e3
(4 1 . . . c;t;>g6 42 d6 .i f6 43 .ixf6 'iff4+ 44 c;t;>g 1
43 'ii'x e2+ e4
'ilVe3+ 45 c;t;>f1 'iVd3+ is also possible) 42 .if6
'ii'g 8 43 .ixe5 c;t;>h8 44 .ixg7+ 'ii'x g7 45 44 'ii' e 7!
'ii'xc4 'ii'e 5+ Black should gain a draw. The simplest. There is no point in calculat
40 d6 e3 ing the more complicated 44 d8'ii' .ie5+ 45
40 . . . c;t;>g6 41 d7 .if6 was rather more g3 (wh ich , however, was also sufficient for a
tenacious, although after 42 .i b6 Black's win) when there is a simple solution .
position is d ifficult. My opponent obviously Black resigned.
CD 1 33

Mark Dvoretsky

Lon g -d i sta n ce D i s p ute

7 e4 i.. g4
S sacrifice?
hould I take a risk? Should I make a
Questions such as these 8 i.. e 3 tLlfd7
qu ite often have to be solved . It is clear that 9 'iVb3 i.. xf3
here there is not and can not be a general
Black wants to develop his knight at c6 , but
prescription . The best that readers ca n be
the immediate 9 . . . tLlc6 ru ns into 1 0 'iVxb7
advised to do is refer to books and articles in
tLla5 1 1 'it'a6 , as in the game Polugayevsky
which this type of situation is analysed . Test
Simag i n , played in Leningrad in the 1 960
them on yourself - try, by deeply analysing
USSR Championship (however, after 1 1 . . . c5
the position, to decide how you would act in
1 2 dxc5 l1b8, according to the Encyclopae
this or that case , and then check your
dia of Chess Openings, the position is
reasoning with the commentator's conclu
unclear).
sions. By acti ng in this way, you will not only
develop your tech nique of calculating varia The prel iminary exchange on f3 , elimi nating
tions, but also learn to determ ine intu itively one of the defenders of the d4-point, does
the degree of acceptable risk. not leave Wh ite time to captu re the pawn on
b7. However, it also has its drawbacks, and
I should like to show you the analysis of a
therefore the main theoretical conti nuation
sharp position, which occu rred in a game of
became 9 . . . tLlb6.
the Soviet master Vlad imir Simagin (he
became a grandmaster much later). Eleven 10 gxf3 tLlc6
years later (without having any knowledge 1 1 1:[d1 ?
of that previous game) the same position Now Simagin's idea proves justified . As
was obtai ned by Bobby Fischer. The opin later practice showed , by playing 1 1 O-O-O !
ions of Simagin and Fischer d iverged . You Wh ite gains an advantage.
have the opportun ity to make a choice , to 11 . . . e5
decide which player's handling of the 1 2 dxe5
position was more correct.
1 2 d5?! tLld4 is unfavourable for Wh ite.
12 . . . tLlcxe5
Shamkovich - Simagin
1 3 i.. h3
Leningrad 1 95 1
Leonid Shamkovich plays aggressively, hop
GrOnfeld Defence
ing to exploit the pin on the knight at d7. I n
1 d4 tLlf6
the event o f 1 3 i.. e 2 Black has the excel lent
2 c4 g6 reply 1 3 . . . 'it'h4 ! , and if 14 f4 , then 14 . . . tLlg4.
3 tLlc3 d5 13 . . . tLlxf3+!
4 tLlf3 i.. g 7 1 3 . . . 'it'h4! ? 14 i.. x d7 l:tad8 would also have
5 'it'b3 dxc4 g iven Black a good game.
6 'it'xc4 0-0 1 4 e2
1 34 Long-d i stance Dispute

If 1 4 'iti>f1 Simagin g ives the variation I n the game Evans-Fischer (USA Champi
1 4 . . . lDfe5! 1 5 i.. x d7 lDxd7 1 6 'iVb5 c6 1 7 onship 1 962/63) Black d id not risk going in
'ili'xb7 i.. xc3 (the immed iate 1 7 . . .'iVh4! is no for the complications and he restricted
worse) 1 8 bxc3 'ili'h4 ! . Now 1 9 'iVxd7 lIad8 h imself to the simple 1 9 . . . 'ii'x d7 20 l:1xd7
is not possible, while after 1 9 lIxd7 both i.. x c3 . A draw became practically inevitable.
1 9 . . . .:tab8 and 1 9 . . . 'iVxe4 are strong . 21 l:txa7 :te8 22 .l:.a4 i.. b4 23 i.. d4 ':c2 24
14 . . . lDfe5 l:.xb4 c5 25 i.. x c5 ':'xc5 26 'iti>g2 ':c2 27 a4
1 5 i.. x d7 l:td8 28 'iti>g3 ':'a2 29 ':c1 %1dd2 30 I:tf1
l:td3+ 31 f3 %:tda3 32 ':'d1 l1xa4 33 l::t d 8+
If 1 5 ':'xd7, then 1 5 . . . 'iIi'h4! ( 1 5 . . . lDxd7 1 6
'iti>g7 Draw.
l:td 1 is less good ). 1 5 f4 'iVh4 1 6 i.. x d7
lDxd7 1 7 lixd7 'iVg4+ will also not do. Simagin acted differently. He decl ined the
lDxd7 d raw offered at that moment and sacrificed
15 . . .
a piece.
1 6 'iVb5 c6
19 . . . 'ili'f6 ! !
1 7 'ili'xb7 l:tb8
I n fact, i t is also not easy to refute the move
1 8 'ili'xd7
1 9 . . . 'iVh4? ! . 20 'ii' x c6? (or 20 'ili'd3?) is bad
Of course, not 1 8 'iVxc6? l:txb2+ 1 9 'iti>f1 in view of 20 . . . 'iIi'h3+ 21 'iti>e1 'iVf3 , attacking
'iVh4 ! . the rook and threatening mate after 22 . . .
18 . . . lIxb2+ i.. x c3+. I f 2 0 lDa4?! Black can reply
19 'iti>f1 20 . . . l:txa2 21 lDc5 i.. h6 22 'ifd3 'ili'h3+ 23
'iti>e1 i.. x e3 24 'iVxe3 'ii'x e3+ 25 fxe3 .:1fb8
with sufficient cou nterplay. The strongest
conti nuation is 20 lDe2 ! 'ili'xe4 21 ':g1
(weaker is 21 lDg3 'ili'f3 22 ':c1 f5! )
21 . . . l:txa2 2 2 ':'c1 ! ? , intending 23 'ili'xc6 or
23 'iVg4 followed by 'ii' c4 . Wh ite success
fu lly consolidates and retains an advantage.
Fischer considered the sacrifice made in the
game to be completely incorrect. But Simagin
tried to show that Black's combi nation leads
to a win. I th ink that the truth l ies somewhere
in between . Let us examine some varia
tions.
1 . 20 lDa4? This is what Shamkovich played
in the game. After 20 .. J::txa 2 21 lDc5 'iVf3 22
Here is the position in which I invite you to 'iti>g1 (22 1:[g 1 'ii' e 2+ 23 'iti>g2 'ili'xe3)
take a decision for Black. He has a choice 22 ... i.. h6! Black's attack became irresist
between regaining the knight, transposing ible. There followed 23 i.. d 4 'ili'xd 1 + 24 'iti>g2
into a roughly equal ending, and the attempt 'iVd2 25 'iVd6 i.. e 3! 26 lDd7 'ili'xf2+ 27 'iti>h3
to attack a piece down by either 1 9 . . . 'iVh4 'iVg2+, and Wh ite resigned .
(from here the queen controls the h3-square II. 20 lDe2? 'iVf3 21 lDg3 i.. h61 It is
and attacks the pawn on e4), or 1 9 . . . 'iVf6 apparently not possible to defend the wh ite
(aiming at the weak f3-square). Which king, for example: 22 i.. x a7 (noth ing is
would you prefer? changed by 22 i.. c5 ':c2 23 i.. x a7 lla8! 24
Long-d i stance Dispute 4:J 1 35

l:.e 1 .l:t.xa2) 22 . . . l:ta8! (threatening 23 . . . .l:t.xa7) The only q uestion is whether Black should
23 :e 1 (23 i.c5 i.f8 24 i.e3 ':'axa2 ; 23 be satisfied with a d raw, or whether he has
l:.a 1 l:1xf2+! 24 i.xf2 i.e3) 23 . . . l:.xa2 24 the right to continue the attack with 21 ... c5! ?
i.c5 i.d2! 25 l:1b1 l:ta 1 26 'ii' b 7 i.c3! with Simagin thinks that he does. He g ives the
the decisive threat of 27 . . . l:txb 1 + 28 'ii'x b 1 variation 22 l:tg3 'ii' h 1 + 23 l:tg 1 'it'xh2 24
:a 1 (analysis by Simag i n ) . 1:Ig2 'ifh 1 + 25 J:tg 1 'ifh4! with an attack.
III. 2 0 i.d4 ! ? 'it'f3 Wh ite can play more strongly: 22 1i.xc5!
1i.xc3 23 'ii'd 31 'iff6 24 1:Ig3 . By allowing the
20 . . . 'it'h4 is weaker in view of the excellent
opponent to restore material equal ity, he
reply 21 ttJ d 5 ! , pointed out by F ischer. Then
activates his forces. 24 . . . 1i. b4?? loses im
21 . . . i.xd4? 22 ttJe7+ is completely bad .
mediately to 25 1i.d4, and Black resig ned
After 2 1 . . . 'il'xe4 22 ttJe7+ 'iti>h8 23 i.xg7+
(McLellan-Kokori n , correspondence 1 968).
'iti>xg7 24 'iVd4+ (24 :g 1 ) 24 . . . 'ii'x d4 25
J:txd4 Wh ite should be able to convert his 24 . . . 1i.e5 25 ':'f3
piece advantage. But even here Black is by
no means doomed - he plays 21 . . . cxd5 22
1i.xb2 i.xb2 23 'il'xd5 'il'h3+ 24 'iti>e2 'ii'g 4+
with a probable d raw.
21 :g1
21 'iti>g 1 ? will not do because of 21 .. Jlc2 or
2 1 . . . c5 .

25 . . . 'ii' h 4 suggests itself, when 26 1i.xf8?


'it'xh2 27 ':'d2 .l:t. b 1 + 28 'iti>e2 'ii'g 1 29 l:t d 1
lIb2+ 30 l:. d 2 l:t b 1 leads only t o a d raw.
However, the simple move 26 h 3 ! , pointed
out by Larry Evans, sets Black insuperable
d ifficulties. For example: 26 . . . :tfb8 (26 . . .
l:.c8? 2 7 'it'd7 ':xc5 2 8 'it'xf7+) 2 7 'ii' d 5!
( Evans suggested 27 ':'xf7!? 'iti>xf7? 28
It was because of this variation that Fischer 'iVd7+ 'iti>g8 29 'it'e6+, but Black has a
rejected the piece sacrifice. But this was tougher defence: 27 . . . l1c2! 28 'it'xc2? 'ii'x h3+
wrong - after a l l , if Black wishes, he can 29 'iti>e2 'iti>xf7 with equal ity; however, after
force a draw by 21 . . . l:tc2 ! ? 22 l:tg3 (forced) 28 i.e7! 'it'h5 29 ':'f3 White retains the
22 . . . 'ifh 1 + 23 1:[g 1 'ii'f3 . It is also probably a advantage) 27 . . . i.f4! 28 e5! with a winning
draw in the endgame arising after 23 . . . 'ifxh2 position (only not 28 i.d6? ':xf2+! 29 ':'xf2
24 ttJe2 'ifh5 25 'ifg4 'ifxg4 26 J:txg4 l:td8 27 'it'xh3+ with a draw).
'iti>e1 l:txe2+ 28 'iti>xe2 i.xd4 29 1:Ig5 'iti>f8 30 I have been able to fi nd a way of strengthen
l:tc5 1i.xc5 31 l:txd8+ 'iti>e7. ing the attack: 25 ... 'it'g5 ! ! 26 1i.xf8 1i.xh2 27
1 36 \t> Long-distance Dispute

'ito>e1 , and now not 27 . . . 'ii'g 1 +? 28 'ii'f 1 'ii'g 4


(28 . . .''' g 5 29 .i.h6! 'ii'x h6 30 'ii'c4) 29 .i.h6!
'ii'x e4+ 30 .i.e3 'ii'xf3 3 1 l1d8+ 'ito>g7 32
.i.d4+ 'ito>h6 33 .i.xb2 , but 27 ... .i.c7 ! ! with
the th reats 28 . . . 'ito>xf8 and 28 . . . 'ii'g 1 + 29 'ii'f 1
.i.a5+. 28 .ltb4 'ii' g 1 + 29 'ii'f 1 'ii'g 4 leads to
a repetition of moves. The clever move 28
.i.g7 ! ? would be justified after 28 . . . 'ito>xg7+?
29 "'c3+ i.. e 5 30 .l:txf7+ ! ! , but Black plays
28 ......g 1 + 29 Wf1 .i.a5+ 30 .i. c3 'ii'g 4 31
i.. x a5 'iVxf3 32 .i.d2 We4+ 33 'ii'e2 (33 .i.e3
'ii' b4+ ) 33 ... Wh 1 + with perpetual check.
IV. 20 lId3 ! ? 'ii'f3 . The less accu rate move
20 .. Jlc2?! is justified after 2 1 ltJe2?! 'ii'f3 22
ltJg3 .l:tb8! (22 . . . .l:txa2 or 22 . . . h5 is weaker After t h i s S i m a g i n exami ned 2 1 . . . .i.xc3?! 22
because of 23 .i.d2; if 22 . . . .i.h6 there l:txc3 l:tbb8 (an unexpected retreat: Black
follows 23 'ii'd 4 with the threat 24 .i.xh6) 23 creates the threat of 23 . . . .l:tfd8) 23 l:tc1 .:tfd8
l:.b3 lIxb3 24 Wd8+ (24 axb3 h5) 24 . . . .i.f8 24 'ii'xc6? l:td 1 + 25 J:[xd 1 Wxd 1 + 26 'ito>g2
25 axb3 l1b2 26 'ito>e1 h5, and the in itiative is 'ii'g 4+ 27 'ito>f1 l:td8! and wins. However,
seized by Black. instead of the captu re of the c6-pawn , 24
I n the variation 2 1 .i.d2 l:txd2 22 l:txd2 'ii'x c3 'ii'c7 ! is far stronger. By retu rn ing with his
23 .l:t g 1 ! followed by ':g3 a position slig htly queen to g3, Wh ite parries the attack. Black
better for Wh ite is reached . (23 'ito>g2?! stil l retains some in itiative, but it should
suggests itself, but this encou nters the g radually evaporate.
u npleasant reply 23 . . . We5 ! , when defend Black's prospects are hardly improved by
ing the e4-pawn is awkward : 24 'ii'x c6? other attempts on the 22nd move :
Wg5+, or 24 :e1 ? Wg5+ and 25 . . . .i.c3). 22 . . Jie2 23 'ii'd 1 ! Wg4+ 24 'ito>f1 'ii'f3 25
2 1 ltJd 1 ! Wf3 22 .l:tg 1 'ii'xe4 23 l:td2 l:txd2 24 lId3! l:txf2+ 26 i.. xf2 Wxh 1 + 27 .i.g 1 ;
'ii'x d2 is more promising for White - the 22 . . . :t.fb8 23 l:txc6 l:t b 1 + 24 .l:tc1 l:txc1 + 25
compensation for the piece is probably .i.xc1 .
insufficient. The exchange on c3 is bad - instead
2 1 'ito>g1 ! 21 ... lIc2 ! can be recommended . In reply 22
.ltd2? is a mistake in view of 22 . . . ':xd2 ! .
(see diagram)
After 2 2 .lt c5?! Wf4 23 Wd6 Black achieves
The natural 2 1 :g 1 ? is incorrect in view of a favourable ending by 23 . . . 'ii' x d6! 24 .i.xd6
2 1 . . . .i.xc3 22 .l:txc3 l:tfb8 (th reate n i n g 1:.d8 25 ltJd 1 .lte5. I n the event of 23 'ii'e 7
23 . . . l:txf2 + ! ) 23 Wxc6?! (more tenacious is Black's resou rces are illustrated by the
23 i.. d 2 Wd 1 + 24 'ito>g2 'ii'x d2 25 'ii'x d2 :txd2 following curious variation : 23 . . . Wg4+ 24
with an extra pawn for Black in a double l:tg3 'ii' c8 ! 25 'ii'x a7 (25 . . . l:te8 was th reat
rook endgame) 23 . . . .l:.d8 24 1:.c1 'ii'e 2+ 25 ened ) 25 . . . l:td8 26 ltJa4 l:txa2 27 'ito>g2 'ii'e6
'ito>g2 Wxe3, and Black wins (Simag i n ) . 28 :e1 'ii'e 8! followed by 29 . . . l:ta8, and the
knight at a4 is lost.
22 .ltxa7 !? is possible, although after
22 . . . l:tc1 + 23 ltJd 1 Wxe4 24 .i.e3 l:t a 1 Black
retains qu ite good counter-chances.
Long-dista nce Dispute ctJ 1 37

The same assessment applies to the posi though he overestimated his position. One
tion arising after 22 d 1 ! ? Wxe4 (or can argue about the analytical correctness
22 . . . 1:[e2 23 d2 Wxe4 24 h3 c5). Here the of the piece sacrifice, but from the practical
outcome remains u nclear. point of view it is certainly justified . The
V. 20 l:tc1 ! This move, suggested by the probabil ity of the opponent fig uring out the
Brazilian g randmaster Gilberto M i los, may complications and finding all the strongest
cast dou bts on Simagin's bold idea. Having moves at the board is pretty smal l . Wh ite is
defended his knig ht, at the same time Wh ite in far more danger - after the slig htest
does not al low the reply 20 . . JIc2 , wh ich inaccu racy the attack will become i rresist
gave the opponent counterplay after 20 ible.
l::t d 3 . He is not afra id of 20 . . . l:1d8 in view of
It is curious that Bobby Fischer, a fig hting
21 'ii' h 3, while in the event of 20 . . . .th6 he
player who always aimed only for a wi n , did
can choose between 2 1 Wh3 and 2 1 'it'xa7
not risk sacrificing the piece and satisfied
l:txf2+ 22 xf2 xc1 23 Wd4 . There only
h imself with a d raw. The American grand
remains 20 . Wf3 21 g1 ! (but, of course,
master valued clarity, did not l i ke to lose
..

not 21 1:[g 1 ? .txc3 22 1:[xc3 l:fb8), when


control of what was happen ing on the
2 1 . . . .th6 22 'ii'd 1 is hopeless, while 21 . . .
board , and therefore mistrusted irrationa l ,
.txc3 2 2 1:[xc3 leads to a situation favour
intu itive sacrifices o f material. I n t h i s respect
able for White, familiar to us from the 20
his style d iffered sign ificantly from that of
l:ld3 variation .
brill iant chess artists such as Vlad imir
I t is time to s u m u p . I n the long-d istance Simag i n , M ikha il Tal and Alexey Shirov, for
dispute between Simagin and Fischer, it is whom risk, involving problematic sacrifices,
Simag i n who was the more correct, al- is natu ral and usua l .
1 38
Mark Dvoretsky

Attacks with oppos ite-s i ded Castl i ng

M th is topic - after a l l , it is discussed in


ost of you will probably be familiar with exchange on d3 and recaptu re with the c
pawn. Should two tempi be wasted in the
many books about the middlegame. It is open ing, even for the sake of exchanging
u n l i kely that I will be able to say anyth ing the opponent's strong bishop?
new, and in fact I am not aiming for th is. We 9 exf6 lhxf6
will simply analyse a few games, in the 1 0 'iVe2 a6
cou rse of which we will recal l some impor
could simply have castled, but I was
tant features, typical of positions with castling
attracted by a positional trap. The tempting
on opposite sides, and do some training on
1 1 f5?! would have allowed Black to
their practical appl ication .
advantageously sacrifice a pawn : 1 1 . . . e5!
( 1 1 . . . 0-0 is also not bad ) 1 2 lhxe5 lhxe5
Geo rgad ze - Dvorets ky ( 1 2 . . . lhd4 ! ? ) 1 3 \!kxe5+ -.tf7 , and the wh ite
USSR Spartakiad , Moscow 1 967 king , which is caught in the centre, comes
French Defence under a dangerous attack.
1 e4 e6 11 .ltd2 0-0
2 d4 d5 1 2 0-0-0 'ii'c 7
3 lhc3 lhf6 With opposite-sided castling the two
4 e5 lhfd7 players usually conduct pawn storms on
5 f4 c5 opposite wings, trying as soon as possi
lhc6 ble to weak the enemy king's defences.
6 lhf3
From this point of view the move 1 2 . . . b5!
7 dxc5
seems logica l . I thought that Wh ite would
This move does not promise Wh ite any reply 1 3 g4 b4 1 4 lha4 .ltd6 1 5 g5, and if
advantage. The critical continuation is 7 1 5 . . . lhh5 - 1 6 lhe5. Alas, this conclusion is
e3 ! ' i ncorrect in view of a l ittle combination :
7. . . .ltxc5 1 6 . . . lhxf4! 1 7 .ltxf4 lhxe5 1 B xe5 'iVxg5+.
7 . . . lhxc5 is also not bad . Having failed to calculate the variation to the
8 .lt d3 f6 end, I decided to make a preparatory move,
B . . . O-O? is a mistake because of the strengthening Black's position in the centre.
standard bishop sacrifice 9 .ltxh7+! -.txh7 I n principle, also a sound idea : don 't forget
1 0 lhg5+. The move in the game is perfectly about control of the centre even when
logical - I prepare castl ing and at the same sharp wing attacks are in progress.
time exchange the strong central pawn on 1 3 g4 b5
e5. Black has also played d ifferently here: Of course, the captu re of the g4-pawn was
B ... lh b4 or B ... a6 followed by 9 ... Wc7. I don't not even considered .
like the knight move - White simply retreats
his bishop to e2, but he can also allow the
Attacks with opposite-sided Castl ing ltJ 1 39

variation 1 8 . . .t'Dxf4? 1 9 i.xf4 l:lxf4 20 t'Df6+


i.xf6 21 gxf6 ':'xf6 22 h5) is refuted
tactically: 1 8 . . . dxe4! 1 9 i.xe4 t'Dg3 ! . And if
1 8 %1df1 ? ! (with the idea of 1 9 t'Dg3) there is
the un pleasant reply 1 8 . . :ii'c7 ! .
1 8 . . . t'Dxf6 1 9 gxf6 lIxf6 2 0 h 4 e 5 ! (otherwise
Wh ite's attack becomes dangerous) 21 fxe5
l:lf2 22 'ii' e 1 i.g4, and now Wh ite must
either sacrifice the exchange by 23 h5, or
choose 23 i.e3 'ii' c5 24 i.xd4 'fixd4 25 l:Id2
:xd2 26 'ii' xd2 , parting with the e5-pawn ,
but in retu rn retaining attacking possibil ities.
14 l:tdf1 ?!

Question : what would you now play as A seemingly sensible move - Wh ite sup
Wh ite? ports his f4-pawn in advance and vacates
With opposite-sided castling one must the d 1 -square for the retreat of his knig ht.
act as energetically as possible, trying at But even such a m i n i mal delay is already
any cost to seize the initiative. Here the sufficient for Black to be the fi rst to lau nch
slightest delay is usually fatal. his assault.
The principle itself is perfectly clear, but Here I should like to take the opportun ity to
sometimes it is not easy to fol low it. For quote an idea of Alexander Kotov regarding
example, the attempt to undermine the mutual attacks with opposite-sided castl ing,
enemy centre by 14 f5?! exf5 1 5 g5 is bad in which he thought was important. 'When
view of 1 5 . . . t'De4 1 6 t'Dxd5 'fif? beginning a pawn storm, you should bear in
I th i n k that the correct continuation was the mind that it is of a forcing nature and you
sharp 1 4 g5! t'Dh5 1 5 t'De5 ! . Now it is should calculate it as accurately as you
extremely dangerous to accept the pawn would calculate a combination '.
sacrifice: 1 5 . . .t'Dxf4 1 6 i.xf4 ':xf4 1 7 t'Dxc6 I don't agree with Kotov's idea . I ndeed , the
'ii'xc6 1 8 'fih5 ( 1 8 :hf1 ! ? or even 1 8 outcome in such cases sometimes hangs by
i.xh7+ !? also comes into consideration) a thread , and depends on a single tempo.
18 . . . g6 19 i.xg6 hxg6? 20 'fixg6+ h8 2 1 The calculation of variations plays an
lLlxd 5 . important role, but nevertheless not the
Tamaz Georgadze was probably concerned lead ing one - it helps specific problems to
about the reply 1 5 . . . g 6 ! , after which the be solved , but usually (as in the given
weakness of the f4-pawn is very percepti game) it does not enable the fate of an
ble. In sharp situations with opposite attack to be accu rately determined before
sided castling, for the sake of the hand . Therefore you should not be too
initiative you sometimes have to go in for carried away by calculation, and, of cou rse,
positional or material concessions, and you must not be restricted to it. It is
you should not be afraid to do this. Let us important to sense the spirit of the position ,
continue 1 6 t'Dxc6 'ii' xc6 1 7 t'De4! i.d4 1 8 and to be able to assess intu itively the
lLlf6+ ! ' prospects of the two sides, whatever d i rec
The attempt to prepare this check, by tion events may take.
playing 1 8 h4? (wh ich is justified in the 14 . . . b4
1 40 Attacks with opposite-sided Castl ing

1 5 ttJd1 d6 Where did the mistake lie? In the assess


16 ttJ e5 b3! ment of the fi nal position . I did not take into
This is far stronger than the prim itive account the weakness of the f3-square (if
captu re on e5, which would have led to the wh ite pawn had been on g2, the position
unclear play. Black solves his main strateg ic would indeed have been unclear). As soon
problem - he weakens the pawn defences as the knight goes to e3, the rook will
of the enemy king . immediately occupy the f3-point. There it will
exert unpleasant pressure on the oppo
1 7 axb3 ttJd4
nent's position , and the exchange on f3 is
1 8 'ii'e 1
completely hopeless for Wh ite.
Here I made perhaps my only serious
Of course, when you look at the diagram it
mistake in the game - after a long think I
all seems obvious, but during a game you
accepted the draw offered by my opponent.
may miss someth ing at the end of a long
Such mistakes should be analysed , to
variation . However, is this any justification?
understa nd why they were made. There
It is important to think about how to avoid
may be purely chess reasons, as wel l as
such mistakes in the futu re .
psycholog ical ones, or sometimes these
and others are interwove n . If in the process of calculating you
First t h e purely chess reason. I n calculating sometimes do not have a very clear
the variations, I decided that the following impression of the position (a part of the
position would most probably arise: board falls out of your field of view, you
forget the exact placing of certain pieces,
1 8 . . . ttJxb3+ 19 'it>b1 ttJxd2+ 20 'ii'x d2 xe5
and so on), special training is needed. As
21 fxe5 'ii'x e5 22 :e 1 ? ! (22 g5 is stronger,
often as possible you should analyse
forcing the retreat of the knight to d7)
positions which interest you without
22 . . . ttJe4 23 xe4 dxe4 24 .l:.hf1 b7.
moving the pieces, look through games
printed in books or magazines without
using a board, and play blindfold with
other players.
And now about the psychological reason for
my mistake. At that time I was a young and
inexperienced player, but I played (on the
j u n ior board ) in the same team as such
g reats as M ikhail Botvi n n i k and Vasily
Smyslov, and, understandably, I was very
anxious and afraid of letting the team down .
As a result I spent an un usually long time
checking and re-checki ng variations, and by
the point when peace was ag reed I had less
than half an hour for the remaining 23
Black is a pawn u p , but it is doubled , and the moves. I ncidentally, after the game I showed
wh ite knight is ready to occupy the excellent Botvinnik the fi nal position and the possibili
blockading square e3. I was not sure that ties in it which I had considered . After asking
the advantage was on my side here, and I how much time I had remaining, the ex
did not find anyth ing better. Therefore I champion said that I had done right to agree
ag reed a draw. a draw.
Attacks with opposite-sided Castl ing l2J 141

Even s o , i t is clear that a cool-headed , self Bronstein - Dvoretsky


confident player would never have accepted USSR Championsh ip, First League,
a draw offer in such a position. If he was not Odessa 1 974
satisfied with the concluding position of the French Defence
va riation calculated , he would have care
1 e4 e6
fully looked for an improvement earl ier. And
2 d4 d5
he would surely have found that the simple
20 . . . .1l.b7! (instead of 20 . . . xe5?! ) guaran 3 ttJc3 ttJf6
tees Black a g reat advantage in view of the 4 e5 ttJfd7
th reats of 21 . . . d4, 21 . . . ttJe4 and 21 . . . .1l.xe5 . 5 f4 c5
You must tirelessly develop your fight 6 ttJf3 ttJc6
ing spirit, resistance to disturbance, and 7 dxc5 .1l.xc5
ability not to lose your composure in any
8 a3?!
situation, even the most complicated,
I n my view, this is anti-positiona l . I will
and not to give in to strong opponents.
explain why:
Without this you won 't achieve any great
success in chess. 1 ) In the opening you should qu ickly
develop the pieces , not wasting time on
A striving at important moments to act
insign ifica nt pawn moves.
with particular care and safety often has
an adverse affect on a chess player's 2) It is difficult for Wh ite to castle kingside,
actions. A fter all, he betrays his custom and i n the event of queenside castl ing the
ary mode of behaviour and thinking, and advance of the a-pawn will make it easier for
this is rather dangerous. Black to open li nes for the attack. It is
extremely dangerous to go in for a
Why? ! I will give a simple analogy. Anyone
position with opposite-sided castling, if
fi nds it easy to walk along a log wh ich is
there are defects in the pawns covering
lying on the grou n d . But if that same log is
your king.
placed across a sheer d rop, then an
unprepared person will most probably fal l . Wh ite wants his bishop to feel comfortable
On t h e grou nd, when there is no danger, our at d3, without having to fear an attack by the
movements are largely automatic, and they knight from b4 . But he pays too high a price
are d irected by our sub-conscious, which for th is.
does this q uite wel l . But over the sheer drop 8. . . 0-0
our fear of fal l i ng forces us to control every 9 .1l.d3 f6
step, to try and avoid the slightest mistake. 1 0 exf6 ttJxf6
As a result, the natural combination of
1 1 'iWe2 a6
conscious and sub-conscious is d isrupted ,
1 2 .1l.d2 .1l.d7
and an unusual mode is always more
difficult to follow. I delay the advance of my b-pawn , to avoid
'frightening' my opponent. If 1 2 . . . b5 he
would probably have answered 1 3 ttJd 1 .
1 3 0-0-0 b5
1 4 g4? !
This cou nterattack on the kingside is clearly
too late. It was better to play 14 ttJe5.
14 . . . b4
1 42 Attacks with opposite-sided Castl ing

1 5 axb4 lLlxb4 did not satisfy Black. In the middlegame his


advantage is far more sign ificant.
24 lLlxd5 l:txd4
25 lLlc3 f4+
26 'ifi>b1 d2
26 . . . l::t d 2 27 'ili'e4 nb8 was tempting, but
after 28 'ili'xf4 I did not fi nd a way of mating
the opponent.
27 l:[d1 xc3
28 bxc3 l:l b8+
29 'ifi>c1

The defects of the move a2-a3 are now


obvious. Black has opened l i nes on the
queenside and gai ned the advantage.
16 g5?!
Pawn advances often lead to the creation
of weaknesses in your own position. If in
return you obtain a strong attack, it is
worth going in for this. But if the attack
does not succeed, the weaknesses will
tell. I n the g iven instance it is clear that only
Black will be attacki ng, so why weaken the
A very simple l ittle exercise: suggest the
f4-pawn?
most accu rate continuation of the attack.
16 . . . lLlxd3+
I did not want to exchange a pair of rooks.
1 6 . . . lLlg4 is also possible, as well as the This can be avoided with the help of a
immed iate 1 6 . . . lLl h 5 , si nce the bishop sacri simple tactical idea .
fice on h7 is not dangerous. 29 . . . 'ili'c6 1
17 cxd3 30 'iVc2 l:la4
If 1 8 'ifxd3 the simplest reply is 1 8 . . . lLlg4 . 31 l:[hg1 l:r.a3
17 . . . lLlh5 3 1 . . . 'ifb6 was also strong .
18 d4 d6 32 'ifi>d2 'ili'c4
1 9 lLle5 lLlxf4 33 'ifd3 l:1b2+
20 -txf4 ':'xf4 34 'ifi>e1
21 h4 'ili'c7 Or 34 'ifi>e3 lbc3 . Now the exchange on d3
22 lLlxd7 'ili'xd7 followed by a check on the 1 st ran k would
23 l:tde1 :e8 1 have won a rook, but the move made by me
Of course, going into an endgame a pawn leads to a forced mate .
up by 23 . . . l:txd4 24 'ifxe6+ 'ili'xe6 25 l:1xe6 34 . . . 'ili'xh4+
Attacks with opposite-sided Castling ltJ 1 43

35 'ii'g 3 To be honest, I did not sacrifice the pawn ,


3S :g3 'ii' h 1 + 36 'ii'f 1 'iVe4+. but simply blundered it, and so a q uestion
35 . . . 'ii'e4+ mark is attached to my move. But the
exclamation mark added to it reflects the
Wh ite resigned .
fact that Black nevertheless gains some
Even great players have bad days, when (although probably insufficient) positional
they play below their usual strength . Appar compensation for the lost pawn . With
ently David Bronstein had such a day - and opposite-sided castling it is very impor
so victory came to me easily. tant to be the first to begin an attack. For
the sake of rapidly opening lines in the
Here is another 'light' game. vicinity of the enemy king it sometimes
makes sense to sacrifice a pawn.
Maryasin - Dvoretsky In the su bsequent stage of the game my
Kiev 1 970 opponent played uncertainly, and the ad
Pirc-Ufimtsev Defence vantage gradually passed to me.
9 ..ta4 'ii'c 7
1 e4 g6
1 0 h4
2 d4 ..tg7
1 0 g4 came into consideration , intending 1 1
3 tLlc3 c6
g S tLl h S 1 2 tLlge2 followed by 1 3 tLlg3 .
4 ..tc4 d6
10 . . . h5!
5 'ii'f3 e6
11 ..tg5?! tLlh7
6 ..te3
12 ..t e3 tLld7
Here 6 .1i.f4 and 6 tLlge2 have also been
1 3 g4 hxg4
played .
1 4 'ii' x g4 tLldf6
6. . . tLlf6
Of course, h4-hS can not be allowed - the h
7 0-0-0
pawn must be securely blocked . When
preparing an attack on one wing , don't
forget about necessa ry prophylaxis on the
opposite wing.
15 'ii' g 2 tLlh5
1 6 tLlge2 b8
1 7 tLlg3 tLl7f6
1 8 tLlxh5?
Wh ite is hoping to l ift the blockade, by
transferring his other knight to g3, but he
does not have time for this. 1 9 ..tgS was
stronger, with the idea of 20 .1i.xf6 tLlxf6 2 1
h S . And 1 9 e S dxeS 2 0 'ii'xc6 also came into
consideration .
18 . . . tLlxh5
7. . . b5?! 1 9 tLle2 c5
8 .1i.xb5! 0-0 20 c3 "a5
8 . . . cxbS? 9 eS is hopeless for Black. 21 ..t b3 .1i.a6
1 44 Attacks with opposite-sided Castl ing

22 lId2 31 e2
Wh ite wa nts to consolidate, by playing f3,
'it'f1 and g2 . But during this time I am able
to blow up the enemy centre.
31 . . . 'it'e4+ !
32 f3
If 32 lId3, then 32 . . . xd4 ! , while if 32 e1
there follows 32 . . . 'ifb5! (intending si mply to
advance the a-pawn), and 33 'it'f1 ? 'ifb 1 +
followed by 34 . . .'it'xe4 is bad for Wh ite.
32 . . . d5
33 e5 xe5!
34 'iff1 f6
35 'ifxe4 lIxe4
How should Black continue the attack? 36 e2?
To me it seemed d ubious to go chasing the 36 ': a 1 was essential , with the idea of
a2-pawn : 22 . . . c4 23 c2 'it'xa2 24 b 1 doubling rooks on the 7th rank as soon as
(th is is why t h e opponent played 2 2 lI d 2 - possible. After the move in the game Black
the b2-pawn is now defended) 24 . . . 'it'a 1 25 wins without any d ifficulty.
ttJg3. Wh ite h imself is threatening to lau nch 36 . . . ttJg7 37 lIa1 ttJf5 38 lIxa7 xd4! 39
an attack, for example, after 25 . . . ttJxg3 26 lIa8+ g7 40 f4 e3 (40 . . . xf2 ! ? 41
fxg3! followed by h4-h5 . However, if I had n 't e5+ f6) 41 lId3 lIxf4 42 J:txe3 ttJxh4 43
been too lazy to calculate fully the variation ];t e7 g5 (the fu rther play proceeds in
25 . . . l:tb3! 26 ttJxh5 ':xc3+ 27 bxc3 l;lb8 28 accordance with a well-known endgame
d 1 l:txb 1 + 29 e2 ':xh 1 30 ttJxg7 xg7, principle formulated by N imzowitsch : 'the
lead ing to a great advantage for Black, I col lective advance') 44 lIaa7 g6 45 l;ld7
could have gone in for it. ttJf5 46 lIab7 f6 47 ':b8 ttJd4+ 48 e3 f5
22 . . . xe 2 1 ? 49 lIf8 e5 50 ':xd5 l:tf3+ 51 d2 l:1xf2+ 52
I preferred not t o block l ines on the e3 ':f3+ 53 d2 e4 54 11d7 g4 55 :g8
q ueens ide, but on the contrary, to open f5 56 ':e8 ':a3 57 ':de7 ttJf3+ White
them immediately. resigned.
23 ':xe2 exd4
24 exd4 lIxb31 Pawns attacking the enemy king position
Such sacrifices do not requ i re any calcula can not themselves g ive mate. The aim of a
pawn storm is to open lines for the
tion . It is clear that now only Black has
pieces. Mainly for the queen and rooks,
winning chances.
although it is not uncommon for an
25 axb3 'it'a 1 +
important role in the attack to be played
26 d2 'it'xb2+ by the minor pieces.
27 e1 'it'a 1 +
I n the fol lowing game the main hero was my
28 d2 'it'b2+ dark-square bishop. After occupying the
29 e1 'it'xb3 long d iagonal as early as the second move,
30 :td2 :te8 it su bsequently d id not i n fact move from its
Attacks with opposite-sided Castl ing ttJ 1 45

post. But its influence on the development of 1 3 . . . tLle4! 1 4 tLle2 fS , and Black seized the
events was enormous . in itiative.
Alanakian-Dvoretsky (Moscow 1 97 1 ) : 1 2
.i.xc6!? bxc6 1 3 a4? ! ( 1 3 'ii'f4) 1 3 . . . aS 1 4
Dvoretsky - Khramtsov 'ii'fS .l:.fe8 ( 1 4 . . . 'ii'd 6) 1 S 1:[ae 1 1:[ad8 1 6 1:[e2
Moscow 1 970 cS 1 7 'ii'x eS 1:[xeS 1 8 1:[ef2 c6 1 9 1:[fS 1:[de8
Simagin-Larsen Opening 20 h3 1:[8e7 21 1:[xeS 1:[xeS 22 1:[f4 f8 23
1 b3 e5 g4 e7 with a good endgame for Black.
2 .i. b2 tLlc6 7. . . .i.xf3
3 e3 d5 8 'ii'xf3 tLlf6
4 .i.b5 .i.d6 I n correct is 8 . . . e4? 9 'ii'g 3 f6 1 0 tLlc3 .f7 1 1
5 f4 tLlxd S ! , as i n the game Dvoretsky-Makarov
Of cou rse, the opponent can not be allowed (Moscow 1 970). There followed : 1 1 . . . 0-0-0
to set up a powerful pawn centre u n h i n ( 1 1 . . . xdS 1 2 .i.c4 'ii' h S 1 3 . xg7 or
dered . The attack on the other side by c2- 1 2 . . .gS 1 3 'ii' x gS fxgS 14 .i.xg7) 1 2 .i.c4
c4 looks steadier, but I several times 'ii'd 7 1 3 0-0-0 a6 1 4 .g4 tLlaS? 1 S tLlb6+!
successfully employed the more risky move Black resigned .
in the game.
5. . . 'ii'e 7
If S . . .f6 Wh ite was i ntending 6 'ii' h S+!?
(provoking a weakening of the a 1 -h8
diagonal) 6 . . . g6 7 'ii' h 4.
6 tLlf3 .i.g4
In the event of 6 .. .f6 ! ? it is dangerous to win
a pawn : 7 fxeS?! fxeS 8 xc6+ (8 tLlxeS?
.i.xeS 9 .i.xc6+ d8! is completely bad )
8 . . . bxc6 9 tLlxeS . h4+ (9 . . . .i.xeS? 1 0
'ii' h S+) 1 0 g3 'ii' h 3 ( 1 0 e4 is worse in
. . .

view of 11 O-O ! ) 1 1 'ili'e2 tLlf6 with a


dangerous attack for Black. But after the
correct 7 0-0 the enemy centre remains
vulnerable. What do you think, how original is this
position? I was staggered to d iscover that
7 h3
many decades earl ier it was analysed by the
I also had occasion to play this position with well-known theoretician Vsevolod Rauzer in
Black. My opponents invariably chose 7 his notes to his game (with Black) agai nst
fxeS .i.xeS 8 iLxeS iLxf3 9 'ili'xf3 'ili'xeS 1 0 Vyacheslav Ragozin, played in the 1 936
tLlc3 tLlf6 1 1 0-0 0-0 . Objectively the young masters tournament in Len ingrad . He
chances here are roughly eq ual, but this pointed out that after 9 0-0 exf4 1 0 exf4 0-0
does not mean that a draw is inevitable. The 1 1 .i.xc6 bxc6 1 2 tLlc3 .:tfe8 Black has the
player who acts more pu rposefully is the better chances . However, by playing 1 0
one who will be successfu l . xf6! (instead of 1 0 exf4? ! ) 1 0 . . .'ii'xf6 1 1
Semeniuk-Dvoretsky (Sverdlovsk 1 987): tLlc3 , Wh ite gains the advantage, so Black
12 'ilih3 tLle7 !? 1 3 .i.d3?! ( 1 3 ':f4! is better) does better to choose 9 . . . 0-0.
1 46 \t> Attacks with opposite-sided Castl ing

Rauzer recommended 9 g3 with approxi is den ied this possibil ity. As a result the
mate equal ity. I think that 9 lDc3!? also bishop on b2 becomes fearfully strong.
deserves consideration. 15 i.. x e6 "xe6
9 f5?! 1 6 lDe2 lDh5
An over-comm itting conti nuatio n . I sensed
how risky it was, but I wanted to engage my
opponent i n a complicated and u nusual
fight. I n the end my idea was justified .
9 . . . e4
1 0 "f2 h5
Wh ite wants to castle on the q ueenside, and
therefore the most u npleasant move for him
was 1 0 . . . lD h S ! , forcing kingside castl i n g . It
is true that after 11 0-0 it is not possible to
clamp the kingside by 1 1 . . . lDg3 because of
the reply 1 2 f6! , but 1 1 . . . .. gS or 1 1 . . . 0-0 , for
example, is not bad .
1 1 lDe3
I also thought about 1 1 g3 h4 1 2 gxh4, but I
decided that it was too provocative. 17 f6 !
11 . . . h4 Weaker was 1 7 lDxg3 lDxg3 1 8 "f4 f6 ! (but
12 0-0-0 i.. g 3?! not 1 8 . . . lDxh 1 ? 1 9 i-xg7 or 1 9 f6 with
13 "f1 0-0 deadly mati ng th reats).
14 b1 17 . . . g6
What would you now have played as Black? I did not even consider the acceptance of
To answer this question, it is useful to the pawn sacrifice. After 1 7 . . . lDxf6 White
consider the point of my last move . would have had a pleasant choice between
1 8 lDxg3 hxg3 1 9 "f4 (or 1 9 "fS), 1 8 lDd4
I was not averse to the exchange of several
pieces: 1 4 i.. xc6 bxc6 1 S lDe2 i.. e S 1 6 (with the idea of 1 9 lDfS ) and, fi nally, the
i.. x eS "xeS 1 7 'iWf4 . After 1 7 . . .'ifxf4 1 8 primitive 1 8 i.. xf6 'ifxf6 1 9 'ii'xf6 gxf6 20
lDxf4 Wh ite has the better endgame. U nfor lDxg3 hxg3 2 1 l::t hf1 . If this move had been
tu nately, this idea did not work because of made, then I would have had to choose, but
the mate on a 1 , but now Black has to reckon there was no point in spending time before
with it. hand .
I recommend the attacking but also prophy 1 8 lDxg3 hxg3
lactic reply 1 4 . . . a S ! , wh ich disru pts Wh ite's Of course, not 1 8 . . . lDxg3 1 9 'iff4 lDfS 20
pla n . If 1 S i.. x c6?! bxc6 1 6 lDe2 i.. e S 1 7 "gS followed by 2 1 1:[df1 or 2 1 g4 hxg3 22
i.. x eS WxeS 1 8 Wf4 there follows 1 8 . . . We7 ! h4.
1 9 a 4 ( 1 9 "xh4 a 4 ) 1 9 . . . lifb8 , a n d things 19 "e2 1
become u npleasant for the wh ite king . The advantage is with Wh ite, but it is not so
14 . .. 'ii'e 5? easy to breach the opponent's defences.
A serious positional mistake . Black should On the kingside he has erected someth ing
always have been able to meet lDe2 with resembling a fortress. If l:. hf1 with the idea
the exchange of bishops on eS, but now he of l:.fS ! , Black repl ies . . . "e6 , and the rook
Attacks with opposite-sided Castling CD 1 47

has no invasion square on the f-file. The 23 . . . gxh5


queen can be played via g4 to g5, threaten 24 'ifxh5+ 'iiti> e 7
ing to invade at h6, but the threat will be
24 . . . 'iiti> g 7 25 1:[f1 was completely bad .
parried by . . . 'iiti> h 7 . Wh ite would have had to
24 . . . 'iiti> e 6 was more tenacious, after which I
open a 'second front' , by underm i n i ng the
was intending 25 'ifg4+ (25 1:[f1 I?) 25 . . .f5 25
opponent's centre at a n appropriate mo 'ifg6+ 'iiti> d 7 26 'ifg7+ 'iiti> e 6 27 h4! , and Black
ment with d2-d3.
has no way of opposing the advance of the
19 . . . liJxf6? rook's pawn - the entire board is raked by
Alexander Khramtsov made things much the bishop.
easier for me. G reed in such situations is 25 i.. a3+ 'iiti> d 7
completely inappropriate.
Here too 25 . . . 'iiti> e 6 was more tenacious.
20 1:[df1 liJh5
26 'ifh7+ 'iiti> e 6
21 1:[f51
27 'ife7+ 'iiti>f5
Not 2 1 "g4 f5 22 'ifg5 1:[f7. Now the threat
is 22 1:[xh5 gxh5 23 'ifxh5 f6 24 'ifg6+ 'iiti> h 8 28 'ifg7 1
25 l:[f1 . If 2 1 . . . liJg7, then both 22 1:[g5 and More accu rate t h a n 28 1:[f1 + 'iiti> g 6 29 1:[f4 f5.
22 ltf6 'ifd7 23 1:[hf1 are strong . 28 . . . 1:[fe8
21 . . . f6 29 1:[f1 +
22 'ifg4 'iiti>f7 Black resigned.
23 1:[xh5
An obvious exchange sacrifice, from which I n conclusion I invite you to practise finding
it is hard to refrai n . But 23 1:[hf1 would the best conti nuations in some positions
possibly have decided matters more simply. with opposite-sided castl ing.

Exercises

1 . Wh ite to move 2. Black to move


1 48 Attacks with opposite-sided Castl ing

3. Wh ite to move 4. Black to move

5. Wh ite to move 6. Black to move

7. Wh ite to move 8. Black to move


Attacks with opposite-sided Castling 'ZJ 1 49

Sol uti ons

1 . Fori ntos-Zedek ( I mperia 1 99 1 ) . set-ups, a strange pictu re is sometimes


It is important for Wh ite t o open t h e h-file for observed . After castl ing long, Wh ite then
his attack, but if 1 7 h4? there follows mounts on offensive on the q ueenside, by
1 7 . . . g41 . The target should first be fixed. advancing the pawns in front of his king,
1 7 g41 xg4 while Black attacks on the kingside. (A
classic example of such strategy is the
18 h41
game Kotov-Szabo, played in the 1 953
Black has no defence. If 1 8 . . . xf3, then 1 9 Candidates Tournament in Switzerland).
hxg5+ xh 1 20 'ii' h 2+ (of course, 20 ':xh 1 + The actions of the two sides, which at
also mates). first sight seem paradoxical, are easy to
No better is 1 8 . . . g6 1 9 xg6 ( 1 9 hxg5+ explain. Each conducts an offensive on
g7 20 lDf6 is also good ) 1 9 . . . xf3 the wing where he is stronger: where he
( 1 9 . . . fxg6 20 'ifxg6; 1 9 . . . g7 20 xf7! l::txf7 controls more space (which is deter
2 1 lDxg5) 20 hxg5+ g8 2 1 l::t h 8+ (2 1 'ifh2) mined by the central pawn structure) and
21 . . . g7 22 l::t h7+ g8 23 xf7+ l::t xf7 24 has more pieces.
'ii'g 6+. I n the position offered to you Wh ite has a
I n the game Black resigned after first clear plan: to advance his pawns to a4 and
playing 1 8 . . . f5 1 9 hxg5+ g8 20 g6 xe3+ b5 and then captu re the d3-pawn. But first
21 lDxe3 l::tf6 . he must take away the c4-square from the
enemy bishop. This problem can be solved
2. Khol mov-Naumkin (Moscow Champi only by the king , which boldly advances.
onsh ip 1 983). 23 b31
12 . . . h61 The advance of the wh ite pawns can not be
Black intends . . . g7-g5, not only beg i n n ing prevented . Alexander Bel iavsky's desperate
an attack on the king , but also preparing to attempts to complicate the play proved
hit the central e5-pawn by . . . lDg6, . . . g7, unsuccessfu l .
and the n , if necessary, . . . g5-g4 or . . . h6- 2 3 . . . l::t c 8 2 4 a 4 c 6 25 dxc6 l::t x c6 2 6 b 5 d 5
h5. It is impossible to prevent this pla n . 27 lDxd5 f8 28 'ii'x d3 'ii'd 6 2 9 b21 ':b8
1 3 h4 g51 30 lIhc1 , and Wh ite won .
14 hxg5 lDg6
Black has gai ned a great advantage. 4. Lyublinsky-Simagin (Moscow 1 939).
15 g3 e71 1 6 d3 :dg8 1 7 c4 dxc4 1 8 12 . . . h81
xc4 hxg5 1 9 'ii' b 3 lDf4 20 l::t fd 1 l::t g 61 2 1 An exchange sacrifice typical of such struc
xf4 gxf4 2 2 a 4 l::t hg8 23 f1 b 8 ! 2 4 a 5 tures, which was employed many times by
c 6 25 a6 l::t x g2 26 xe6 h4 27 l::t d 2 Vlad imir Simag i n . Black needs the bishop
xf3 28 xg8 l::t g 1 +! Wh ite resigned . far more than his passive rook. It is useful
both for the defence of his own king, and for
3. A. Petrosian-Bel iavsky (Riga 1 973). the attack on the opponent's king .
I n the Samisch Variation of the King's I ndian 1 3 xf8 'ifxf8 1 4 a3?
Defence, as wel l as certain other open ing Without extreme necessity you should
1 50 Attacks with opposite-sided Castl ing

not advance pawns where you are weaker. 6. Pchiolkin-Tolonen (Russian Corre
1 4 liJa4 or 1 4 h5 was better. spondence Championship 1 980/83).
1 4 ... Ub8 One of the most difficult problems in
The rook coord inates excellently with the chess is how to correctly combine attack
bishop - the two pieces exert terrible and defence, avoiding both excessive
pressure on the b2-point. Wh ite's position is caution, leading to passivity, and ultra
already d ifficult. aggression, bordering on recklessness.

1 5 .1i.d3 c5! 16 exd5?! liJxd5 1 7 liJa4 .1i.d7! 23 . . . g6!


18 'iWa5 .1i.xa4 19 'iVxa4 'ii' h 6+ 20 f4 liJxf4 By defending against g5-g6, Black retains
21 :td2 .1i.xb2+ 22 'it'd1 .1i.c3 23 Uf2 .l:!.b1 an excellent position. The opponent has to
mate reckon very seriously with the th reats of
24 . . . d5 and 24 . . . liJb5.
5. Ochoa-Vera ( Havana 1 98 1 ). The impatient attempt to carry out one of
17 b5! these th reats immed iately allows Wh ite to
mount a dangerous attack on the king,
By sacrificing a pawn , and then also a piece,
which outweighs Black's activity on the
Wh ite destroys the opponent's defences on
queenside.
the q ueenside and obtains a decisive attack
on the king . 23 . . . liJb5? 24 g6! liJc3+ (24 . . . liJa3+ 25 'it>a1
liJxc4 26 'iWh4, threatening to g ive mate or
17 . . . hxg3
captu re with the q ueen on c4) 25 'it'c1 d5 26
1 8 hxg3 cxb5 'iVh4 fxg6 27 hxg6 h6 (Black has a difficult
1 9 liJxb5! axb5 position after 27 . . . hxg6 28 Uxd5 liJxd5 29
20 a6 .l:!.d 1 or 27 . . . 'iVa3+ 28 'it'd2 hxg6 29 'it'e1 ) 28
.1i.xh6 gxh6 (28 . . . 'iVa3+ 29 'it'd2 gxh6 30 g7
20 . . . bxa6 2 1 l:!.xa6 .1i.xg3!? 22 l:!.a8+ liJb8 23
.1i.c5 3 1 'iWxh6) 29 .:txd5! (less good is 29 g7
'iWg4+ f5 24 'iVxg3 'ii'x g3 25 fxg3 , and ,
.1i.c5, threatening a check on e3) 29 . . . liJxd5
despite the exchange of queens, Wh ite's
30 g7 %:tb6 (31 'iYxh6 was threatened) 31
attack continues.
gxf8'iW+ 'it'xf8 32 'iVg3 with decisive th reats.
20 . . . liJe5 21 axb7+ (2 1 'iVd5 ! ? is also not
23 . . . d5? 24 g6 ! !
bad ) 21 . . . 'it'd7 22 Ua6! (weaker is 22 f4?
liJd3 ) 22 . . . liJd3 23 'iVg4+ f5 24 'iVd4 liJxe1 I n the game White d id not risk the piece
25 .1i.f4 and wins. sacrifice and he chose 24 .1i.d3? There
followed 24 . . . g6 25 l:tg3 liJb5 26 l:!.h3 'iVa3? !
20 ... liJc5 2 1 axb7+ (2 1 a7!? 'it'd7 22 .1i.xc5
(26 . . . liJc3+ 27 'it'd2 d4 would have won ) 27
'iVxc5 23 .1i.xb7) 21 ... 'it'd7 22 'iVg4+ ! f5?
.1i.xb5 axb5 28 c3 bxc3 29 llxd5 l:td8 30
22 . . . liJe6 was far more tenacious. .l:!.xd8 .l:[xd8 3 1 .1i.c1 'iVa6? (31 . . . 'ii'a 5!,
23 'ii' b4 liJe4 24 'ii'x b5+ 'it'e6 25 :a6 (25 i ntend ing 32 ... .1i.a3, was correct - it is
.1i.d4 ! ? ) 25 ... 'iWxc3 26 l:!.c1 ! 'iWe5 27 'ilt'xe5+ important that the c3-pawn is defended) 32
'it'xe5 28 .1i.f4+ 'it'd5 29 l:!.d1 + (another way f4! %:tc8 33 'ilt'c2 exf4 34 l:!.xc3 with roughly
to the goal was 29 .l:!.xd6+ l:txd6 30 .1i.xd6 equal play.
followed by nc8) 29 . . . 'it'c5 30 .1i.e3+ 'it'b4
24 ... dxc4
31 .l:!.b6+ Black resigned .
If 24 . . . fxg6 25 hxg6 hxg6, then Wh ite should
not play 26 .l:!.xd5?! liJxd5 27 .l:!.d 1 in view of
27 . . . .l:!.b5! 28 .1i. b6 (28 .:txd5 .l:!.xd5 29 .1i.b6
Attacks with opposite-sided Castl ing ttJ 1 51

'ifb5! ) 28 . . ...xb6 29 'ifxb6 l:1xb6 30 .1t.xd5+ his extra pawn . Let us see how the game
h7 31 .1t.xa8 with a probable draw. 26 concluded .
'iVh4! l:td8 (there is noth ing better) 27 .1t.d3 31 . . . .1t.c5? 32 %:tc1 "f7 33 b4 .1t.d4 34 .1t. b3
is much stronger, with an irresistible attack. 'iVe7 35 .1t.f4 b5
25 'ii' h 4 fxg6 26 fxg6 h6 Black's lot is not eased by 35 . . . d2 36 .1t.xd2
26 . . . hxg6 27 "xc4+ h8 28 ':'g3 is no .1t.xf2+ (36 . . . tDd3 37 l:tc6) 37 h 1 ! .1t.d4
better. (37 . . . %:txd2 38 Was + ! ) 38 .1t.f4 .
27 .1t.xh61 gxh6 28 g7, and Black has no 3 6 "c6 (36 .1t.e6! was even stronger,
defence. preparing the invasion of the q ueen at c6)
36 . . . 'ii'd 7 37 "e4 .1t. b6 38 %:td1 :e8 39
J:[xd3 "c8 40 .1t.f7 %:te7 41 .1t.xh5 "c4 42
7. Simagi n-Petrosian (Moscow 1 956).
'ii' h 7! c7 43 l:td2 "xb4 44 'ifg8 1:td7 45
17 h41 l:tc2+
An example of skilful prophylaxis with It was possible to win the queen by 45
opposite-sided castling! 'This move seems xe5+ fxe5 46 l::t x d7+ xd7 47 .1t.e8+ d8
risky, but in this way White parries Black's 48 xb5+ e7 49 "xg7+ e6 50 .1t. c4+!
attack on the kingside ' (Simag i n ) . It is Wxc4 5 1 "g8+ .
important to deny the opponent the possibil
4 5 . . ..1t. c 5 4 6 "a8 d 6 4 7 .:td2+ .1t.d4 48
ity of . . . h5-h4 or . . . g7-g5. For the sake of
.1t.e3 e6 49 "e8+ f5 50 g4+ e4 51
th is, one can even violate the principle,
"a8+ l:td5 52 %:txd4+ Black resigned.
mentioned in the notes to another game by
Si mag in ( Exercise 4). Wh ite , who has two By playing 3 1 . . . l:tc8, Black would have
strong bishops, has the better chances. The prevented the enemy rook from occupying
th reat is c3-c4-c5 . I ncidentally, the immedi the c-file (32 l:tc1 ? 'ifxc1 + 33 .1t.xc1 ':xc1 +
ate 1 7 c4 g5! 1 8 c5 .1t.e 7 1 9 .1t.xf4 gxf4 34 h2 tDg4+ 35 h3 l:th 1 + ! 36 'ii'x h 1
( 1 9 . . ...xf4 ! ? ) 20 tDe2 is sufficient only for tDxf2+), but after 32 .1t.b5!? his position
equal ity. would have remained d ifficult in view of the
insecure position of his king and the lack of
1 7 . . . tD4d5 1 8 tDe4 tDxe4 1 9 .1t.xe4 tDf6 20
counterplay.
.1t.c2 tDg4 21 g3 .l:.he8 22 a51 e5 23 .1t.g5! f6
24 .1t.d2 (th reatening 25 .1t.g6) 24 . . . exd4 25 Only if you sense just how strateg ically
cxd4 ltxe1 + 26 ':'xe1 c5! 27 a6! cxd4 28 dangerous Black's position is can you
.1t.a5?1 decide on the compl ications beg i n ning with
31 . . . b5, which was suggested after the
A tempting move, but not the best. 28 .1t.e4!
game by Tigran Petrosian. After all, in this
bxa6 29 "d3 tDe5 (29 . . . 'ifb6 30 .1t.g2! with a
case you have to reckon with the seemingly
decisive attack) 30 "xa6 'ifb6 3 1 'ii'a4 was
powerful .1t.a5. However, in winning the
stronger.
exchange, Wh ite l ifts the blockade on the
28 . . . b6 29 .1t.d2 tDe5 30 'it'g2 d3 31 .1t.a4 d3-pawn .
The position of the next exercise has been 31 . . . b5! 1
reached.
32 .1t.a5
The following variation is i nteresting : 32
8. Simagin-Petrosian (Moscow 1 956). .1t.xb5 "b6 33 .1t.a5 'ii'x b5 34 .1t.xd8 d2! (not
It only remains for Wh ite to play 32 1:tc1 , and 34 . . . tDc6? 35 %:te8; 34 . . . 'ifxa6 35 'iVe4 leads
th ings will be bad for the opponent, despite to an unclear position) 35 l:td 1 'ii' b 3 36
1 52 \t> Attacks with opposite-sided Castling

1:[xd2! (this sacrifice is forced: 36 'iVf1 'iVd5 bxa4 ! 35 1:[e8 lZ'lxd8 36 1:[xd8+ <j;c7 or 35
is bad for Wh ite, and he has a hopeless 1:[e6 e5.
endgame after 36 'iVb7+ 'iVxb7 37 axb7 b4 33 . . . lZ'lf3+
with the threats of 38 . . . lZ'lc6 and 38 . . . lZ'lc4) Weaker is 33 . . . bxa4 34 a5, when 34 . . . lZ'lf3+?
38 . . . lZ'lf3+ 39 'iVxf3 'iVxf3 40 1:[xd6 'iVf5! (the no longer works because of 35 'iVxf3!.
only defence, but a sufficient one, against
34 <j;h1 bxa4
the threat of 4 1 a5), and White faces a
fight for a draw. 35 a5 e5
The active placing of Black's pieces and his
32 . . . 'iVc61
strong passed pawn compensate for the
33 xd8 sacrificed exchange. A good example of a
Wh ite loses after 33 'iVxc6? lZ'lxc6 34 xd8 timely cou nterattack.
CD 1 53

PART V

Defe n ce

I gor Belov

P ractical Exercises i n the Taki n g of d ifficu lt


Decisions

- Vau l i n
M demand a deep penetration into the
any moves that we make d o not Beloy
Katowice 1 99 1
position . It is sufficient to make use of
standard eval uation considerations and to
check a few variations. Any yet in nearly
every game there invariably comes a
turning-point (sometimes several), when
the solution is by no means obvious, and
on it depends the entire course and
perhaps the result of the subsequent
play. It is a few such situations, which
occu rred in my games, that I wish to offer for
your attention. Try in a restricted time to
solve those problems which I encountered ,
and then we will compare our conclusions.
We will beg in with a few relatively ( only
relatively! ) simple examples, and conclude
with some that are very complicated , al most Exercise 1 . We have a position with a n
irrational . un usual material balance. Who is playing
for a win? How should Wh ite continue?

Rook, knight and pawn are roughly equ iva


lent to a q ueen , but in the positional sense
Black is stronger. Wh ite has nothing to aim
at, whereas the opponent is threatening to
create pressure on the weak d4-pawn . Here
White should not try to be too ambitious.
1 54 Practical Exercises in the Taking of d ifficult Decisions

The clearest way was pointed out by I Iya Black makes an escape square for his king,
Makariev. and then plays . . . b7-b6 and . . J:tb7.
1 .lixd5! cxd5 The idea of I n na Gaponenko seems ques
If 1 . . . exd5, then 2 'ikc3 a6 3 l:te 1 . tionable: 1 a6 bxa6 2 'iVc4. Black's rook
2 ':d1 :fc8 immediately comes into play on the b-file,
and his bishop endeavours to get to the d4-
3 I::!. d 3!
pawn and attack f2 .
The rook is head ing for c3. It may even be
U nfortunately, at the board I too failed to
possible to seize the in itiative . The side
figure out the position. I real ised that I
with the queen should aim for ex
should aim for exchanges and I stud ied the
changes/ The power of the queen is
move 1 .lixd5. But I did not see the rook
easier to exploit, when it is opposed by
manoeuvre to c3, and considered only
fewer pieces - the chances of breaking
1 . . . cxd5 2 h4 .lif6 3 'ii'e 3. Then 3 .. .l:Ibc8!
into the opponent's position are im
(with the threat of 4 . . . l:tc4) is strong, and if 4
proved.
Itc1 Black has 4 . . . .lixd4 ! .
[Instead of 2. . . I::!. fcB Black does better to play
1 l:ta4? .lie7 !
2. . . b6, hoping for 3 a6?! b5 followed by
4 . . . 1J.b6 or 4 . . . b4. But after 3 axb6 I::!. xb6 4 O f course, Black prevents 2 .lixd5 and 3
'it' c3 followed by 5 .l:ta 1 White would seem to 1:tb4 . Even now it was not yet too late to
be out of danger - Dvoretsky.] captu re on d5, but I decided fi rst to occu py
Peter Svidler was intend ing to bring up his the c-file with my rook.
rook along another route : 2 'it'b4?! l:tfc8 2 11c4? ! liJc7!
(2 . . J:tfe8 3 h4 .lie7 4 'iVd2 and 5 ttc1 ) 3 .l:!.a3. Alas, I completely overlooked this simple
This is too intricate. Try to put your plans into move. With the retreat of the knight, White's
effect in the simplest and most rel iable way, position immed iately becomes d ifficult. The
otherwise you risk making some blunder, as opponent wants to make a concerted attack
i n fact occurred with Peter: 3 . . . .t!.c4! 4 'ilYd6? on d4. He has more pieces than me, and
l:td8. therefore the pawn essentially ca nnot be
[If it is clearly realised that the rook must defended . All I can hope for are chance
definitely aim for the c-fi/e, it is even tactical opportu nities.
possible to consider a pawn sacrifice: 2 h4 3 'it'd3 l:tfd8 4 .lie4 h6 5 'iVe3 a6!
.lif6 3 1:1 c 1 ! .lixd4 4 l:tc7. But after 4 . . . b6 or
Before the knight is moved to b5, the a5-a6
4 . . . b5 White still has problems, so that the
th rust must be prevented .
manoeuvre l:ta 1-d1-d3-c3 is more con
vincing, in my opinion - Dvoretsky.] 6 .lic2 .li f6 7 .lia4 liJb5 8 .lixb5 axb5 9 l:tb4
All the remaining plans are weaker. For J:ta8 1 0 'it'f4 lId7 1 1 'it>g2 lIad8 , and Black
example, Maxim Boguslavsky suggested 1 won .
'ikc4 with the threat of 2 a6. Black replies
1 . . . a6, and what now? Exchange on d5?
This is illogical - after all, Black recaptu res
with gain of tempo. Vasya Emelin continued
the analysis : 2 .lixd5 cxd5 (2 . . . exd5 3 'it'b4
and 4 11e1 ) 3 'it'c7 .l:[fc8 4 'it'd6 .lif6 5 l:td 1 .
Of course, the wh ite queen is active , but the
exchange of rooks has had to be deferred.
Practical Exercises in the Taking of d ifficult Decisions CD 1 55

Kamshon kov - Belov squares (there was a choice: a7/b6 or a61


Podolsk 1 99 1 b5). But the opponent's pawns on f4 and h2
are 'wrongly' placed and are therefore
vul nerable. Nevertheless, the drawing ten
dencies of opposite-colour bishops are
very great, and in analysis I found a forced
draw for my opponent. But I was hoping that
he would not find it, since he did not know
my sealed move, and the saving path was
not altogether obvious.
Diana Darchiya suggests playing 1 .i. h 3 , in
order to prepa re f4-f5 . Correct! We will
either exchange pawns, or force . . . e6-e5,
but then the wh ite f-pawn will be on a
'correct' square, and the black e-pawn on an
'incorrect' square of the same colour as its
bishop. The chances of a successfu l block
This was the position a t the adjourn ment, ade will be improved . I am pleased to note
with Black having sealed . . . .i.a5-c7! , which that one half of the participa nts in the
came as a surprise to my opponent. competition assessed the position correctly
Exercise 2. Remember the positional prin and made the same choice.
ciples which apply in such situations, and I will show the variation that I fou n d .
with their help choose a plan of defence for 1 .i.h3! d8
White . 1 . . . ltd5 2 .i.xe6+ .
2 f5 e5
We have a position with opposite-colour
3 l:td2 1 ?
bishops. It would not be bad to exchange
rooks and transpose into a pure bishop T h e exchange o f rooks leads b y force to a
ending, but for the moment, alas, this is not draw. 3 .i.g2 is also not bad .
possible. 3. . . e7
When defending an endgame where the 4 ':xd6 xd6
opponent has a material advantage, you 5 .i.g2 f6
should exchange pawns. A usefu l ru le. 6 .i. b7 1
Are there any other general considerations?
It is important for Wh ite that on the
Many years ago I attended a lectu re by queenside too Black's pawns should be on
Dvoretsky on opposite-colour bishops, and squares of the colour of his bishop. See how
th is gave me a fi rm grasp of the main easy it is to make good moves, if you know
principles for playing such endings. One of this ru le.
the ru les formulated by him states: 'The
6. . . a5
stronger side should keep his pawns on
7 .i.c61 b4
squares of the colour of the opponent's
bishop; the weaker side should keep his 8 f3 !
on squares of the colour of his own The simplest. I also analysed the sharper
bishop '. Gu ided by this rule, in time-trouble continuation 8 .i.e4 g5 9 f3 h4 1 0 f6
I unhesitatingly placed my pawns on light h6 1 1 f7 h3 1 2 f5+ xh2 1 3 g4 -
1 56 \t> Practical Exercises in the Taking of d ifficult Decisions

however much I tried , I also failed to find a vated his forces.


win here. 1 J.. e 4? lId4 2 'it e3 (2 J.. b7!?) 2 . . . lIb4 3
S. . . 'it g 5 lIc3 h5
8 . . . 'itxf5 9 J.. e4+ and 1 0 J.. x h7. 3 . . . h6 is weaker - the stronger side should
9 'ite4 J.. c 7 not place his pawns on squares of the
1 0 J.. d 7 colour of his own bishop. Wh ite's position
The d raw has become obvious. has become d ifficult, because his pieces
are tied to the defence of his weak pawns.
The defensive plan which we have analysed Here is it appropriate to remember another
consistently carries out one of the ideas i mportant principle. If, apart from the
typical of endings with opposite-colour opposite-colour bishops, there are also
bishops, involving the correct deployment of other pieces on the board, on no account
the pawns . Of course, a practical game is should you remain passive - you must
not a study, and a position can sometimes seek counterplay, and fight for the initia
be approached in d ifferent ways. But in any tive at any price.
case accu racy is required , and by no mea ns 4 h3 J.. b6+ 5 'itf3 .i.d4 6 l:d3 'ite7 7 f5 e5 8
all methods of defending are equally good . J.. d 5 'itf6 9 J.. e6 a5 1 0 l:.d2 e4+ ! 1 1 'ite2 a4
For example, 1 'ite3 is tempting, depriving 1 2 bxa4 bxa4 1 3 l:tc2 a3 1 4 'itd 1 'ite5 1 5
the rook of the d4-square and intending after :e2 e 3 1 6 l:tc2 'ite4 1 7 'ite2 krb1 White
1 . . . J.. b 6+ 2 'ite2 lId4?! to conti nue 3 l:td2 or resigned.
3 l:tc6. But Black has the unpleasant reply
1 . . . l:td 1 (and if 2 J.. e4 , then 2 . . . :e1 +,
winning a second pawn ). Lempert - Beloy
Sash a Chernosvitov and I n n a Gaponenko Katowice 1 990
recommend 1 lIc5. What for? You want to
attack the h7-pawn? I will happily give it up,
provided I can get to the b3-pawn . I reply
1 . . . l:td4 2 lIh5 lIxf4 3 ':xh7+ 'itd6. Or if 2 f5,
then 2 . . . exf5 3 lIxf5 lIb4 4 :f7+ 'itd8. After 5
lIxh7 lIxb3 it is not easy to save the game,
but otherwise how does Wh ite avoid being
two pawns down (5 lIf3 J.. x h2; 5 J.. d 5 :h4).
[A good defensive idea was found by
grandmaster Evgeny Bareev: 1 Ir.c3! (with
the threat of 2 lIh3) 1 . . Jld4 2 J.. b 7!, for
example, 2. . . J.. xf4 3 J.. xa6 b4 4 :1d3 with a
draw. The same idea can be put into effect
slightly differently: 1 ':c5 r:1d4 2 J.. b 7!. In
connection with this I should like to remind
you of another principle in endings with Exercise 3. In whose favou r is this posi
opposite-colour bishops: attack the enemy tion? What would you play as Black?
pawns with your bishop - Dvoretsky.]
l Iakha Kadymova suggested 1 .i.e4. This is U nfortunately, two of you did not have time to
what my opponent played , but it did not turn reach any conclusion . Vova Baklan sug
out well - immediately Black greatly acti- gested a move which had not even occurred
Practi cal Exercises in the Taking of d ifficult Decisions ltJ 1 57

to me: 1 . . . e3. His idea looks very dubious. After a move by the bishop this assessment
White gains an obvious advantage, for will be completely correct: the two rooks
example, by 2 fxe3 lLle4 3 ':'xcB ':'xcB 4 .i.g2 combined with the dangerous passed d
or 3 . . . 'iVxcB 4 d6 'iVd7 5 .i.g2. pawn are stronger than the queen . Bad is
All the rest of you found the strongest 4 . . . 'iVd7? 5 gxf4! 'ii'x eB 6 d6 'ii'd 7 7 .i.c4
possibility. followed by B e6, and Wh ite wins.
1 . . . .i.h6! Emelin calculated fu rther than anyone - he
suggested 4 ... .i.xg3 ! . If 5 1:[e6, then 5 . . . 'ii' c5
If the strong bishop on f4 can be exchanged
or 5 . . . 'iff4 is possible. We must look at 6
without detriment to Black's position , he will
fxg3 'ifxg3+ 7 g2 - what happens here?
stand better.
7 . . . e3? is anti-positional: after B l::tf 1 the
But if you analysed only 2 ':'xcB lLlxcB 3
black pawns are blocked . But the recom
'iVc7 , this reply is only worth th ree points out
mendation 7 . . . 'iVxb3! is very interesti ng.
of five. The critical continuation is 2 'iVxd6!
Where should the rook move to? Here
'iVxd6 3 llxc8 .
Black's chances are certainly not worse.
I have to admit that I only considered
7 . . f4! 1 8 1:[xe4 f3 9 11d2 fxg2. If 1 0 d6 there
.

follows 1 0 . . . 'iVh3 1 1 :txg2 'iVd3 1 2 ':'e7+


h6 1 3 ':'e6 'iVd4+ with a d raw. Wh ite can
try 1 0 1:[ee2!?, aiming to keep his rook on the
d-file, but this too is unclear.
We are now able to make an objective
assessment of the initial position . Black
stands worse , but after 1 . . . .i.h6! he can
hope to save the game.
Let us now see how the game proceeded .
My opponent, al most without th inking, ex
changed the rooks.
2 l:txc81! lLlxc8
In the event of 3 . . :ii'fB?! 4 ':'c7+ hB 3 'iVc7 l:td8
(4 . . .1le7 5 d6 ':'xc7 6 dxc7 .i.xf4 7 l:tdB) 5 3 . . . :e7 was also not bad .
.i.xh6 'iVxh6 6 d6 Wh ite has the advantage.
4 ':'c1 .i.xf4
3 . . . .i.xf4! must be played .
5 'iVxf4? !
[There is also a third possibility: 3 . . . 'iVxf4!!.
In the endgame arising after 4 gxf4 I1xc8, on 5 'iVxd7+ l:txd7 6 l:txcB was safer, and in the
his next move Black captures on f4 and ending with opposite-colour bishops Wh ite
emerges a pawn up. Or if 4 ':'xe8 there would probably not have lost.
follows 4 . . . 'fif3! 5 'fle 7+ h8 6 ne 1 (the only 5. . . lLld6
move) 6. . e3 7 ': 1xe3 .i.xe3 8 ':'xe3 'iVxd5,
.
6 'ife5
and it is now White who has to fight for a
Here my opponent offered a draw, but I de
draw Dvoretsky.]
clined. Black has excellently-placed pawns ,
-

4 l:[xe8 restricting t h e enemy bishop, a n d the


Svidler reached this position in his calcula ' N imzowitsch knig ht' , blockading the passed
tions and he assessed it in favour of Wh ite. d-pawn , is very strong.
1 58 Practical Exercises in the Taking of d ifficult Decisions

6. . . %:teS Mainly it was suggested that 1 . . . d3!? should


7 Wd4 1:e7 be played , by some - without any analysis.
S 'ii' b 6 1:[f7 Many gave the variation 2 ':'xa4 ':xa4 3
i.. xa4 dxe2 4 'ii'x e2 'ii'a 8. Black wins back
9 i.. h3 We7
his pawn and gains the advantage.
1 0 'ii'c 5 'ii'e 5
Svidler analysed 2 It)f4 i.. c3 3 ':xc3 'ii'x a5.
Black has a clear advantage. U nfortu nately, After 4 ':'xa3 'ii'x b5 the e4-pawn is attacked .
in the su bsequent play I twice blundered [However, by continuing 5 ':'b3, White gains
and th rew away the win . a great advantage: 5. . . 'ii' c 6 6 It)d5 or
5. . . 'ii' a 6 6 'ii' b 4. This means that 2. . . i.. c3 is a
Yachmennik - Belov bad move. The sacrifice of two minor pieces
Smolensk 1 989 for a rook, suggested by Volodya Baklan, is
better: 2. . . i.. b 2!? 3 ':xa4 i.. xc 1 4 'ii'xc 1
':xa4 5 i.. xa4 'ii'a 5 followed by 6. . . i.. xe4.
But the strongest continuation was pointed
out by grandmaster Dolmatov: 2. . . i.. xe4! 3
':xa4 1:1xa4 4 i.. xa4 e5, and Black, at any
event, is not worse - Dvoretsky.]
Only Makariev considered 2 lt)g3 ! , but even
he stopped after 2 . . . i.. c 3! 3 ':xc3 'ii'xa5.
[ The pretty counter-stroke 3 .. J1xa2 would
have led to equality in the event of 4 'ii'xa2?
It)xc3 5 'ii' a 1 d2. Unfortunately, it is refuted
by the prosaic 4 ':'xa4! ':xd2 5 It)xd2 -

Dolmatov.]
In fact the variation should be conti nued : 4
l:txa3 Wxb5 5 'ii' x d3 'ii'x d3 6 ':'xd3 It)c5 7
Exercise 4. The opponent's last move l:td4 (7 :e3 lIa8 8 a3 :a4 9 It)d2 is also
i.. d 3-b5 set me a d ifficult problem : how to possible) 7 . . . :a8 8 .l:tc4 ':'a5 9 It)d4. After
save the pin ned knight on a4. You (just as I calculating this far, I real ised that I would be
had to d u ring the game) have to: a pawn down in a d ifficult position.
a) assess the position; [By playing 9 . . lt)xe4, Black regains the
.

b) fi nd various possibilities for Black and pawn. Then 1 0 It)xe4 (10 It)b3 :e5)
weigh u p the necessary variations; 1 0. . . i.. xe4 1 1 lt)b3 l:.e5 1 2 f4 1:1e6 13 l:tc8+
c) choose the most promising course. rt;h 7 1 4 lt)c5 ':'c6 1 5 %:txc6 i.. xc6 leads to a
drawn endgame. And in the event of 7 :e3
Opinions regarding the assessment varied : (instead of 7 r1d4) 7. . . 1:1d8! it is vel}' difficult
'White is better', 'Wh ite is worse' , 'equal ity' . for White to convert his material advantage
Nearly all of you poi nted out that 1 . . .:xf3 is - the opponent's pieces are really too
bad because of 2 l:1xa4 ! . [After 2 .. J lc3 3 active. It can be concluded that 1 . . . d3!?
It)xc3 dxc3 Black retains some positional would have given excellent saving chances
compensation for the lost exchange (strong - Yusupov.]
passed pawn on c3, and two bishops). He [On the other hand, by continuing 5 ::'xd3
can go in for this position, if nothing better is (instead of 5 'ikxd3 ?!) 5. . . lt)c5 6 'tJ.d4 (or 6
found - Dvoretsky.] :e3!?) 6. . : ikb 1 + 7 lt)e 1 !, White would retain
Practical Exercises in the Taking of d ifficult Decisions ttJ 1 59

his extra pawn, for which Black has no real convert his advantage. If 6 lDf3 or 6 lDb3
compensation (he loses after 7. ..liJxe4 ? 8 there would have followed 6 . . . a8, aiming
lDxe4 xe4 9 11b4). In addition, as grand to break through onto the 2nd rank with the
master Bareev pointed out, instead of the rook.
capture on d3 White also has the interposi [After 6 lDb3 :l.a8?! there is the simple reply
tion 5 11b3!, and after 5. . . a6 - not 6 'ilih6?! 7 lDc5. Black does better to try 6 . . . b2!?,
f6, but simply 6 1:!xd3 lDc5 7 :d8, retaining a having in mind the variation 7 :e 1 :a8 8
serious advantage. For example: 7. . . f6 (cap lDc5 ?! xg2! 9 <:J;xg2 a3 1 0 c6 :c8. In
turing on e4 loses a piece) 8 :Xxf8+ <:J;xf8 9 the event of 6 lDf3 :a8?! White has 7 c6 -
'ikd8+ <:J;f7 1 0 'ikh8 - Dvoretsky.] therefore it makes sense to play 6 . . . xf3,
Have we taken all the candidate moves i nto spoiling the opponent's pawn structure -

accou nt? Vad i m Zviagi ntsev mentioned Dvoretsky.]


(however, without any analysis) 1 . . . lDc3. 6 lDc6 :d2
After 2 llxa3 lDxb5 3 :b3 there is no 7 lDc3 xc6
compensation for the lost exchange.
8 xc6 d4
[For my part I should like to suggest one
9 lDe4 b2
more idea: 1 . . . xe4!? 2 :txa4 l::txa 4 3 Ji.xa4
1 0 :d1 b6
'iVa8 4 Ji.b3(d1) Ji.xf3 5 gxf3 "ilxf3. Objec
tively, the resulting position probably fa Threatening 1 1 . . .f5 .
vours White - his passed a-pawn may 1 1 :d2 ':xd2
become extremely dangerous. But for the 1 2 lDxd2
moment he has to worry about his broken
kingside and reckon with the threat of a
black pawn advance in the centre. To be
honest, this continuation seems to me to be
more promising than that which occurred in
the game - Dolmatov.]
After weig hing up the variations, I came to
the conclusion that after a normal develop
ment of events I would most probably lose.
Not wish ing to reconcile myself to such a
dismal fate, I continued my search ing. I n the
end I managed to find a surprising chance.
I n principle, Black's position does have
some pluses. For example, the two bishops
and a compact pawn chain . The idea of
creating a fortress occurred to me . . . I anticipated this position wel l in advance
1 . . . lba2 ! ? and judged it to be drawn . The only target
2 'ikxa2 'iVxa5 that Wh ite ca n attack is the f7-pawn . But I
couldn't imagine how two pieces would
3 'ilkxa4 'iVxa4
simultaneously be able to attack it - si nce
4 Ji.xa4 xe4
the approaches to it on the dark squares are
5 lDfxd4 d8 guarded by my king and bishop. Of course, I
For the sacrificed piece Black has only one was not fully confident of a successful
pawn . However, it is not so easy for Wh ite to outcome, but I think that from the practical
1 60 Practical Exercises in the Taking of d ifficult Decisions

point of view the decision was justified . I had change the character of the play, even
a clear impression of how I would be sacrificing material if necessary. The
outplayed with an extra pawn , whereas I opponent will most probably cope suc
could not imagine how Wh ite would win cessfully in a standard, technical posi
here . tion (say, with an extra pawn). It will be
12 . . . i.d4 far harder for him in a situation with an
1 3 f1 h5 unusual material balance - here the
1 4 e2 g7 probability of a mistake sharply in
creases.
1 5 f3 e6
I ncidentally, by fi nding this d ifficult and
We soon adjourned the game. Analysis
unexpected defensive idea , I not only saved
confirmed that my assessment of the posi
half a point, but also experienced an
tion was correct. I will show you a curious
enormous emotional lift, thanks to which I
episode which occu rred during the resump
began winning game after game.
tio n .

Mityaev - Belov
Moscow 1 989

Wh ite tried 1 f5! ? After t h e g a m e m y


opponent pointed o u t that even 1 . . . exf5 ! ? 2
tDxf7 f6 would not have left h i m any
chances of success. It has to be said that I was close to a wi n , but when play went into
the drawi ng tendencies in herent in oppo an endgame, I relaxed and completely
site-colour bishops are exceptionally great! forgot that I could be mated . I only woke up
The game conti nuation also led to a draw: in the position which I am offering for your
1 ... gxf5 2 i. b5 i.xg5! 3 hxg5 g6 4 f4 attention.
e5+! 5 xe5 xg5 6 iLd3 f6+ 7 e6 f4 8 Exercise 5. How to combat f2-f4 ? At fi rst I
iLe2 h4. thought that things were completely bad ,
What is the main conclusion that should be but then . . . See if you can find a way of
drawn from the example we have just defending .
analysed?
When defending a difficult position, you Emelin did not fi nd anyth ing better than
should consider the most improbable 1 .. .fS 2 f4 tDd7, but he rightly judged Black's
resources, trust yourself and boldly position to be d ifficult.
Practical Exercises in the Taking of d ifficult Decisions tLJ 161

A clever idea ( i n the spirit of the previous cluded that 1 . . . f8!? 2 lDe4 Iha2! would
game) was devised by Zviagintsev: 1 .. .f5 2 allow Black to defend successfully -

f4 f7!? 3 fxe5 xe5. But after 4 .i.e1 I Dvoretsky.]


doubt whether Black has real compensation One more possibil ity must be considered .
for the sacrificed piece. Besides, capturing
1 . . . g5!
the knight is not obligatory - 3 lDb5!? is also
strong . 2 iLxg5
Serezha Movsesian settled on 1 . . . f8. But But what next? Chernosvitov rightly poi nted
this is bad ! Nearly all of you gave the out that the exchanging combination 2 . . .
variation 2 lDe4 lDc4 3 l:tb3 with advantage :t:!xf2+ 3 xf2 lDg4+ 4 f3 lDxe3 5 i... x e3
to Wh ite, and Svidler conti nued it: 3 .. J:tb2 4 xc3 leads after 6 .i.f4 to a hopeless
l1xb2 xb2 5 f6 ! ' I don't know whether he endgame for Black.
saw the counter-stroke 5 . . . lDd2+, which , Most of you restricted yourselves to the
however, does not change the assessment modest 2 .. .f6 3 i... h4 f7 (3 . . . lDg6 is
of the position . completely bad : 4 lD b5! lDxh4 5 11e8+ f7 6
To me it seemed more natu ral to play not 2 lDxd6+ g6 7 :t:!g8+! iLg7 8 hxg7). How
lDe4 , but 2 lDb5 (why block the e-file?). should the resulting position be assessed?
However, after 2 . . . lDc4 3 i... e 7+ e8 In the opinion of Svidler and Makariev, it is
(3 . . . g8? 4 1:1e1 i... e5 5 i... f6! with the threat unclear, perhaps slightly more pleasant for
of 6 lDxd6!) 4 lDxd6+ lDxd6 5 i... x d6+ d7 6 Wh ite . Wel l , compared with me you are
f8 i... d 4 7 l:te7+ d8 8 ];t xf7 e8 9 1:If3 g reat optimists. I assessed it as hopeless.
':xa2 the outcome still remains unclear. Wh ite is a pawn up, and the bishop on h8
Apparently the knight move to e4 is never has no moves. Let us analyse it in concrete
theless stronger. terms : 4 lDe4 lDc4 (5 . . Jlxa2 6 lDxd6+ g6
[ To both knight moves Black replies 2. . J:!xa2! 7 f4 lDg4 8 l:te8 leads to the loss of a piece)
3 lDxd6 1:1a6 4 I!b3 lDd7. The knight retreat 5 11b3? f5 6 1:!b7+ g6 - here Black does
5 lDc4 would give an advantage, were it not indeed acquire excellent counter-chances.
for the double attack 5. . . 1:1a4!, while after 5 7 lDxd6 lDxd6 8 1:Ib6 h5 will not do for
lDe4 or 5 g3 White has almost nothing. Wh ite. But why should he go in for these
complications? I think that the reply 5 ':c3 ! ,
Instead of 4 1:1b3 White can try 4 f4!? 1:!xd6 5
suggested b y Dvoretsky, will become a cold
:a3! f5 6 1:Ia8+ f7 7 1:!xh8. Black plays
shower for you . After 5 . . . 1:!xc3 (5 . . . lDd2+ 6
7. . . lDg4, and after 8 11xh 7+ g8 9 11g 7+
e2) 6 lDxc3 you will probably regain the
h8 he is threatening 10 . . . lDxh6 or 10 . . .
pawn on h6. But how to hold the outside
lDe3+ and 1 1 . . . lDxd5 with equality. In the
passed a-pawn and at the same time not
event of 8 e2!? l:1xd5?! 9 1:Ixh 7+ g8 1 0
lose the d6-pawn?
l:tg 7+ h 8 1 1 i... g 5 White still has some
hopes, although the opponent can go into a Let's look more closely at Wh ite's main
rook endgame a pawn down by 8 . . . lDf6! 9 threat f2-f4 . This move may win the game,
xf6+ xf6 1 0 l:txh 7 1:1xd5 1 1 1:.c7 1:!d8 1 2 but in itself it is anti-positiona l . It seriously
l:txc5 :t:! h 8 1 3 Ilc6+ f7 1 4 l::t c 7+ f6 1 5 h 7 weakens the second rank, and reduces the
g5. Then neither 1 6 l::t c6+ g 7 1 7 fxg5 mobil ity of the wh ite bishop. Is it not possible
%:.e8+! 18 f3 xh 7 nor 16 fxg5+ xg5 1 7 somehow to exploit th is, and set up a
f3 g6 1 8 f4 1:Ia8 leaves White any real counter-attack against the opponent's king?
chances of success. 2 . . . f5 1 ?
If this analysis is correct, it can be con- 3 f4
1 62 Practical Exercises in the Taking of d ifficult Decisions

[As Bareev pointed out, 3 0,b5!? was also [It made sense to try 7. . . 0,g4!. The obvious
strong. But the attempt to exclude this reply 8 g3 ? would allow Black to put into
possibility by transposing moves: 1 . . . f5 2 f4 effect a well-known drawing mechanism
g5 did not work in view of 3 fxe5! gxh4 4 with rook + knight by 8 . . . lId2!! (it is
exd6 or 3. . . f4 4 11e2! - Dvoretsky.] inaccurate to begin with 8 . . . 0,h2? in view of
3 . . . 0,g4 9 fLe8! followed by lIe3). After 9 lie8 there
c;i;>f7 follows 9. . .1:1d 1 +! 1 0 c;i;>g2 lid2+ 1 1 c;i;>f1
4 11e8+
0,h2+! 1 2 c;i;>g 1 0,f3+ with perpetual check.
5 lIxh8 0,e3+
Only 8 c;i;>f1 !! 0,e3+ 9 c;i;> e 1 0,xg2+ 10 c;i;>f1
6 c;i;>g1
0,e3+ 1 1 c;i;>g 1 would have won. As a result
of this White gets rid of the g2-pawn, the
return of the knight to g4 no longer
threatens mate, and he gains an important
tempo to destroy the drawing mechanism -

Dvoretsky.]
7. . . c;i;>h5
8 0, b5?
Here the opponent spent some fifteen
minutes , but he was unable to come to the
correct decision. Apparently he very much
wanted to keep all his extra material. But if
he wanted to go after the d6-pawn , it would
have been more logical to do this a move
earl ier, without d riving the black king for
If Black takes the knight, the outcome is ward .
settled by the far-advanced h-pawn : 6 . . . I was most afraid of 8 :1g7 , si nce I d id not
':'xc3? 7 ':xh7+ c;i;>g6 8 lig7+ c;i;> h 5 9 h7. No see how I could counter the power of the h
better is 6 . . .0,g4? 7 ':xh7+ c;i;>g6 8 lig7+ pawn . For example: 8 . . . l:.xg2+ 9 c;i;> h 1 lIc2
c;i;>h5 9 0, d 1 ! lIc1 1 0 c;i;>f1 . Final ly, 5 . . . lIxc3 1 0 l:txh7 ':'xc3 1 1 ':'e7 c;i;>g4 1 2 h7 c;i;>g3
(instead of 5 . . . 0,e3+ ) is also hopeless: 6 ( 1 2 . . . c;i;>f3 1 3 ':'xe3+!) 1 3 h4+ ! .
':'xh7+ c;i;>g8 7 lig7+ c;i;>h8 8 l:td7 l:tc2 9
[Instead o f 1 1 . . . c;i;>g4 Black has the more
c;i;>e1 !? followed by lixd6.
tenacious 1 1 . . .11c 1 + 12 c;i;>h2 0,g4+ 13 c;i;>g3
6 . c;i;>g61
11c3+ 14 c;i;>g2 l:tc2+ 15 c;i;>f1 %1xa2 1 6 h 7
. .

Now the main strategic basis of the pawn lIa8, when the win still has to b e demon
sacrifice . . . g6-g5! becomes clear - it is strated.
important to g ive the black king air and
Black is not obliged to capture on g2.
enable it to break free.
8 . . . 0,g4!? suggests itself, for example: 9 g3
7 ':'g8+ (9 c;i;>f1 0,e3+) 9 . . . .1:.xc3 10 ':'xh 7 ':' c 1 + 1 1
I n the event of 7 0,b5!? I was intend ing to c;i;>g2 l:1c2+ 1 2 c;i;>f1 0,e3+ 13 c;i;>g 1 (if 1 3 c;i;>e 1
play 7 . . J:bg2+ 8 c;i;> h 1 l:txa2 9 0,xd6 0,xd5 the same reply follows with even greater
1 0 ':'g8+ c;i;> h 5 , and if 1 1 0,xf5?, then strength) 13 . . . c;i;>g4! 14 .1:. e 7 c;i;>f3 (threaten
1 1 . . . c;i;>g4! with cou nterplay, si nce the king ing 15 . . . l:t c 1 + 16 c;i;>h2 0,g4+) 15 ':'xe3+
breaks through to h3. U nfortunately, after 1 1 c;i;>xe3 16 h7 .1:.c 1 + 1 7 c;i;>g2 ':c2+ 18 c;i;>h3
d8! (Dol matov) or 1 1 f6! (Dvoretsky) lic1 with a draw. White nevertheless retains
Wh ite nevertheless wins. chances of success, by playing 9 0,d1! 11c1
Practical Exercises in the Taking of d ifficult Decisions ttJ 1 63

1 0 cJ;; f1 r1xd 1 + 11 cJ;; e2, and if 1 1 . . . 'l:.g 1 ?, A draw would have resulted from 12 l:[xf3
then 12 i.e 7!, winning - Dvoretsky.] cJ;; xf3 .
The king feels a little more comfortable on The mistakes made by my opponent to
the h-file than on the Bth ra nk. The attempt wards the end of the game are easily
to play for mate looks tempting: B i.e7!? If explained. He thought that his position was
B . . . .:.xc3 , then White decides matters with won , and he took the pawn sacrifice . . . g6-
the q u iet move 9 g 3 ! ! found by Dolmatov, gS to be sign of desperatio n . The move
after which the king can no longer escape 6 . . . cJ;; g 6 ! , complicati ng the position , came
from the mating net. This means that Black as a su rprise to h i m . In severe time-trouble
has to defend with B . . . ':'c1 + 9 cJ;; h 2 lDg4+ 1 0 (wh ich was mutual, incidentally) and under
cJ;; h 3? ':'xc3+ 1 1 g3 lDf2+ 1 2 cJ;; g 2 lDe4, or 9 the psychological effect of the sharp change
cJ;; f2 lDg4+ 1 0 cJ;; e 2 l:txc3 . There is little joy of situation , Wh ite loses his bea rings and
here, of course, but even so it is possible to even loses the game.
fight o n . 12 . . . cJ;; x g3
Probably t h e simplest way, and t h e most 1 3 l:lxf3+ cJ;; xf3
unpleasant one for me, was the win of two
1 4 cJ;; e 1 'iii> e 3
pieces for a rook either in the previous
variation ( 1 0 lbg4 with the king on h2 or f2), 1 5 cJ;; f1 c4! ?
or by B i.f6!? l:tc1 + (B ... lDg4 9 :gS+ cJ;; h 6 1 6 lDxd6 'iii>f3?!
1 0 i.g7 mate , or 9 . . cJ;; h 4 1 0 J:lxfS+ cJ;; g 3? 1 1
. 1 7 cJ;; e 1 c3
lDe4 mate) 9 cJ;; f2 lDg4+ 1 0 lbg4 cJ;; xg4 1 1 1 8 lDxf5??
a4. White should be able to convert his
1 B d 1 was essential, with a probable
advantage, although he still needs to over
d raw.
come some technical difficulties.
[Black would still have had to work for the
8 . . . lDg4
draw, by playing 1 B. . .1:!xa2 (weaker is
9 g3?! lDh2? 1 B. . . I!d2+ 1 9 cJ;; c 1 'l:.xd5 20 lDc4 cJ;; e 4 2 1
Black wants to set u p the afore-mentioned cJ;; c2 cJ;; d4 2 2 lDb6! o r 2 0. . . 1:.d4 2 1 lDe5+
drawing mechanism with . . . ':d2 and . . . lDf3+. e4 22 cJ;; c2 1:.d2+ 23 cJ;; xc3 1:lxa2 24 lDd7)
[Here exactly the same motifs operate as 1 9 lDc4 lta4 20 lDe5+ cJ;; e 4 2 1 d6 :!d4+ 22
were mentioned earlier. 9 . . . J:ld2!! would 'iii> c2 1:!xd6. Therefore it made sense to
have given a draw, while White could have advance the c-pawn a move earlier: 1 6. . . c3!
retained winning chances by choosing 9 1 7 lDxf5+ 'iii> e 4 with equality Dvoretsky.]
-

cJ;; f1 ! lDe3+ 1 0 cJ;; e 1 lDxg2+ 1 1 cJ;; f1 lDe3+ 1 2 18 . . . 1:!c1 mate


cJ;; g 1 cJ;; g 4! 1 3 1:.eB! cJ;; f3 1 4 l:xe3+ cJ;; xe3 1 5
The decision taken by Black ( 1 . . . gS and
lDxd6 - Dvoretsky.]
2 . . .fS) was, I think, the correct one from the
Here my opponent used up his last few
practical point of view, despite the fact that
mi nutes, since he simply could not find a
in analysis it proved possible to fi nd a
forced win . I n fact there no longer is one. He
refutation , and not only one. As in the
should probably have gone in for the
previous example, I wanted to divert the
variation 1 0 i. f6 lDf3+ 1 1 f1 ltd2 1 2 ltgS+
opponent from a purely tech nical cou rse,
lDxgs 1 3 fxgS l:xdS, in which accu rate
aiming at any cost to complicate the position
defence would have been req u i red of Black.
and create active cou nterplay. The new
10 ]::te 8?! cJ;; g4 situation, with its mutual attacks and unu
1 1 :e3 lDf3+ sual balance of materia l , proved not at all
1 2 cJ;;f1 ?! easy to fig u re out.
1 64 Practical Exercises in the Ta king of difficult Decisions

In seeking decisions, don 't restrict your And now the results of the competitio n . The
self only to an analysis of variations. strongest today was Svidler, with Zviagintsev
Often general positional considerations in second place , and Emelin third . However,
also come to your aid. I n the given all of you did q u ite well . I hope that the
instance I looked to see how I could exploit experience accu mulated in the solving and
the defects of the move f2-f4, and I also discussing of the exercises will come in
aimed to release my king from i mprison useful at the board , where it is certain that
ment. you will constantly encou nter equally diffi
cult problems.
ltJ 1 65

Mark Dvoretsky

Vi rtuoso Defe n ce

W tion of Grigory Sanakoev (the topic of


hen I was studying the games collec 4 ttJxd4 ttJf6
5 ttJc3 a6
the chapter 'Thoughts about a book' my 6 g5 e6
attention was d rawn to the fol lowing excep
7 f4 e7
tionally tense and fasci nating encou nter.
Wh ite built up a very dangerous attack on 8 'it'f3 'iVc7
the king , wh ich , it appeared , would inevita 9 0-0-0 ttJbd7
bly achieve its goa l . However, the oppo 1 0 d3 b5
nent's cool-headed actions enabled h i m not 11 .:the1 b7
only to pa rry the immed iate threats , but 1 2 'ii' g 3 b4
even to seize the in itiative and in the end
1 3 ttJd5 exd5
wi n .
The piece sacrifice has to be accepted -
The colossal complexity o f the problems
incorrect are both 1 3 . . . ttJxd5? 1 4 exd5
facing the players is indicated by the fact
xd5 1 5 xe 7 (or 1 5 ttJxe6 xe6 1 6 1:[xe6
that even in play by correspondence they
- G ipslis) 1 5 . . . xe7 1 6 ttJf5+, and 1 3 . . .
were unable to avoid some errors (true, only
xd5? 1 4 exd5 ttJxd5 1 5 xe7 ( 1 5 e4!?)
a few). As sometimes happens, the stand
1 5 . . . ttJxe7 16 'it'xg7 1:[f8 1 7 :xe6! fxe6
ard of play proved sign ificantly hig her than
1 8 ttJxe6 (Chudinovskikh-Semenov, USSR
the standard of its analytical interpretation .
1 974).
This is not surprising: the maximum concen
tration of thought is attai ned during play (all 14 e5! ?
the same, whether it be by correspondence The main theoretical continuation is 1 4 exd5
or over the board ) . I was able to correct d8 with u nclear conseq uences.
some of Sanakoev's comments in a joint 14 . . . dxe5
analysis with g randmaster Vadim Zviag in ttJh5
1 5 fxe5
tsev, and in the preparation of the present
The game Petrushin-Chudinovskikh, USSR
book practically all the key moments in the
1 973, went 1 5 . . . ttJe4? 16 xe4 xg5+ 1 7
game were revised .
'it'xg5 dxe4 1 8 ttJf5 'it'xe5 1 9 ttJd6+ f8 20
'it'xe5 ttJxe5 21 ttJxb7 ttJg4 22 ttJd6 g6 23 h3
Sanakoev - Maeder ttJf6 24 ttJxe4 ttJxe4 25 :xe4 with a d ifficult
1 0th World Correspondence endgame for Black.
Championship 1 979-1 984 1 5 . . . 0-0-0 has also been played . Here are
Sicilian Defence two examples:
1 e4 c5 1 6 ttJf5 c5 1 7 exf6 'it'xg3 1 8 hxg3 gxf6 1 9
2 ttJf3 d6 f4 ttJe5 20 xe5 fxe5 2 1 :xe5 %1de8 22
3 d4 cxd4 ':xe8+ :'xe8 23 : h 1 l:th8 24 :h6 a5 25
1 66 Virtuoso Defence

d2 b8 26 1:tf6 l:tf8 27 .l:.h6 l:th8 28 :f6 Y:z 21 . . . liJf8 ! ? 22 f5 c8 23 %:te8 f7 24 :f7+


Y:z (Adams-Levitt, London 1 984); g8 25 1:[e8 with a draw. Later that has how
1 6 liJf5 c5 1 7 liJxg7 liJe4 1 8 xe4 dxe4 several 'duels' concluded , in particu lar
1 9 xd8 1:[xd8 20 e6 "xg3 (20 . . . liJf8 ! ? was Hellers-OeFirmia n , Biel 1 989. It is a pity
probably stronger) 2 1 exd7+ l:txd7 22 hxg3 that no one has risked trying 22 . . . g6!? 23
l:[xd 1 + 23 xd 1 f2 24 1:[ h 1 e3 25 l:[xh7 liJxf8 "xe7 24 xe7 gxf5 25 1:[xf5 with a
xg2 26 e2 c6 27 .l:.h5 b5+ 28 l:[xb5 position that is d ifficult to assess.
axb5 29 liJf5 with a won endgame for Wh ite 16 . . . xg5+
( N u n n-Kuczynski , Germany 1 995). 1 7 'ii'x g5 g6
1 8 e6
1 8 g4 is weaker, after which Black can play
either 1 8 . . . liJg7 1 9 e6 liJc5 20 exf7+ xf7
(Weigel-Hauernherin, correspondence 1 977),
or 1 8 . . . h6!?

1 6 'ili'h4
The sharp attempt 1 6 e6?! is interesti ng. I n
the opin ion of Vlad i m i r Lepyoshki n , Wh ite
gains the advantage after the cautious
1 6 . . . xg5+ 1 7 "xg5 "f4+ 1 8 Wxf4 liJxf4
1 9 exd7+ xd7 20 f5+ d6 2 1 g3 liJg6 22 I n this sharp position Black has tried various
h3 liJe7 23 :e3 or 1 7 . . . liJdf6 1 8 exf7+ continuations:
xf7 1 9 liJe6 'ili'd6 20 e2 g6 2 1 xh5 A) 1 8 . . . liJdf6 1 9 exf7+ xf7 20 :f1 1:[e8 21
1:[ae8 22 g4 c8 23 lIf1 l:txe6 24 xe6+ g4 We5 (weaker is 21 . . . g8 22 gxh5 liJxh5
'ii'x e6 25 1:[xd5. 23 xg6 with a dangerous attack) 22 'ii' h 6
The q ueen sacrifice should be accepted : (22 'iWh4!?) 22 . . . liJg7 23 g5 'ii'e 3+ 24 b1
1 6 . . . liJxg3 1 7 exf7+ r:Ji;xf7 1 8 1be7+ g8 1 9 g8 25 xg6? ! (25 l'1f3 would have
hxg3 'ii'x g3 ( 1 9 . . ...e5 20 f5) 20 liJe6 'iWe5 maintained equality) 25 . . . hxg6 26 'iWxg6
2 1 1:[f1 . The correspondence game Baluev (Rodriguez Talavera-Nedobora, Seville 1 992)
Vad i kan ( 1 976), where 1 6 e6 was first and, by playing 26 . . . liJg4 ! ? , it would appear
employed , did not last long : 2 1 . . . liJc5? 22 that Black could have claimed an advan
f5 liJe4? (22 . . . liJxe6 23 xe6+ 'ii' x e6 24 tage.
:xe6 h6 with advantage to Wh ite) 23 xe4 B ) 18 . . ...f4+ 1 9 "xf4 liJxf4 20 exd7+ xd7
dxe4 24 1:[f6! 'ii'x g5+ 25 liJxg5 d5 26 1:[xa6 2 1 1:[f1 liJxd3+ 22 l:1xd3 f5 23 g4 - th is was
Black resigned . In Informator it was an no first played in the correspondence game
tated by Lepyoshki n , who gave the variation Shakarov-Zhuravlev, 1 976. Later practice
Virtuoso Defence ttJ 1 67

showed that the endgame is eq ual . For A tempti ng move . After defending his d7-
example: 23 .. J:taf8 24 gxf5 gxf5 25 11g3 pawn , White then wants to double heavy
l:.hg8 26 tt)xf5 lbg3 27 hxg3 h5 28 'iii> d 2 a5 pieces on the h-file. The immed iate attempt
(Lechtynsky-Sch mid, Germany 1 994) . to implement this attacking set-up is ineffec
C} 1 8 . . . tt)c5 1 9 exf7+ 'iii> xf7 20 :f1 + 'iii> g 8 2 1 tive: 2 1 'ii' h 6 ::txd7 (but not 2 1 . . . 'iWxd7? 22
tt)f5 1:[f8 ( o r 2 1 . . . tt) e 6 22 tt) h 6 + 'iii> g 7 2 3 :e3 'ii'x g4 23 e2 ! and 24 l::t h 3) 22 :e3
tt)f5+ 'iii> g 8 with a d raw, as in Luther-P. N iel tt)e6, and if 23 ':h3?! there is 23 . . . 'iWf4+,
sen , Malmo 2002) 22 tt)e7+ 'iii> g 7 23 xg6 while 23 f5? is refuted by 23 . . . tt)xd4 24
tt)e6 24 tt)f5+ l:txf5 25 'ii'xf5 tt)hf4 26 h5 l:txd4 ::te7 25 l:.h3 f6 (Sanakoev). A draw
l:.f8 27 'iVg4+ 'iii> h 8 28 g3 l::t c8 ! 29 ':'f2 'ii'c5 results from 23 xg6 hxg6 24 xe6 fxe6 25
30 l:tfd2 :g8 (30 . . . tt)xh5 31 'ii'x e6 tt)g7 is of 'ii'x g6+ 'iii> h 8, while the position arising after
equal merit - Berelovich-Dvoiris, Hoogeveen 23 l:txe6 fxe6 24 tt)xe6 'ii'd 6 25 tt)xf8 'iWxf8
200 1 ) 31 'ii'f5 tt)xh5 32 'ii'x e6 tt)g7 with 26 'ike3 is merely slig htly more pleasant for
roughly equal chances (Hakki-DeFirmian, White.
Hamar 1 983). However, there was another, stronger offen
I n two games from the 1 0th World Corre sive plan, involving the advance of the h
spondence Championship, Karl-Heinz Mae pawn . It was employed in the game Estrin
der chose another, more risky method of Maeder, played in the same World Corre
defence. spondence Championship.

18 . . . O-O?! 21 h4! Ibd7

1 9 exd7 l:.ad8 Wh ite's attacki ng possibilities are illustrated


by the followi ng variation : 2 1 . . . 'ii' b 6 22 h5!
20 g4 tt)g7
'ii'x d4?! 23 xg6 'iWb6 24 xf7+! .:txf7 25
Black has noth ing better: after 20 .. .'.f4+ 2 1 l::t e 8+ l::t f8 26 l:txf8+ 'iii> xf8 27 ':f1 + 'iii> g 8 28
'ii'xf4 tt)xf4 2 2 ':e7, 2 0 . . . tt)f4 2 1 f5 'iWe7 (th reatening 29 h6) 28 . . . h6 29 g5 tt)f5
(followed by :e7) or 20 . . .f6 2 1 'ikh4 tt)g7 22 (29 . . . hxg5 30 h6) 30 'ii'e 8+ 'it>g7 31 'ike5+
l::t e 7 'ii'f4+ (22 . . . ':xd7 23 tt)e6) 23 l:1d2 'iii> g 8 32 'iWxf5 'iWe3+ 33 'iii> b 1 'ii'x g5 34 'fie6+
'iVxd4 24 xg6 'iWxd2+ 25 'iii> x d2 hxg6 26 g5 'it>h8 35 'fie8+ 'fig8 36 'iWe7 , and there is no
he is in trouble. satisfactory way of parrying the threat of 37
l::t e 1 followed by 38 'ii'f6+ or 38 'ii'xd8.
22 h5 'iWd6 (22 . . . 'iWb6 ! ? ) 23 'ii' h 6 iLe8 24
l:t h 1 tt)e8 25 :df1 ':e7 (25 . . . tt)f6 26 tt)f5!
gxf5 27 xf5) 26 g5 'iWe7 27 tt)e6! lIxe6 (if
27 . . . 'iWd6 Yakov Estri n gives 28 hxg6 fxg6
29 l::t xf8+ 'iWxf8 30 'iWxf8+ 'it>xf8 31 tt)xb4
b7 32 xa6 , when the endgame is
hopeless for Black) 28 hxg6 f6 29 gxf6
l::tfxf6 (29 . . . tt)xf6 30 ':hg 1 ! 1:[e8 3 1 g7) 30
l::t fg 1 ! e6 31 gxh7+ 'it>h8 32 'ii'g 5! tt)g7 33
'iWxg7+ 'iWxg7 34 ':xg7 , and White won .
The following analysis (if, of course, it is
correct) shows that after the move in the
game White no longer has an advantage .
B u t t o foresee t h i s is q uite impossible even
21 f5 1 ? in a game by correspondence, to say
1 68 Virtuoso Defence

noth ing of over-the-board play. 23 . . . gxf5 24 gxf5 f6 25 ':'g 1 (but not 25 1:e7
21 . . . i.c6 :tf7 26 ti)e6, as recommended in the first
edition of Sanakoev's book, because of the
Maeder aims to eliminate the dangerous d7-
counter-stroke 26 . . . ti)xf5 ! pointed out by
pawn as soon as possible. The wh ite bishop
John N u n n ) 25 . . . :tf7 (25 . . . i.c8? 26 l:th3) 26
is immune: Black loses qu ickly after both
ti)e6 ! i.xe6 27 fxe6 ':e7 28 :th3 with a
21 . . . gxf5? 22 ti)xf5, and 2 1 . . .f6? 22 'ikh6
powerful attack;
gxf5 23 Ite7 Itf7 24 Itde 1 Itdf8 (24 . . . l:txd7
25 1:[e8+) 25 1:xf7 xf7 (25 . . . 1:xf7 26 1:e8+ 23 . . .i.xf5 24 l:th3 ti)h5 25 l:.xh5 gxh5 26
l:.f8 27 l:.xf8+ xf8 28 ti)e6+) 26 ti)xf5 ti)xf5 gxf5 'ii'x h2 27 ti)f3, and Black has to give up
27 gxf5 . queen for rook;
22 11d3 ! ? 23 . . . l:.fe8 24 l:th3 gxf5 25 'ii'x h7+ f8 26
'ilfh8+ e7 27 'ikxg7 'ilff4+ (after 27 . . .fxg4,
Which rook should b e brought t o h3?
as considered by Sanakoev, the simplest is
Wh ite's choice looks illogica l , since after
28 'iVg5+) 28 b 1 fxg4 29 lIh6, and the
22 . . . i.xd7 23 'iVh6 (or 23 l:th3) the move
black king is in trouble.
23 .. .l::tfe 8 will be made with gain of tempo:
Wh ite will have to lose time moving his rook The best defence is 23 . . . ti)e6 ! (with the idea
from e 1 . Of course, Sanakoev took this into of 24 .l:th3? 'ilff4+). The wh ite player thought
account, but nevertheless, after delving that 24 ti)xe6 (24 i.xe6?! fxe6 25 ':xe6!
deeply into the concrete variations, he lIde8 26 l:txg6+ is sufficient only for a draw)
preferred the move in the game. Later he 24 . . . i.xe6 25 i.xg6 fxg6 26 l:txe6 'iVf4+ 27
concluded that his decision was incorrect 'iVxf4 1:xf4 would lead to an equal endgame
and that 22 1:[e3 should have been played . I (i ncidentally, after 26 . . . 1:[f2 ! 27 1:[d2 .l:.f1 + 28
carried out a joint analysis with Vadim 1:d 1 .uxd 1 + 29 xd 1 Black's position is
Zviagintsev, after which we d isagreed with even to be preferred , since the enemy king
the conclusion of the author - in fact the two can no longer feel secure).
moves are roughly equivalent. Later Sanakoev real ised that the simple 25
Let us examine the position arising after .l:.h3 would guarantee h i m a very strong
22 1:[e3 i.xd7. attack, for example: 25 . . . 1:fe8 26 'iVxh7+
f8 27 'ii' h 6+ e7 28 'iVg5+ etc.
However, this entire variation is of no
importance, since it is based on the errone
ous assumption that after 24 . . . fxe6 (instead
of 24 . . . i.xe6?) 25 i.xg6 White wins. Noth
ing of the sort! - the obvious 25 . . . i.c8 26
l:th3 l:td7 parries the opponent's attack.
Thus the exchange on e6 does not give
Wh ite any advantage. Sanakoev wanted to
play 24 l:[f1 , having in mind the winn ing
variations 24 . . . ti)xd4 25 l:[h3 l:[fe8 26 i.xg6
or 24 . . . l:tfe8 25 ti)xe6 (25 l:th3!?) 25 . . . i.xe6
26 l:th3. Alas, there is a refutation: 24 . . . 'iVc4 ! ,
a n d t h e rook on f1 is hang i n g . B u t with the
wh ite rook on d3 the q ueen move could be
23 'ii' h 6 suggests itself. Sanakoev gives the met by b2-b3 - this is why Sanakoev played
following variations: 22 l:[d3.
Virtuoso Defence ttJ 1 69

The attempt to transpose moves by playing


23 l:.h3 (instead of 23 'ilh6) is interesting. If
now 23 . . . tDe6?, then 24 tDxe6 , and if
24 .. .fxe6 not even 25 i.xg6 'ilVf4+ 26 'ilVxf4
':xf4, but 25 ':'xh7 ! . The sad consequences
for Black of the variation 24 . . . i.xe6 25 'ilVh6
are already known to us.
Wh ite retains the advantage in the event of
23 . . . h5?! 24 i.xd7 'ilVxd7 25 l:.h4.
However, with the rook on h3 Black can
accept the piece sacrifice : 23 . . . f6! 24 'ilVh6
gxf5 (or 23 . . . gxf5 24 'ii' h 6 f6 ! ) . The point is
that if 25 gxf5 there is the simple reply
25 . . . i.xf5, while after 25 'ii'x h7+ f7 the What should Black do now? 25 l:th3 is
black queen acq u i res the important f4- th reatened , and the captu re on f5 allows the
square : 26 gxf5 (26 . . . .l:th8 was threatened) opponent to develop a mati ng attack
26 . . . 'ilf4+ 27 b1 i.xf5 . Even after the (24 . . . gxf5? 25 gxf5 f6 25 .l:tg 1 ).
comparatively best 25 1:f1 'iVe5 ! ? or If 24 . . . 'ilVe5 Sanakoev had prepared a
25 . . . 'ii' c4 !? the compensation for the sacri complicated combi natio n : 25 :h3! 'ifxd4 26
ficed piece is clearly insufficient. i.xg6 i.. e 6! 27 i.. x h7+ f8 28 :d3 'ife5
22 . . . i.xd7 (28 . . :i!fc4!? comes into consideration; how
23 'iVh6 ]:tfeS ever, this resou rce can be excl uded by
transposing moves: 27 l:td3! , when 27 . . .
Black develops his rook with gain of tempo,
'ii' c4? 2 8 'iVxh7+ f8 2 9 i.xf7 i.xf7 30
although here too 23 . . . tDe6 !? came into 'ifh8+ e7 31 'ii'x g7 is bad for Black) 29
consideration . 24 l:th3? 'iff4+ is hopeless l:te3 'ifd4 30 i.f5 :d6 3 1 l:td 1 'ifc5 32 ee 1
for White, as is 24 11f1 ? ! 'ii'e 5! (24 . . . 'ii'c4 !? is
followed by 33 'iVh8+ , and Wh ite regains the
also possible, since 25 b3 Wc5 26 l:[h3 will piece, achieving a decisive advantage (32
be met by 26 . . . tDg5!). The regaining of the l:[de 1 is even more energetic).
pawn by repeated captu res on e6 gives
The defence can be improved by 30 . . . e7!
Black the initiative.
(instead of 30 . . . :d6?) with chances for both
The critical variation is 24 tDxe6 fxe6 sides.
(24 . . . i.xe6? 25 l::t h3) 25 l:[h3! i.c8 26 Wh ite can try 30 i.. g 6 (instead of 30 i.. f5). If
i.xg6. A similar position arose after 22 l:te3, 30 . . . l:[d7 it is simplest to defend the g-pawn
but there the wh ite rook stood at d 1 . It is by 31 h 3 ! , when it is unclear what Black
better placed at e 1 , as is mainly seen in the should do. However, it is not apparent how
variation 26 . . . .l:td7? 27 :xe6 .l:tg7 28 i.xh7+ 30 . . . 'ifxg4! can be refuted . For example, 3 1
.l:txh7 29 .l:tg6+. Black is forced to reply i.xf7 'ifg2 ! 3 2 :ef3 i.. xf7 3 3 'ifh8+ e7 34
26 . . . 'iff4+ 27 'iVxf4 l:[xf4 28 i.xh7+ g7 - :xf7+ d6 35 'ii' x g7 'ifxg7 36 :xg7 :e2
here there is some advantage (although it is leads to a double rook endgame, which is
probably only slig ht) on the side of Wh ite . most probably drawn . 31 .l:tg3 'ifd4 or 3 1
24 f1 ! .l:tef3 .l:t d 7 is also unconvi ncing (but not
31 . . . e7? 32 nxf7+ i.xf7 33 l:txf7+ d6 34
i.f5+ :e6 35 'ifxe6+ tDxe6 36 i.xg4 with a
big advantage).
1 70 Virtuoso Defence

Even so, in these variations Black's position It was only when the present ed ition was
looks shaky, and it is dangerous to go in for being prepared for publ ication that I d iscov
such play - one would l i ke to fi nd someth ing ered both the refutation of Maeder's idea ,
rather more safe. and the defensive improvement 24 . . . 1Ii'c4 ! .
I suggest playing 24 .. .'iWc4 ! ' Sanakoev 25 l:th3?
writes that after this 'I could simply reply 25 A bold, but objectively i ncorrect decision !
b3 and continue the attack in comfort. Alas, '
Sanakoev tries to break through immedi
after 25 . . .1Ii'c5 to attack 'in comfort' does not ately on the h-file, but at a high price : the
prove possible, since 26 l:1h3?? no longer black pawn is now on the threshold of
works in view of 26 . . .'ii' x d4 27 11i'xh7+ 'itff8 q ueen ing. However, subseq uently it may
(th reatening an extremely u npleasant check still be possible to stop it, by playing 1:1a3 or
at a 1 ). In the variation 26 iLxd7 llxd7 27 It'lb3 .
.l:.h3 1Ii'xd4 28 1Ii'xh 7 + 'itff8 29 11i'h8+ 'itfe 7 30 25 . . . bxa2
.l:.e 1 + 'itfd6 31 ':'xe8 1Ii'g 1 + ! 32 'itfb2 It'lxe8
26 Wxh7+ 'itff8
33 'ikxe8 'ikd4+ Black is guaranteed a draw.
And after 26 iLxg6 he can reply either 27 'ii' h 8+
26 . . . hxg6 27 l:.h3 f6 28 1Ii'h7+ 'itff7 29 l:th6 In the event of 27 1:1a3 Black would have
'ii'x d4 30 'ili'xg6+ 'itfg8 with a draw, or gained an advantage with the spectacular
26 .. .fxg6 27 .l:.h3 lt'lh5 28 l:txh5 %1e7! with an 27 . . . l:te1 +! 28 l:txe 1 Wf4+ 29 'itfd 1 Wxd4+
u nclear position . 30 'itfc1 lIc8! (th reatening 3 1 . . . 1:1xc2 + ! ) 31
Maeder preferred a completely different 'ii' h 6+ 'itfg8 32 1Ii'e3 Wc4 or 3 1 ':xa2 iLxf5
method of defence. 32 'itf b 1 ':'xc2 33 'itfa 1 ':'e2 .
24 . . . b3?! 27 . . . 'itf e 7
An u nexpected reply! Any captu re on b3 28 Wh4+
would seem to have its d rawbacks. If 25
It'lxb3?, then 25 . . . iLb5 26 l:tc3 'ili'xc3 27
bxc3 iLxf1 (Sanakoev considers this posi
tion to be 'completely unclear', but in fact
here Black has a big advantage). If 25 axb3
there follows 25 . . . 'ika5 (th reatening not only
26 .. .',.'a 1 + , but also 26 . . . :1e 1 + ) 26 'itfb1
iLb5, and it is now Wh ite who has to gain a
d raw by 27 iLxg6! fxg6 28 l:t h 3 ! .
However, i t is not altogether clear what
Black had in mind in reply to 25 :xb3 ! Wa5
26 l:t d 1 ! (26 c3 is much weaker, since the
rook is cut off from the kingside). I do not
see any satisfactory defence against the
threatened l:th3, for example: 26 . . . gxf5 27
gxf5 f6 28 l:th3, 26 . . . Wxa2 27 l:th3, or Sanakoev had aimed for this position . If
26 . . . lt'le6 27 It'lxe6 iLxe6 28 l:th3 Wxa2 29 28 . . . 'itfd6 he was intend ing to reply simply
iLxg6 ! . 29 :a3, stopping the pawn and reta i n i ng a
I n Sanakoev's games collection and i n strong attack.
previous editions o f my book, Black's last The king move to f8 did not concern h i m , if
move was awarded two exclamation marks. only because if Wh ite wishes he can satisfy
Virtuoso Defence 1 71

h imself with a repetition of moves (28 . . .'i'f8 or 34 . . . lbh5? 35 gxh5 J:.c8 36 .i:tf3) 35 ..td3
29 'ii' h 8+ e7 30 'iWh4+ ). He can also l:lxd3+ 36 xd3 g8 with approximate
consider 29 iVf6!? a 1 iV+ 30 d2 . However, equal ity.
here with correct play the game should Alas, Black has available a much sounder
conclude with the same result: defence, secu ri ng him the advantage.
A) 30 . . . 'iif4+? is incorrect: 3 1 l:lxf4 iVe 1 + 32 28 . . . f6!
d3 'ii'd 1 + 33 c3 l:lc8+ 34 b3, and after 29 1:.e3+
moving along the 3rd rank the king hides
Now, in Sanakoev's opinion, Black loses
from the checks at a2 (Sanakoev). Noth ing
after both 29 . . . 'iii f7? 30 ..txg6+ and 29 . . . f8
is changed by 32 . . . ..tf5+ (in the hope of 33
30 ..txg6! ' However, the second variation
gxf5? lbh5 or 33 .i:txf5?! iVd 1 + 34 c3
must be conti nued : 30 . . . a 1 'ii + 31 'iii' d 2
iVe 1 + 35 b3 1I b8+ 36 a2 iVa5+ 37 .ll a 3
'iica5+ 32 c3 'iix b2+ 33 lbc2 1Wxc2+! 34
iVc7) 33 lbxf5! 'iid 1 + 34 c3 etc.
..txc2 f5 ! 35 ..txf5 ..txf5 36 'it'h8+ 'iii' f7 37
B) I n Sanakoev's opinion , 30 . . . iVaa5+ loses ':xf5+ lbxf5 38 iVh5+ 'iii' g 7 39 'iWg5+ ri;f7 40
to 31 c3 lbh5 32 l:[xh5 gxh5 33 iVh6+ rJ;; e 7 'ii'xf5+ rJ;; g 7 41 Vi'g5+ ri;f7 42 'iWh5+ (there is
34 l:le1 + ..te6 35 ..txe6 xh2+ 36 l::t e 2 no mate after 42 l:lf3+ e6) 42 . . . rJ;; g 7 43
iVxe2+ 37 xe2 l:[d6 38 lbf5+ d8 39 l:lxe8 'iWa2+ 44 e3 l:lxe8 45 Vi'xe8 'ii'x h2,
lbxd6 .i:txe6+ 40 'ii' x e6 fxe6 41 lbb 7 + with a and the result is a drawn q ueen endgame.
won ending for Wh ite . I n fact, the evaluation However, Black is no longer satisfied with a
of the endgame after 41 . . . c7 42 lbxa5 d raw.
hxg4 43 f2 is sti ll not altogether obvious, 29 . . . lbe6 ! !
but this is immaterial, since instead of 39
2 9 . . . ..te6 was weaker - in the variation
lbxd6 Wh ite wins far more simply by 39
which occu rred in the game 30 lbb3 'ii'c4 3 1
'iWf6+! ri;c7 40 'ixf7+ ri;b8 4 1 lbxd6. On the
iLd3 'ii'x b3+ 3 2 'ii'xf6+ the knight o n g 7
other hand , Black is by no means bound to
would b e vul nerable.
place his rook on d6, where it is immed iately
lost - 37 . . . l::t b 8! is stronger, when Wh ite , 30 lb b3
apparently, has no advantage. Besides, No combinations are apparent, and there
Black can play d ifferently on the 35th move . fore Wh ite has to retreat.
True, 35 . . J:td6? does not work in view of 36 30 . . . 'iVc4!?
..tc8+! (with the idea of giving mate in the A good move, but by no means Black's only
variation 36 . . . 'iii d 8 37 .i:txe8+ xe8 38 option. 30 . . . .i:th8 31 'ii'e 1 gxf5 32 gxf5 l:lxh2
h8+ e7 39 lbf5 mate) 36 . . . l:le6 37 ..txe6 was very strong, for example, 33 l:lxe6+
d8 38 'iWg5+ 11e7 39 .i:te5. However, the f7 ! 34 .i:te7+ rJ;;f8 35 .i:te2 l:tb8!?, or 33 .i:tc3
simple 35 . . .fxe6! ? forces Wh ite to be satis 'ia5 34 'ii' a 1 d4.
fied with perpetual check: 36 'iWxe6+ (36 31 iLd3 'ii'x b3
llxe6+? d7) 36 . . . f8 37 .i:tf1 + 'iii' g 7 38
32 'ii'xf6+ 'iii d 6
lbf5+ h7 39 'ii' h 6+ g8 40 iVg6+ 'iii h 8 .
33 'ii' e 5+ c6
C ) 30 . . . 'ii'c a5+ 3 1 c 3 iVxb2+ (perpetual
check results from 3 1 . . . ri;g8 ! ? 32 ..txg6! 34 cxb3 a1 'ii' +
xb2+ 33 ..tc2 ..te6 34 lbxe6 'ii'x c2+ ! ) 32 35 'iii c2 'ii'a 5
..tc2 iVbxc3+ 33 l::t x c3 .i:te7 (here Sanakoev 36 ..txg6 .i:tf8
terminated his analysis, thinking that in this 37 .i:tf7
way White's attack was refuted) 34 'it'h4!
1:1e3 ! ! (Black loses after 34 . . . 'iii g 8? 35 'ii'x e7 (see diagram)
1 72 Virtuoso Defence

'iVc3+ '>t>b7) 39 . . . l:txe6 40 'iVc3+ in view of


40 . . . '>t>b7 4 1 'iVb4+ : b6 42 'it'xf8 (42 ':xf8
'ii' c 7+ 43 'ili'c3 l:tc6) 42 . . . 'ii'xf8 43 l:txf8
l:txg6.
I th ink that Black would have retai ned a
significant advantage by conti nuing 37 . . .
'it'b4!? Maeder found another, also very
strong conti nuatio n .
37 . . . ttJc7 ! !
The rook on f8 is indirectly defended : 38
'ili'f6+ '>t>b7 39 11xf8 'iVc5+. Wh ite must finally
forget about any ambitious plans and beg in
fig hting for a draw, although it is now unclear
how to attain it.
At this point Wh ite was still feeling optimis
38 'iVe7 l:t g 8 !
tic. His pieces are active, and the oppo
Perpetual check results from 3 8 . . . :tfe8 39
nent's material advantage is not too g reat.
l:tf6+ '>t>b7 40 i.. x e8 .uxe8 4 1 'iVxd7 .uxe3 42
However, analysis shows that, in al most all
'ili'c6+ '>t>b8 (42 . . . 'it'a7?? 43 b4 ! ) 43 f8+
of the continuations available to Black a
cj;; a 7 44 Itf7 (44 b4? 'iVb6) 44 . . . .l:!.e2+ 45
d raw is the l i m it of White's dreams, an d in
'>t> b 1 'iVe 1 + (45 . . . .l:!.e1 + ) 46 cj;; a 2 'iVa5+. The
some cases it is difficult to ach ieve.
outcome is the same after 38 . . . d4 39 .l:!.e5
An i nteresting drawing variation was sug
.l:!.xf7 (of course, not 39 . . . ttJd5? 40 .:txd 5 ! ) 40
gested by Zviag intsev: 37 . . . '>t>b6 38 'iVd6+
i.. e4+ ! ? (40 i.. xf7 d3+ 4 1 '>t> b 1 ttJb5)
i.. c6 39 'ili'xe6 l:tde8! 40 'iVd6! (40 l1xf8?
40 . . . ttJd5 4 1 'ili'xf7 d3+!? 42 cj;; b 1 (42 'iit x d3
l:txe6 41 .l:!.xe6 'ikc5+; 40 l:te7? .i:tf2+ 4 1
.ixg4) 42 . . . 'ike 1 + 43 'it'a2 'iVa5+ . A roughly
b 1 l:!.f1 + 4 2 c2 'iVc5+) 4 0 . . . l:txf7 4 1 l:!.xe8
equal position is reached after 38 .. J:txf7 39
.l::f.f2+ 42 '>t>b 1 .i:tf1 + 43 'it'c2 l:!.f2 + .
.ixf7 (39 l:tc3+?? 'ikxc3+) 39 . . . ttJb5 40
After 37 . . . .l:!.xf7 38 i.. xf7 ttJ c 5 ( 3 8 . . . ttJc7? is iVf6+, when Black has to give up his d5-
bad : 39 'ikf6+ '>t>b 7 40 'iVxd8 'ikc5+ 4 1 l:tc3) pawn.
39 .ixd5+ '>t>b6 Sanakoev was planning 40
However, it would appear that 38 . . . .:th8!?
'iVd6+ '>t>a7 41 b4 (4 1 l:te7 is better)
was not inferior to the move in the game: if
41 . . . 'iVa4+ 42 .ib3 ttJxb3 43 'iVc7+ '>t>a8 44
39 b4 there follows not 39 . . . l:txh2+? 40 '>t>b3
'iWxd8+. However, instead of the losing
'iVa 1 ! 41 l:!.c3+ '>t>b7 42 l:txc7+ xc7 43
42 .. .'Jxb3?? Black achieves the opposite
'iVc5+ with a d raw, but 39 . . .'a4+! 40 'it'b1
result with the simple 42 . . . 'iVxb4 (th reaten
'iVd 1 + 4 1 a2 cj;; b 7.
ing 43 . . . .if5+) 43 'iWc7+ ttJb7. Wh ite has to
39 b4 'iVa4+
play differently: 40 'ii'f6+ '>t>a7 41 b4 'ika4+
(4 1 . . . i.. a 4+? 42 b3) 42 '>t>c3 'iWd 1 ! 43 bxc5 In the event of 39 . . .'iWb5? 40 i.. f5 the sharp
'ii' c 1 + 44 '>t>d4 'iVxb2+ 45 11c3 'iWd2+ 46 d3 skirmish would have ended in a draw:
'ili'b2+ with perpetual check. 40 . . . i.. xf5+ 41 gxf5 'if'c4+ (or 41 .. :a4+ 42
l1b3! .l:!.g2+ 43 '>t> b 1 l:tg 1 + 44 '>t>c2 ) 42 .l:!.c3
In the event of 37 . . . l:tde8 Wh ite must aim for
g2+ 43 '>t>d 1 ':g 1 + 44 '>t>d2 l:tg2+ with
a draw by 38 l:tc3+ ttJc5 39 'iWd4 .l:!.e2+ 40
perpetual check (not 45 '>t> e 1 ? l:.e8).
'it' b 1 . Sanakoev's recommendation is sig
nificantly weaker: 38 'iWf6 'it'd8!? (38 . . . 'iWb4!?) 40 d2 '>t>b7
39 l:!.xe6+? (Wh ite is also worse after 39 41 i.. f5
Virtuoso Defence tLJ 1 73

A bad sign for White : he is forced to With the rooks on (43 . . . xc6?! 44 h4) the
exchange pieces. But after 4 1 11f6? (with position would become unclear - Wh ite's
the threat of 42 :a3) 41 . . . xg4! it is now his passed pawns are rather dangerous. I n
king that comes under attack. order to neutralise them, i t is i mporta nt
It is more d ifficult to evaluate 41 l:tc3 :ge8 !? above all to exchange the strong wh ite
42 'ifc5 (but not 42 'ifd6? :e2+! 43 xe2 bishop, which explains the captu re on c6
xg4+ 44 f2 :xd6 45 :cxc7+ b8! 46 with the king . It is true that the opponent
.l:.b7+ c8 47 :a7 'ifc6) 42 .. :iVc6 . In the gains the opportun ity to regain part of the
event of 43 'iff2 a draw results from material deficit, but the position is sign ifi
43 . . . 'ifxg6? 44 l:.xc7+ xc7 45 'ifa7+, but cantly simpl ified , and all the same Black's
Black can choose between 43 . . . 'iVd6!? and position remains sufficient for a win .
43 . . . 'ifb6 ! ? 44 'ifxb6+ xb6 45 :xd7 :xd7 44 l:tc3+ d6
46 .l:.xc7 xc7 47 xe8 :h7 48 h5 d6 - 45 ':f6+ e51
with a big advantage i n both cases. 46 l:bc7 xf6
43 h4!? is more promising for Wh ite. I n tu rn , 47 ':xd7
Black's play can b e improved : 4 1 . . . 'iVb5! I n the event of 47 xd7 :h8 Black soon
(instead of 41 . . . :ge8 ) 42 'ife5 (42 'iVd6 creates a decisive attack by the un ited
'iVb6) 42 . . . 'ifb6 43 .l:r.f6 lDe6, and the efforts of his two rooks and king .
advantage remains on his side.
47 . . . ':xd7
41 . . . :ge8 48 .i.xd7 l1h8
42 'iVd6 49 h3 e51
The king must stay i n the centre, while the
rook itself deals with the kingside pawns. If
49 . . . g5? there could have followed 50 b5
axb5 51 d3 l:[xh3+ 52 d4 :b3 (52 . . . b4
53 .i.a4 J:[h 1 54 xd5 ':c1 55 b5 with the
idea of 56 c4 ) 53 xd5 ':xb2 (53 . . . :d3+
54 c6 b4 55 e6) 54 c5 b4 55 a4 and
56 c4 with a draw.
50 g5 :g8
51 h4 ':h8

42 . . . 'iVc61
Not 42 . . . .:xe3? 43 :xd7! or 42 . . . 'ifd 1 +? 43
xd 1 a4+ 44 b3 :xd6 45 :h3! b5 46
:hh7 :c6 47 d7.
43 'iVxc6+
43 'ifxd7? :xd7 44 xd7 did not work in
view of 44 ... 'ifh6! 45 xe8 d4 .
43 . . . xc6!
1 74 Virtuoso Defence

52 96 6 0 <it>c4 d3
I n one way or another Wh ite must exchange 61 .ta4 d2
the q ueenside pawns. The immediate 52 b5 62 .t b3 :91
is refuted by 52 . . . <it>d6! 53 bxa6 (what else?) 63 b5 J:!c1 +!
53 . . . <it>xd7 54 a7 1:[a8! - the rook eliminates
Of course, not 63 . . . d 1 'ii' ? ? 64 .txd 1 l:lxd 1
the a-pawn , wh ile the king stops the passed
65 b6 with a draw.
pawns on the kingside.
64 'it>d5 l:[b1
I ncidental ly, another, less successfu l alloca
tion of the roles of the black pieces Wh ite resigned .
54 . . . <it>c7?! 55 h5 <it>b7 56 h6 <it>xa7 57 <it>e3
Let us sum u p . I n playing for a win both
<it>b6 is also good enough to wi n :
players willingly took g reat risks. From the
5 8 <it>f4 d4; opening Wh ite did not gain any advantage,
58 b4 <it>c6 59 <it>f4 <it>d6 60 <it>f5 <it>e 7 (but not but Black's d u bious castl ing on the 1 8th
60 . . . d4? 61 g6); move put him in an extremely dangerous
58 <it>d4 <it>c6 59 b4 (in the event of 59 <it>e5 position . He committed another error by
Black wins by both 59 . . . <it>d7 60 b4 <it>e7 6 1 being tempted by the spectacular 24 . . . b3?!
<it>xd5 l:tg8, and 5 9 . . .1le8+ 6 0 <it>f6 d4) instead of reta i n i n g the bala nce with
59 . . . 'it>d6 60 b5 'it>e6 6 1 b6 l::t d 8 ! ! 62 h7 <it>f7 24 . . . 'ifc4! . Sanakoev also went wrong twice:
63 h8'it' l:[xh8 64 'it>xd5 <it>e7 65 <it>c6 .l::r.c8 + ! ' on the 2 1 st move he did not choose the
52 . . . ':xh4 strongest plan of attack, and on the 25th he
53 b5 axb5 played too straight-forwardly for mate, miss
ing a q u iet way of refuting his opponent's
54 .txb5 <it>d41
idea . Later Black defended accu rately and
The most accu rate - the king supports the at no point d id he let his advantage slip. On
advance of the d-pawn . The conseq uences the whole, the qual ity of the play (taking into
of 54 .. .1lb4 55 .td3 (55 .te8) 55 . . . .:xb2+ 56 account the irrational natu re of the very
<it>e3 .l:lg2 57 <it>f3 are less clear. sharp situation which arose in the game)
55 <it>c2 can be assessed as very high.
In the variation 55 .te8 l:th2+ 56 <it>c1 (56 'I have played about 300 games by corre
'it>e1 <it>e3 57 <it>f1 <it>f3 ! ) 56 .. .1lg2 (56 . . . 'it>e3? spondence, the majority of which I have
57 g7! .l:tg2 58 .tf7 ; 56 . . . <it>d3!?) 57 .tf7 won, but few of these victories brought me
<it>e4 58 b4 d4 59 b5 d3 60 b6 <it>d4 (or such creative satisfaction as this unsuc
60 . . . <it>e3) Black wins, as is usually the case cessful attack. The excitement of the crea
in the endgame, by one tempo. tive process took such a hold on me, that at
55 . . . ':94 some point the actual result became not so
56 .te8 'it>e3 important - creativity came to the fore . . .
57 b4 d4 'Heaven knows, in this game I did every
thing in my powers. My opponent played
Of course, not 57 . . . .:xb4?? 58 .tf7 .
better - all praise to the winner! But I
58 97 l:tx97 conducted the attack without heed for the
59 <it>b3 ':95 circumstances, and in the end a person is
A good move , although not essentia l . The responsible for his actions, but not for their
immed iate 59 . . . d3 was possible, for exam result . . . Of course, it was crazy to allow the
ple, 60 .th5 .1:.g5 61 .td 1 1:[g 1 62 .th5 %:t h 1 black pawn to reach a2, but "he who has
(62 . . . d2 also wins) 6 3 .tg4 <it>f4 . never done anything reckless is less wise
Virtuoso Defence ctJ 1 75

than he thinks" (La Rochefoucauld). Is the g6-pawn (usually this is done by a


creative pleasure really worth less than pawn).
pitiful half points or even a whole point? And Possibly Mestel overlooked something here,
has not Caissa repaid me one hundredfold since he lost this position in four moves. And
for that glorious recklessness, which I since he thought for a whole hour over his
permitted myself not only in this game, but 26th move in an unsuccessful search for a
also others, which did not end so sadly?' defence, it is at the given moment, on the
(Sanakoev) 24th move, that the persistent reader should
When I was analysing this game I couldn't seek the last hope for Black. We will return
help remembering a colourful a rticle by Bent again to the diagram position, but first let us
Larsen, devoted to the same theme (it was see what happened in the game.
published in the 1 982 No.5 issue of the 24 . . . .:te8??
Danish magazine Skakbladet and trans 25 'ilg511 l:. e5
lated into Russian by Valery Murakhveri ) .
What else ? The threat was i.. xg6+ and h4-
Here for t h e readers is an extract from the
h5.
article.
26 CiJe71

Rivas Pastor - Mestel


Marbella 1 982

26 . . . "il g8
The main variation is 26. . . :1xg5 27 hxg5
When I annotated this game for the newspa "ikxe 7 28 fxe 7 iLe6 29 iLd1 g7 30 i.. a 4. Or
per 'Ekstrabladet', I was under the influence 29 iLe2 and 30 iLb5. Why I also mention
of analyses by Rivas and under pressure this second possibility, we will see later. The
from the editor Dinesen . . . And I believed d1-a4 diagonal could have been blocked!
that Black was doomed: 24 . . . gxh5? 25 "ilg5; Apart from this variation we will also analyse
24 . . . i.. e 6? 25 'Wig5 i.. xd5 26 i.. xg6+ fxg5 27 26. . . "ilf8 27 iLxg6+ h8 (27. . . fxg6 28 h5)
h5, and White wins (according to Rivas, 28 iL f5 i.. xf5 29 exf5 h7 30 CiJg8!!.
there can follow 27 . . . "ilxf6 28 hxg6+ g8 29 Perhaps it was this that Mestel overlooked.
"iixd5+ g7 30 : h 7+ xg6 31 "ikh5 mate However, it is not difficult to see 30 CiJg8, if
this in indeed pretty). I should mention one you think about how White can prevent the
instructive feature: the queen on g5 blocks defence 30. . :ikh6.
1 76 \t? Virtuoso Defence

27 i.. d1 h8 The idea of . . . b4-b3, which was constantly


Or 27 . . .'!:J.xe 7 28 h5. on the cards, in combination with the
28 "ikh6+ manoeuvre . . :ii e 8, which has now become
possible, is discovered in a desperate
Black resigned.
search of the position, when you see that all
the natural replies are unsatisfactory. In a
Very pretty. But let's return to the 24th move,
difficult moment you play 24 . . . b3!!, without
when Mestel still clearly had sufficient time
even examining 25 b 1 or 25 axb3.
for thought, and see what moves and what
ideas are contained in this position, apart After 25 b 1 (the king is on a light square!)
from 24 . . . gxh5, 24 . . . .i.e6 and 24 . . . 1:1e8. there is, of course, no point in taking the a2-
pawn. 25. . . "ike8 is good, but 25. . . .i. b 7 is
For example, 24. . . "ike8 is possible: the
also acceptable, in order to capture on e4
queen defends the g6-pawn, and also eyes
with check (25 b 1 ? .i. b 7!? 26 "ikg5 .i.xd5
the squares e5 (like the rook after 24 . . . fJ.e8)
27 .i.xg6+ ? fxg6 28 h5 .i.xe4+ 29 a 1
and e4. Say, 25 e7 c4 26 "ikg5 11b5. But
"iixf6).
here we already know the solution: 27
.i.xg6+! fxg6 28 h5. After 25 axb3 Black does after all have the
open a-file. There is also the defensive idea
Thus the defence 24 . . . "ike8 does not save
. . . 1:1b8xb3-h3, but unfortunately it does not
Black, but it was right to examine it: unusual
work. The simple 25. . . 11a8 forces the reply
moves sometimes lead to correct ideas. I
26 c2, and the king is again on a light
remember it being said about Reshevsky
square. 26. . . .i. b 7 is possible, or even
that he deliberately got into time-trouble,
26. . . c4, but why give White another pawn
after first studying all (!) the tactical subtle
for the exchange ?
ties of the position, and then played very
confidently with his flag about to fall. This is One could spend a long time in analysis. But
clearly an invention, since you cannot cover in a practical game the most likely course of
all the tactical subtleties, as new ones also events was 24... b3!/ 25 a3 "ik e8 t, and the
arise. It is better to think about something optimist playing White, if he has time, will
else: from the 1 7th move onwards, on every have a long think. Gradually his ears will
move there was the possibility of . . . b4-b3. turn red, his breathing will come difficult,
There you have it! Here Black is not and he will begin slightly shaking his knees
threatened with mate in two, and after - and the entire board.
24 . . . b3 the capture . . . b3xa2 is a serious Cool-headed defence saves many points. I
threat. And if 25 a3, then let's return to the have seen Mestel escape from worse
idea of . . . "ike8: 25. . . "ike8 26 e7 'ila4!. scrapes. I am sure that after 24 . . . b3 he
Unexpectedly Black obtains play. His queen would even have won the game. Attacking
can go to c4, d4 or e4, for example: 27 optimists are very bad at readjusting.
xg6? "ikc4+! or 27 .i.f3 "ikc4+ 28 b 1 The problem for commentators is that when
"ikc2+. White has a pawn for the exchange a game ends in a pretty rout for one of the
and some positional pluses, so that the sides, it can be difficult to give an objective
chances are roughly equal. commentary.
CZJ 1 77

Mark Dvoretsky

What l i es be h i nd a M istake

' To err is human'. It is obvious that no Bareev - Kasparov


I player is able to play faultlessly. E rrors Linares 1 992
are sometimes caused by the d ifficulty of
the problem being solved , or by chance
circumstances. But very ofte n , mistakes
conform to a pattern - they are the result of
certain playing or personal ity defects . This
applies not only to ordinary players, but also
to lead ing grandmasters , and even world
champions.
For all of us, the most d ifficult th ing is
performing successfully in unfamiliar situa
tions, in which we have insufficient experi
ence. To some extent this deficiency can be
cou nterbalanced by pu rposefu l training,
but, unfortunately, there are very few who
resort to this.
It is wel l known that Garry Kasparov's main Black has a choice between the exchange
strength is the breadth and depth of his of queens and the sharp knight sacrifice on
open ing preparatio n : for many years he e4 ( 1 7 .. .'ii c7? is anti-positional in view of 1 8
hi mself, the members of his reg ularly u p ttJa3 followed by l::t c 1 and ttJf2). I n Bareev's
dated tea m , and in recent times also opinion, 1 7 . . . 'i!Vxd2+ was correct. Too risky
powerfu l computers, have been tirelessly now is 1 8 Wxd2? ttJfxe4+ or 1 8 ii.xd2 (with
perfecting his opening arsenal . As a result the idea of developing the knight on a3)
he comparatively rarely finds h imself in 1 8 . . . ii.xg4! ( 1 8 . . . ttJfxe4 is less good : 1 9 fxe4
d ifficult positions and in this field he has very ttJxe4 20 ttJc3) 1 9 fxg4 ttJcxe4 (intending
limited experience. It is no surprise, there 20 . . . ttJxd5), and if 20 ii.f3 (20 ttJc3!? comes
fore , that defence is one of his vulnerable into consideration), then 20 . . . ttJxd2 and
poi nts. When he is forced to defend , 21 . . . . e4 . There would most probably have
Kasparov always does this as actively as followed 1 8 ttJxd2 b5 1 9 ttJf2 l::tfc8 with a
possible, immediately aiming to undertake roughly equal ending.
someth ing, to change sharply the character 'Each of us has his own style, his own way
of the play. This trait of his was poi nted out, of playing ' Bareev remarked in his lectu re.
in particular, by grandmaster Evgeny Bareev 'In Black's place I would have agreed to the
in a lecture which he read in 1 992 at a exchange of queens. But Kasparov does
session of the Dvoretsky-Yusu pov school not like positions in which he has no
for talented young players. counterplay. He did not want to go into a
1 78 What lies behind a M i stake

quiet and (as it seemed to him) slightly Anand - Kasparov


inferior endgame, and so he decided on a 9th game of the match
rather dubious piece sacrifice. '
17 . . . tDfxe4?1
18 fxe4 tDxe4
1 9 'iWxaS l:[xaS
20 tDc31
'A simple and very strong move. If 20. . . tDxc3
21 bxc3 11xd5, then 22 tDf2 with a subse
quent blockade on the light squares. Black
has three pawns for the knight, but without
counterplay this is insufficient - he needs
some dynamic factor, such as two con
nected passed pawns. ' (Bareev)
There followed : 20 ... tDg3 21 ':'g1 tDxe2 22
'it>xe2 e4 23 l:tac1 1 fS 24 gxfS 11xfS?1
(24 . . . e8) 2S tDf2 e8 26 ':h1 bS+ 27 27 l:IdS!
tDxbS l:[xbS 28 ':c8+ 'it>h7. And now,
It is obviously extremely dangerous to
instead of the game continuation 29 ':d1 ? accept the positional exchange sacrifice
':xb2+ 30 ':d2 a31 with an unclear position ,
offered by the I ndian grand master - the
Wh ite could have gai ned a decisive advan grouping which Wh ite creates in the centre
tage with the simple 29 l:tb1 ! l:tfxd5 30 b4! and on the q ueens ide is just too powerful.
axb3 3 1 axb3. Black should have played 28 . . . h 5 ! ? , i ntend
In many cases it is active defence which ing to meet 28 c7, attacking the e5-pawn ,
promises the best chances of success, but with 28 . . . 'iWe7! . Black's position would have
this is by no means always the case. Any remained u npleasant, but by no means lost.
one-sided approach is bad . Sometimes you 27 . . . tDxdS?
should calmly parry the opponent's threats,
28 exdS 'iWg6
patiently and accu rately solving the prob
lems which arise. A lack of flexibil ity in his Kasparov nevertheless took the rook. Why?
choice of playing methods makes a player I see the explanation as being that he
vul nerable. hi mself was hoping to obtain some activity.
The queen aims at the rook, and also at the
It is i nteresting that in his match against
c2- and d3-squares; he has the active move
Vishwanathan Anand (New York 1 995)
. . . e5--e4 , attacking the bishop . . . Alas, this is
Kasparov several ti mes chose the tactics of
all an illusion - the strategic pluses of
active defence in situations where they
White's position are far more important.
were completely inappropriate (true, in the
second half of the event Anand was demor 29 cS e4
alised and he was unable to pun ish his 30 e2 :eS
opponent for th is). 31 'iWd7 1 ':gS?
In the opinion of Alexander Chern i n , 31 . . . e3
was more tenacious, for example: 32 l:tf1
':'g5 33 d3 e2! 34 xe2 ':xg2 35 d3
:g 1 + ! 36 %1xg 1 'iWxd3, and the position
What lies behind a M i stake ltJ 1 79

remains sharp. 32 %l g 1 is stronger, but then nervous tension , agg ravated by the emo
instead of 32 . . . %lg5 (which leads to a tions provoked by the loss of the previous,
position from the game) Black can make 1 0th game. And on the other hand - a lack
another, more useful move - 32 . . . 'ii'c2 . of the habit of deeply and carefully checking
32 %lg1 e3 variations. Anand possesses a wonderful
33 d6 %lg3 i ntuition , and many of his decisions (some
times very d ifficult ones) are taken quickly,
34 'ii'x b7 'ii'e 6
but rapid ity of th i n king often goes badly with
35 h21 accu racy and precision in calculation .
Black resigned , since after 35 . . . 'ii'e 5 36
3 0 . . . :1xb4+ 31 a3. Wh ite was expecting to
'ii'x aB he is unable to make favourable use
emerge with the exchange for a pawn after,
of the d iscovered check.
say, 3 1 . . . .l:.bc4, but he overlooked a fearfu lly
strong counter-stroke: 31 . . Jlxc2 ! . He had to
Anand - Kasparov resign immediately in view of 32 %lxc2 l:Ib3+
1 1 th game of the match 33 a2 :e3+, when Black is two pawns up.
Here Kasparov was simply lucky - after all,
the move he made was in fact bad ! The
captu re on e7 leads by force to a double
rook ending with an extra pawn and excel
lent winning chances for White .
28 li'lxe7! :te8
29 li'ld5 xd5
30 b4!
The moves can also be interposed : 29 b4
axb4 30 axb4 l::t c4 31 li'ld 5 .
30 . . . axb4
31 axb4 %lc4
32 l:txd5
This position has been reached by force. Of
The e7-pawn is under attack. After the
course, it is too early to call a halt - a certain
normal 27 . . .fB Black would have retained
calcu lating tech nique (although not too
an acceptable, although slightly inferior
complicated ) is requ i red , in order to take the
position . I n stead of this Kasparov played
variation to its logical end. It is surprising
'actively' .
that neither Anand (during the game) nor
27 . . . e6?
Kasparov (in his analysis in Informator)
In the game his idea proved justified , si nce coped with this problem .
his opponent was tempted by a faulty
32 . . . %lxb4+
combination on the theme of double attack:
28 b4? axb4 29 axb4 %lc4 30 li'l b6?? If 32 . . . .l:.ecB? (suggested by Kasparov), then
33 c3! (weaker is 33 %le2 l::t x b4+ 34 c1
I should mention that Anand thought about
':c6 35 .l:.ed2 l:ta6) 33 . . . l::t xc3 34 :e2
this combination for only a few minutes. and
followed by 35 %lxb5.
the fatal knight move was altogether made
almost instantly. Why? On the one hand, 33 c3
what obviously told was the enormous Of cou rse, not 33 c1 ? f5 with equality.
1 80 What lies behind a Mistake

33 . . . 1:[c4+ 1:[xg5 the g2-pawn is lost, and after 34 g3


34 b3 l:tec8 gxh4 35 gxh4 1:[f5 - the f3-pawn . But Anand
34 .. .f5!? is more tenacious , but after 35 finds an excellent rook manoeuvre, which
l:txb5 1:[d4 36 c3 Black's position remains refutes the opponent's idea .
difficult. 34 1:[a8 1 gxh4
35 11e2 34 . . . d7! 35 g3 (of cou rse, not 35 g4??
Black loses one of his pawns, and his gxh4 36 1:[h8 h3) 35 . . .1:[f5 36 1:[h8 .l:[xf3 37
chances of saving the game are problem hxg5 l:txg3 38 1:[xh5 c6 was stronger -
atic. however, here too Black would stil l have
been a pawn down .
I should mention that if Anand had accu
rately calculated this variation, he would 35 1:1e8+! d7
al most certainly have gone in for it. Here 35 . . . f5 36 1:[e4 g5 37 a4!? was even
Wh ite's advantage is g reater than in the worse for Black.
position with an exchange advantage, which 36 1:[e4 c3
he was hoping to obtain by playing 28 b4? (it
is most probably drawn ).

Anand - Kasparov
1 7th game of the match

Here Wh ite had to make a d ifficult choice.


Anand preferred simply to remain a pawn
up.
37 ':xh4? 1 cxb2+
38 xb2 1Ig 5
Of course, Black should have safeg uarded 39 a4
his pawn : 32 . . . a6. It is important that if 33 But in the resulting position Wh ite's pawns
d2 (with the positional threat of c3-d4) are broken and the opponent reta ins real
he has 33 . . . c3+ ! 34 bxc3 :tc5, after which chances of saving the game (remember the
the game should end i n a draw. But aga i n , d rawing tendencies of rook endings). I n the
instead o f a normal cou rse Kasparov end Kasparov managed to gain a d raw.
chooses an 'active' one. H ad Anand been a l ittle more self-confident
32 . . . 1:[c5? (he was appreciably oppressed by the
33 %1xa7 g5 burden of failure in the preced ing games),
It appears that Black is alright: after 34 hxg5 he would most probably have chosen 37 b4!
What lies behind a M i stake ttJ 181

lIg5 3 8 lIxh4 lIxg2 3 9 'iti>b1 . Here , for the pawn to d4 when the king is on b 1 .
moment, Wh ite is not a pawn up, but he has The moral to be d rawn from these examples
a serious positional plus - two connected (the list of them could have been extended )
passed pawns on the q ueenside. If the is obvious. For a player of any standard it is
wh ite king ma nages to get to b3 (as i n the important to make a thorough analysis of his
variation 39 . . ..:f2 40 ':xh5!? ':'xf3 41 'iti>a2), own games, and disclose the latent, deep
the position will certainly become won . causes of the mistakes he has made, as this
Black's only counter-chance (although I always serves as the first step towards their
doubt whether it is sufficient) is to push his d- elimination .
1 82 <;t>

PART VI

Mark Dvoretsky

Analysis of a Game

Wgame, played by Sasha Chernosvitov.


e are going t o look a t an interesting 1 0 b3, 1 0 d5 and 1 0 l:r. d 1 have also been
played . In his commentary on the game
He annotated it in great deta i l : he gave Sasha described in detail his views on the
numerous variations, and described what theory of the g iven variation . But since his
he was thinking about d u ring the game, opening conceptions are to a certa i n extent
what he was afraid of and what he over a personal matter, there is no need to
looked . His analysis is not error-free. Of discuss them here. We are more interested
course, in such cases mistakes are inevita in the problems which arose later.
ble - after all, when you are examining 10 . . . xcS
complicated variations, it is easy at some
1 1 h3?!
point to become entang led . But beh ind the
mistakes made, defin ite deficiencies in 1 1 e4 is more critical, althoug h then 1 1 . . . lDg4
thinking, in the approach to the game, can (th reatening 1 2 . . . lDd4) or the immed iate
sometimes be see n . The reason I have 1 1 . . . lDd4 has to be reckoned with . Wh ite
chosen this particu lar game for analysis wants to adva nce his e-pawn in comfort, but
was because some of the omissions, both in in the open ing every tempo counts , and with
the moves and in the comments , seemed to such slow play he can no longer expect an
me to be instructive . advantage.
11 . . . 0-0
Den isov - C hernosvitov 1 2 e4 lDd41
Moscow J u n ior Championsh i p 1 99 1 1 3 lDxd4 xd4
Queen 's Gambit Accepted The opening stage has concluded in Black's
1 d4 dS favour. He controls the central squares, the
2 c4 dxc4 bishop on c4 is running up against the e6-
pawn, the wh ite e-pawn has been halted and
3 e3 lDf6
in some variations it can even come under
4 xc4 e6 attack. It only remains for Black to develop
S lDf3 cS his light-square bishop, and his position will
6 0-0 a6 be preferable.
7 a4 lDc6 14 d2?!
8 'ife2 'ifc7 Chernosvitov recommends 14 d3, in order
9 lDc3 d6 to prepare the development of the bishop at
1 0 dxcS e3.
Analysis of a Game tZJ 1 83

14 . . . b6 xg7 21 g3 with equality.


1 5 llac1 b7 18 . . . l:.fd8
1 6 d3 The wh ite pieces on the d-file are beg i n n i ng
to hang . If 1 9 c2 (with the idea of h2 and
g2-g 3 , or 'ii'f3-g3) Chernosvitov suggests
replying 1 9 . . . g5! 20 'ii'f3 e5 2 1 ':'e2 l:.d4 . If
1 9 ttJd 1 (preparing 20 c3), then 1 9 . . . b5 20
axb5 axb5 21 c3 e5 22 'ii'f3 xc3 23 bxc3
(23 ttJxc3 b4) 23 . . . ':'a2, and Black has an
appreciable advantage.
Sasha considers the comparatively best
defence to be 1 9 'ii'f3! ? xf2 20 ':'xf2 ':xd3
21 'ii'x d3 'ii'xf2 22 'ii'e 3!? 'ii'x e3 23 xe3
ttJxe4 (23 . . . ttJd7 24 : d 1 ) 24 ttJxe4 xe4 25
xb6 with chances of equalising (for exam
ple, 25 .. .f5 26 ':'c1 f7 27 d4). However, if
the concluding position of this variation
does not satisfy Black, he can retain the
The threat of 1 7 ttJd5 (or 1 7 ttJb5) is very advantage with the qu iet move 1 9 . . . e5 ! ? ,
obvious. But what will Wh ite do after the seeing as 20 ttJd 1 w i l l b e met b y a little
queen moves off the c-file? Perhaps ex combinatio n : 20 . . . J:txd3! 2 1 'ii'x d3 ttJxe4
change the dark-square bishops by 1 7 e3, with the threats 22 . . Jld8 and 22 . . . ttJg3+.
but most probably prepare f2-f4 by movi ng 1 9 h2?
the king to h 1 . Is it not possible to make it It is a well-known tru ism that everyth ing
d ifficult for the opponent to carry out his should be done at the right time. 'A move
plans? made one move later is often a mistake. '
16 . . . 'ii' g 3! Now the capture with the bishop o n f2 ,
An excellent manoeuvre! 16 . . . 'ii'e 5 1 7 h1 which we saw in the last variation, gains
ttJh5 1 8 'ii'g 4 was much weaker. greatly in strength .
1 7 h1 'ii' h 4
The queen is excellently placed at h4. From
here it presses on the e4-pawn and simulta
neously eyes the king. In some cases the
knight will jump to g4, while if 1 8 f4 , then
1 8 . . . ttJh5 1 9 e1 ttJg3+ 20 xg3 'ii'x g3 with
the better chances for Black.
1 8 l:tce1 ? !
Now Wh ite's position becomes d ifficu lt. He
should have decided on a second succes
sive king move: 1 8 h2 ! . The tactical basis
of it is the elegant variation 1 8 . . . xc3 1 9
xc3 (weaker is 1 9 g 3 e5 20 f4 ttJg4+! 2 1
g2 'ii'e 7) 1 9 . . . ttJxe4 (after 1 9 . . . xe4 20
xa6 White is only slig htly worse) 20 xg7! 19 . . . xf2 1
1 84 Analysis of a Game

20 1lxf2 l:txd3 u n necessary sharpening of the play. It can


21 'iWxd3 lDg4+ ! happen that your advantage hangs by a
thread , and to maintain it you have to exert
22 g1 'ifxf2+
yourself and deeply calculate complicated
23 h1 variations. But here the situation is clearly
Chernosvitov played the first half of the d ifferent: Black is a sound pawn to the good
game very strongly, completely outplayed and he has a safe way of retaining all the
his opponent, and gained an advantage advantages of his position .
qu ite sufficient for a wi n . But from this 23 . . . 'iWh4?
moment it was as though a substitute had 24 'iid 7!
taken his place.
What should Black do now?
What should Black play now? 23 . . . 'iWh4
suggests itself, but then the cou nterattack
24 'iWd7 is unpleasant. Therefore in the fi rst
instance the safer conti nuation 23 . . . lDe5!
should be considered . The reply 24 'iWe3 is
forced . Let us conti nue the variation :
2 4 . . . 'iWxe3 2 5 i.xe3 lDc4 2 6 i. c 1 (26 i.f2
lDxb2 27 i.xb6 ':'c8 is also cheerless, for
example: 28 i.d4 I[c4 29 lIb1 ':xd4 30
J:txb2 l:td7) 26 . . . lId8 27 b3 lDe5 28 i.e3 (28
i.f4 lDd3) 28 . . . lId3 29 i.xb6 J:txc3 30 i.d4
l1xb3 31 i.xe5 f6 followed by . . . e6-e5, and
Black converts his extra pawn without
particular d ifficulty.
Why d id n 't Sasha play this? Through inertia
he wanted to conti nue the attack, fearing 24 . . Jlb8? 25 :f1 is completely bad for
that i n the endgame Wh ite would save Black, and therefore he must choose be
hi mself thanks to the opposite-colour bish tween 24 . . Jld8 and 24 . . . lDf2+ .
ops. But more importa ntly, he underesti On the previous move Sash a committed a
mated the opponent's threat, assuming that typical psychological mistake: realising that
the q ueen move to d7 was not possible due he had gained a decisive advantage, he
to the loss of the e4-pawn . But in fact after relaxed and played carelessly. Now he
23 . . :tIi'h4 24 'iWd7 lDf2+ 25 h2 i.xe4 after sensed that things were by no means as
exchanging on e4 Wh ite can then simultane simple as they appeared earlier. And here
ously attack two pieces with 'iWc6 or 'iWb7. came a second psychological mistake (re
One can lau nch into such adventures only member, incidentally, Tarrasch's famous
on the basis of very accurate calculation. maxim: ' M istakes never occur singly' . ) As
Here I would almost certainly have called a often happens, the sudden d ifficulties spoiled
halt and rejected the entire variation (if I had his mood and prevented h i m from calmly
not done this a move earl ier, on seeing the analysing the variations, which , however,
move 24 'iWd7). are now very complicated . Black also failed
I n the conversion of an advantage you to solve this problem later, i n analysis at
should try to strictly control all the oppo home.
nent's active possibil ities, not allowing any Chernosvitov rejected 24 . . .lId8 25 'iWxb7
Analysis of a Game ttJ 1 85

':xd2 because of 26 1:.e2. But here it is too the position would not be changed (if 32
early to stop the calculation: Black can play 'itt g 3 , then 32 . . . 'iVe 1 + followed by 33 . . . a5 is
26 ... h51 (or 26 . . . h6!), opening an escape possible, if there is noth ing better).
square for the king . Black's rook is immune And in the event of 28 'itt h 2 there is an
and his th reats look rather dangerous. excellent knight sacrifice : 28 ... lbxh31 29
I ncidentally, in similar situations the pawn is gxh3 l:[d31 with an irresistible attack. For
usually advanced not one square , but two , example, 30 'iVb8+ 'itt h 7 3 1 'itt g 1 ':xh3 32
since it may come in useful i n the attack. I n .l:th2 'iVg5+ ! , or 30 'itt g 2 ':xh3 31 'ii'd 6 'itt h 7 ! ,
the g iven instance the two moves are intending 32 . . .f6 and 33 . . . e5 (31 . . . 'ii'g 4+ 32
roughly equ ivalent. 'ittf2 'iVg5! with the same th reat of . . . e6-e5
The opponent's reply is obvious: 27 'ii'x b6 is equally good ).
(27 'ii'x a6? l:1c2! is bad ). The next problem is Attempts to avoid mate lead to a lost
this: how can Black exploit the advantages endgame for Wh ite : 29 ':xd2 lbf4+ 30 'itt g 1
of his position? 'iVe 1 + 3 1 'itt h 2 'ii'x d2 , o r 2 9 'ii' b 8+ 'itt h7 30
'iVg3 'iVxg3+ 31 'itt x g3 ':d3+ ! 32 'itt h 2 lbg5.
I n stead of 26 ':e2 we must consider 26
.l:tf1 ! .

The simplest solution - 27 . . . .l:.xe2 28 lbxe2


'iVe 1 + 29 lbg 1 lbf2+ 30 'itt h 2 lbxe4 - does
not seem convincing to me. After 3 1 lbf3 (31
'ifb8+ ! 'itt h 7 32 lbf3 is even more accurate) After 26 . . . f6 there i s the strong reply 27
the coordination of the black pieces is 'iVe? ! . And if 26 ... h5!?, then White conti n
disrupted somewhat, and Wh ite is threaten ues 27 'ii'x f7+ 'itt h 7 28 'ii'xe6 lbe3 29 lbd5!
ing the rapid advance of his pawn on the lbxf1 30 'ii'f5+ 'itt h 6 31 'ii'e 6+ g6 32 'iVg8 1 ,
queenside. For example, 31 lbf3 'ii'g 3+ 32 and the game inevitably ends in perpetual
'itt g 1 'iVd6 33 'iVxd6 lbxd6 34 b4 followed by check.
lbd4 .
How can Black's play be improved? Grand
Let us try 27 ... lbf2+ ! . If 28 'itt g 1 ? , then master Dolmatov found a rather unexpected
28 . . . lbxh3+! 29 gxh3 :xe2 30 lbxe2 'iVe 1 + idea : after 24 . . .l:ld8 25 'ii'x b7 he suggested
3 1 'itt g 2 'ii'xe2+, and the queen endgame is avoiding the capture of the bishop in favour
easily won . I n this variation the pawn is of 25 ... h6!.
better placed at h5 than at h6. However,
even with the pawn on h6 the evaluation of (see diagram)
1 86 Analysis of a Game

achieve noth ing after 24 ... liJf2+ 25 h2


iLxe4 , nevertheless he quickly rejected
24 . . . .l:.d8. Apparently what showed itself
here was the psycholog ical effect of the
mistake made earl ier, about which I have
already spoke n .
24 . . . liJf2+
25 h2 iLxe4?
As grandmaster Evgeny Bareev poi nted
out, it was not yet too late to switch to attack
by 25 . . . :d8! 26 'ili'xb7 lIxd2. Wh ite loses
after 27 l:r.f1 ? 'ii'f4+ 28 g 1 liJxh3+ or 27
'ifb8+? l:r.d8 28 _g3 liJg4+ ! ' And in the
event of 27 :e2 Black gains an advantage
26 l:le2 l:lxd2 transposes into the variation
by 27 . . . h6 (or 27 . . . h5) 28 'iVxa6 liJg4+ 29
25 .. J:txd2 26 l:le2 h6, in which , accord ing to
g 1 (29 h 1 ? :tc2 ! ) 29 . . . liJe3 ! , for exam
our analysis, Black's attack achieves its aim
ple: 30 llxd2 'ili'e 1 + 3 1 h2 liJf1 + etc. The
(true, we put the pawn on h5, but this is of no
outcome of the complications arising after
sign ificance). And in the event of 26 l:lf1
27 . . . g5!? is less clear: 28 'ili'xb6! (28 'iVxa6? !
Black is no longer obliged to captu re on d2 -
l:txe2 29 'ii'xe2 g4 ! ) 28 . . . liJxh3!? 29 gxh3
26 . . . liJf2+! 27 l:lxf2 'ii'xf2 is much stronger,
l:td3 30 g 1 ! lIxh3 31 'iVb8+ g7 32 g2.
when the wh ite bishop cannot escape.
I ncidentally, with this move order, as in the
The best defence is 26 liJd51 exd5 27 :f1 (if
variation by Dolmatov analysed earl ier,
27 exd5, then 27 . . . liJf2+ and 28 . . . liJxh3! ) ,
Black would have deprived his opponent of
b u t here too Black retains a b i g advantage,
the best defence l:tf1 ! .
by continuing 27 . . . liJf2+ 28 h2 (28 g 1
liJxh3+ 29 gxh3 'ii'g 3+ 30 h 1 'ii'x h3+ 3 1 26 liJxe4?
g 1 'ii'g 4+! 3 2 h 1 'ii' h 5+ and 3 3 . . . dxe4 i s Wh ite misses his chance. He should have
hopeless for White) 2 8. . . dxe4 (weaker i s played 26 :txe4 ! liJxe4 27 'ii'c6 'ifg3+ (after
2 8 . . . liJxe4 2 9 'ii' x f7+ h 7 3 0 iLf4) 2 9 iLe1 27 . . . l:td8 28 liJxe4 the bishop on d2 is
'ii'f4+ 30 g1 e3 31 'ii'x a6 :d1 32 'ii'e2 defended - this is why he should take with
'iix a4 (32 . . . 'ifd4 33 iLxf2 exf2+ 34 'ifxf2 the rook, not the knight) 28 g 1 .l:.d8 29
l:lxf1 + 35 xf1 'ii'd 1 + 36 'ii'e 1 'ili'xa4 is also liJxe4. After 29 .. :"e5 there fol lows 30 'ii'x b6
good , with an extra pawn in a queen 'ili'd4+?! 31 iLe3, while if 29 . . . 'ii' b 8 (as
endgame) 33 'ii'x e3 liJd3 . Chernosvitov wanted to play), then 30 iLe3!
There are good g rounds for once again b5 31 a5 h6 32 'ili'b6 (32 'ii'x a6? 'ii'e 5)
remembering Tarrasch , who maintained 32 .. J:ld 1 + 33 f2 'if e5 34 liJd2 with
that you should resort to a combi nation in excellent drawing chances.
order to repair a mistake committed earlier. I In endings a rook is often stronger than a
would remark, however, that at the board it bishop and knig ht, if it is able to break
is not so d ifficult to decide to go in for tactical through at the right time into the opponent's
complications, if the method of elimination is position to attack the enemy q ueenside
employed - after first satisfying yourself that pawns and create there an outside passed
all other tries are unpromising. Strangely pawn. But here the endgame is stil l a long
enoug h , although Sash a saw that he would way off, and besides, it is not only Wh ite's
Analysis of a Game ltJ 1 87

q ueenside pawns that are under fire , but 40 . . . 1:.d2+? ! ) 41 1:.g2 1:.d 1 or 4 1 'ii'x d6
also Black's. 'it'xc2+ 42 g 1 'ii' b 1 + and 43 . . . 'ii'x b5.
26 . . . tDxe4 In the game Black decided to go i nto an
27 'it'c6 'ii'g 3+ endgame. An incorrect assessment of the
28 g1 'ii'f2+ position! Even if the endgame is won , it is
clear that with the q ueens on it would be
29 h2 'ii'g 3+
much simpler to convert the advantage.
It is usefu l , by repeati ng moves, to save time
33 . . . 'ii'd 3?
on the clock.
30 g1 'it'f2+ 34 1:.e3 'ii'd 6+
31 h2 l:.d8 35 'it'xd6 ':'xd6
32 l:txe4 'ii'x d2 An interesting rook ending has arise n . In the
33 'it'xb6 analysis of it, numerous questions occurred
to me; the answers to them were either
lacking i n the commentary, or did not seem
convi ncing.

Black is a sound pawn to the good . Which is


more correct: to exchange or retai n the
queens?
Of cou rse, it is better to keep the queens on. 36 .l:[b3
After all, the wh ite king is exposed and it can White prepares a4-a5 and 1:.b6. A tempti ng
easily come under attack. Chernosvitov idea , but the more standard plan beginn ing
illustrates this assessment with the follow with 36 b4 should also have been consid
ing sample variations: 33 . . . h6 34 b4 (34 ered (the rook will stand beh ind the passed
'iVxa6 'it'xb2) 34 . . . 1:.d6 35 'it'b8+ h7 36 b5 pawn). After 36 . . . f8 37 b5 axb5 38 axb5
axb5 37 axb5 f5! 38 ltc4 e5 (if 38 . . . 'ii'd 5 the e7 39 1:.b3 it is bad to play 39 . . . d7? 40
only defence is 39 1:.c3! 'ii'e 5+ 40 ':'g3) 39 b6 c8 in view of 41 l:lc3+ b8 42 1:.c7 .
l:.c2 'iVf4+ 40 g3 1:.d2+ 41 1:.xd2 'ii'x d2+ 42 39 . . . l:tb6 is correct, when Black most prob
g 1 'iVe3+ 43 g2 e4 44 'ii'f4? 'ii'e 2+ 45 ably wins, although it is not so simple. If his
g 1 'ii'x b5. king goes to c5, there follows 1:.c3+ ; this
It is unclear whether Black can win after 44 means that he will have to waste a tempo on
b6, but this is not so important: he can . . . 1:. b6-b 7 . During this time White will create
ach ieve his goal with 40 . . . 'ii'e4! (instead of cou nterplay on the kingside. How? Wel l , for
1 88 Analysis of a Game

example, by 40 g3 l:tb 7 41 f4 d6 42 g4 rook is excellently placed to the rear of the


f6 43 g5 c5 44 l:te3, or 42 . . . c5 43 ':c3+ passed a-pawn . I don't see what can be
xb5 44 e5 , intending l:tc8-g8. done to oppose the advance of the pawn
36 . . . g6? armada on the kingside. For example:
42 . . . h5 43 a6 l:ta4 44 g3 h4+ 45 f3 e5 46
A very strange move , on which Sash a
':'a8 (46 e3 l:ta2) 46 . . . f5 47 a7 l:ta3+ 48
makes no comment. It is clear that Black will
f2 g5 and 49 . . . f4 . Note the good position
have to advance his kingside pawns, so why
of the f7-pawn on its i n itial square - after 49
not advance the pawn two squares, why
':'f8 ':'xa7 it will be defended by the rook.
waste a tempo? It is qu ite possible that a
48 .. .f6? would be a typical mistake - now
race will develop, in which every tempo will
after . . . f5-f4 the reply l:ta8-f8 gains in
count. Even if the delay does not affect the
strength .
assessment here, the next time it wil l .
37 . . . g7
O f course, 3 6 . . . g 5 ! ? was stronger than the
38 a5 l:td2
move in the game. 36 .. .f5!? also looks
tempti ng, intending to bring the king out to f6 Black's other plan is to bring his king to the
followed by . . . h7-h5-h4 , and . . . e6-e5-e4 . centre of the board . But in this case he will
For example: 37 a5 (37 l::t b 7 a5!?) 37 . . . f7 have to g ive u p one or two pawns on the
38 l:t b6 e7! (gaining another tempo) 39 kingside. Here is Chernosvitov's analysis:
:b7+ f6 40 l:tb6 l::t d 2 41 b4 l:1d4 . 38 . . . f6 39 l:lb6 e5 40 l:tb7 f5 4 1 l:txh7
'ud5 (4 1 . . .d4 42 l:tb7 e5 43 l:tb6 c5 44
37 :b8+
l:txd6 xd6 45 g4! fxg4 46 hxg4 c5 47
Here Chernosvitov makes the following g2 , and the pawn endgame is d rawn ) 42
comment: 'It would appear that White could b4 l:.b5 43 ':'a7 ':'xb4 44 l:txa6 l:ta4 45 l:ta8
have immediately gained a draw by 37 a5 f4 46 a6 e5 47 a7 e4 (47 . . . g5 48 g3+ e4
:d5 38 :b8+ g7 39 b4 f6 40 l:tb6 l:td4 49 11g8 or 48 . . . f3 49 l:tf8) 48 l:tg8 l:txa7 49
4 1 ':'xa6 l:txb4 42 l:ta 7' . l:txg6 e3 50 ':'e6 with a draw.
This variation is interesting, but by no means
forced . At the very end instead of 49 . . . e3?
there is the far stronger 49 . . . e3 ! ' I am not
sure that White can save himself here - the
e-pawn really is too strong . On the other
hand , it is not altogether clear why he
i n itially wasted time advancing his a-pawn,
and only then went for the g6-pawn . In reply
to 45 . . . f4 either 46 :a6 e5 47 l:txg6 or 46
l:te8 e5 47 g3+ suggests itself.
39 g3 f6
40 b4 ':'b2
41 l:tb6 e5?
It is more natural to advance the kingside
Roug hly such a position can arise in many pawns: 4 1 . . . h5 or first 4 1 . . .g5. I ncidentally,
variations and its assessment is important after . . . h7-h5 the king move to e5 gains
for the correct understanding of the entire in strength - since when the wh ite rook
endgame. But is it really d rawn? The black steps onto the 7th rank, Black simply replies
Analysis of a Game ltJ 1 89

. . . f7-f5 and the h-pawn will no longer be


under attack.
42 1:[b7
Chernosvitov comments: ' Neither White nor
Black saw the "two-mover" 42 ':'xa6 1:[xb4
43 1:Ia7 with a draw' . Wel l , we have already
talked about such a position - in fact after
43 . . .f6 Black should most probably wi n .
43 . . .f5 4 4 ':'xh7 ':'a4 4 5 ':'a7 g5 followed by
. . .1::t a3+ is also tempti ng.
When analysing endings it is very important
to make a correct assessment of key
positions, wh ich are reached from different
variations. Sash a made a mistake in his
44 l:th4+?
assessment of one such position , and as a
result his perception of the entire endgame A decisive loss of time. And meanwhile it is
was distorted . possible that there was no longer a win after
any of the 'normal' conti nuations: 44 1:Ia7
I should mention that Wh ite does not have
f4+ 45 h4 l::t x g2 46 l::tx a6, 44 .l:r.g7 f4+ (or
time to take his king across to the q ueenside.
44 . . . l:txb4 45 ltxg6 e5 46 h4) 45 h2 f3
Here are some sample variations: 42 f3 f5
46 :txg6 e5 47 %:tf3 (47 g 1 ), or 44 h4 f4+
43 e3 g5 44 d3 f4 45 c3 l:txg2 46 ':'xa6
45 h3 e5 46 ':a7 e3 47 l::tx a6 e4 48
f3 47 1:[a8 (47 d3 .:tb2 or 47 . . . h5) 47 . . . h5
%lxg6.
48 a6 :a2 49 b3 :'xa6! 50 :txa6 f2 51 :a 1
g4 52 hxg4 hxg4 53 b5 d5! 54 :1c1 g3 55 44 . . . e3
b6 g2 56 b 7 f1 'it' 57 b8'ii' 'it'xc1 , or 48 b5 :1g 1 45 1:1c4 e5
49 :f8 e4 50 b6 (noth ing is changed by 50 46 l:tc3+
c2 l:ta1 or 50 b2 l:.d 1 5 1 b6 %:td5)
Wh ite also loses after 46 l::t c6 f4+ 47 h2 e4
50 .. J:tb1 51 c2 l:.b5 52 d2 (52 a6 ':'xb6
(th reatening 48 .. .f3) 48 g 1 l::t b 1 + 49 h2
53 a 7 .1:.a6 54 a8'i1i' + l::tx a 8 55 %:txa8 f2 56
l::t x b4 followed by . . . f2 and . . . e4--e3.
':f8 e3) 52 . . . g4 53 hxg4 hxg4 54 a6 %lb2+!
55 d1 (55 e1 e3) 55 . . . g3 56 a7 1:[a2. 46 . . . e2

42 . . . f5 47 ':'c6 1:[xb4

43 l::t x h7 48 :1xg6

The delay i n playing . . . h7-h5 has told. The 48 %1e6 f4+ 49 g4 f3+ 50 g3 f2 51 l::t x e5+
outcome of the game is now in q uestion . f1 , and the f2-pawn inevitably promotes to
a quee n .
43 . . . e4
48 . . . f4+
49 h2 e4
50 :1xa6 e3
51 l:tb6 11a4
52 a6 f2
Wh ite resig ned .
1 90 Analysis of a Game

When analysing complicated endings, we follows 4 . . . l::t h 3+! 5 gxh3 g3+) 4 . . . l::t x a7
sometimes stumble upon positions which , (4 . . . g3! 5 as" f3! mates more qu ickly) 5
possibly, do not arise by force, but which are l::txe4 l::t a 1 + 6 'iti>h2 g3+ 7 'iti>h3 f3 S :f4 l::t h 1 +
i nteresting i n their own right. One such 9 'iti>g4 'iti>xg2 1 0 l::t xf3 l::t x h4+ .
position is examined by Chernosvitov.
It is time to sum u p . Chernosvitov played the
second half of the game uncertainly, and
also his commentary, although detailed,
was not too successfu l . Here two serious
deficiencies of his are clearly see n :
1 ) Weak conversion o f an advantage.
Remember: in a winning position Black
i ncorrectly allowed his opponent to sharpen
the play. Then he did not even try to figure
out the resulting complications. Finally, after
incorrectly resolving an exchanging prob
lem, he went into an endgame instead of
playing for mate . Later any methodical
player would surely have advanced his g
pawn two squares, instead of one, and he
I n his opin ion this is a position of mutual would also have chosen an appropriate
zugzwang. Wel l , it is not hard to see that if it moment to advance his h-pawn , whereas
is White to move he loses (1 g3 f3 , 1 h4 g4, Sasha left it at h 7, where it was lost. As a
or 1 l::t g S ':xa7 2 ':xg5 'iti>f2 3 l::t f5 f3 4 gxf3 result, the opponent gained real chances of
e3). But I can not agree with the conclusion saving the game.
that with Black to move it is a draw. The 2) Uncertain orientation in rook endings.
wh ite king is really very awkwardly placed . Chernosvitov overlooked some typical ideas
Black plays 1 . . . :ta2 2 'iti>g 1 l:Xa6! 3 'iti> h 1 (3 and plans, and his general assessments
'iti>h2 ::t a 1 ) 3 . . . g4! 4 hxg4 'iti>f2 5 g5 'iti>g3 with and specific recommendations often proved
unavoidable mate. incorrect.
There is also other, more spectacu lar Sasha can be advised to make a serious
solution : 1 . . . 'iti>f2 2 .:teS g4! 3 h4 (3 hxg4 study of rook endings, and even better - of
lba7) 3 . . . l:ta3! (another way is 3 .. .f3! 4 as. the theory and tech nique of the endgame as
l:th 1 + ! ! ) 4 'iti> h 1 (if 4 as. or 4 l:txe4 there a whole.
ltJ 1 91

Artur Yusu pov

C reative Ach i evements of P u p i ls


from the School

A t the end of the book it has become an But this is already a serious mistake.
/""\e stablished tradition to g ive examples Accord ing to theory, better is 1 0 .. .fxe6 1 1
of play by pupils from the school (their ages dxe6 'ike7 1 2 t'i)d5 'ikxe6+ 1 3 'ili'e2 'it'xe2+
are g iven in brackets). The jun iors played 1 4 .txe2 0-0 1 5 t'i)c7 t'i)c6 1 6 t'i)xa8 t'i) b4 1 7
and annotated a whole series of interesting t'i)f3 t'i)c2+ 1 8 d 1 t'i)xa 1 1 9 .tc4+ h8, as
games, some of wh ich , with slight correc in the games Shereshevsky-Semenyu k,
tions in the analysis, will now be offered to Vilnius 1 974 , and Lputian-Magerramov,
the judgement of the readers. The author Beltsy 1 979.
faced a d ifficult problem, since nearly every 11 .t b5+ f8
young player has good examples of attack 1 1 . . . e 7 is also dangerous in view of 1 2
ing play. Th is is not surprising : attack, risk .tf4 fxe6 1 3 d6+ f7 1 4 t'i)f3 .
and imagination are natu rally associated
1 2 t'i)f3 fxe6
with youth . However, the games chosen
1 2 . . . a6 looks somewhat more accu rate,
speak for themselves.
although after 1 3 .te2 fxe6 1 4 0-0 exd5 1 5
t'i)g5 g8 ( 1 5 . . . .tf5?? 1 6 ':xf5 is completely
Boguslavsky ( 1 4) - Lepi n
bad ; Black also loses after 1 5 . . . d4 1 6 'iVb3
Moscow 1 989 'ikd7 1 7 .tg4) 1 6 .tc4 b5 1 7 .txd5+ t'i)xd5
Modern Benoni 1 8 t'i)f7 White has a very promising position .
1 d4 t'i)f6 1 3 0-0 exd5
2 c4 e6 14 t'i)g5 g8
3 t'i)c3 c5 1 4 . . . h6 came into consideration, but in this
4 d5 exd5 case after 1 5 'ikxd5 'ili'xd5 16 t'i)xd5 hxg5 1 7
5 cxd5 d6 .txg5 Wh ite regains the piece and reta ins
the in itiative in the endgame. Now, however,
6 e4 g6
the king's rook is shut in the corner, and
7 f4 .tg7
Wh ite is able to carry out his attack al most
8 e5 u n h i ndered .
This is typical of Maxim's style: already in 1 5 t'i)xd5
the open ing Wh ite chooses the sharpest
Also not bad was 1 5 .tc4 ! ? b5 1 6 .txd5+
conti nuatio n .
t'i)xd5 1 7 t'i)f7 with the idea of answering
8. . . dxe5 1 7 . . . 'ikd7 with 1 8 t'i)h6+ .txh6 1 9 .txh6,
The alternative is the immed iate 8 . . . t'i)fd7. with decisive threats.
9 fxe5 t'i)fd7 15 . . . t'i)xd5
1 0 e6 t'i)f6? Usually when defending the king you should
1 92 Creative Achievements of Pupils from the School

aim for the exchange of queens. In the given A fantastic position , where Wh ite has only
instance this would not have brought any one piece for the quee n , but one of the
particu lar relief: 1 5 . . . 'ii'x d5 1 S 'ii'x d5+ lZ'lxd5 opponent's rooks is out of play and his king
1 7 ..tc4 ..td4+ 1 8 'ii? h 1 'ii? g 7 1 9 ..txd5 fU8 20 is in a mating net. The fol lowing variations
':xf8 (20 lZ'lf7!? is also not bad ) 20 . . . 'ii? xf8 2 1 are based on Boguslavsky's analysis.
lZ'leS+ ..txeS 2 2 ..txb7, winning material. A) 20 . . . 'ii'f7 loses immediately to 21 ..tc4;
16 lZ'lf7 B) 20 ... lZ'ld7 2 1 :ae1 'ii' d 5 (or 2 1 . . . 'ii'x e 1 22
..tc4+) 22 1:.e7 'ii'd 4+ 23 'ii? h 1 lZ'lfS (23 ... lZ'lbS
24 ..te8) 24 ':xfS with unavoidable mate;
C) 20 . . . lZ'lcs 2 1 :ae 1
C 1 ) 2 1 . . . 'iVxa2? 22 1:.fS! (not 22 ..txcS? bxcS
23 :e7 'iix b2 24 :ef7 because of 24 . . :iWa 1 ! )
2 2 . . . 'ii'f7 23 :ef1 lZ'le5 2 4 llxf7 lZ'lxf7 25
..tc4 and wins;
C2) 21 . . . 'iIi'd5 22 b3 lZ'la5 (if 22 . . . lZ'le5, then
23 l:txe5) 23 ':'e7 'ii'd 4+ 24 'ii? h 1 , and Black
can not parry the threat of 25 ..te8;
C3) U nfortu nately, Maxim does not consider
the best defence: 21 . . . 'iVf7 ! . Here noth ing
decisive is apparent. For example, 22 :txf7
'ii? xf7 23 ..tc4+ 'ii? fS 24 .l:.f1 + 'ii? e 5 25 g7+
'ii? d S with eq uality, or 22 b3 lZ'la5 (but not
16 . . . ..td4+?
22 . . . 'iVxf1 +? 23 ':xf1 lZ'le5 24 1:.e 1 as 25
In a d ifficult position Black goes wrong and ..tf1 lZ'lf7 2S ..tc4, and all the black pieces
is elegantly mated . A subtle queen sacrifice, are tied up) 23 ..te2 .:te8.
which Boguslavsky had prepared , remained
Objectively, Wh ite would have done better
off-screen . After the best move 1 S .. :ii'e 7 1 7
to reject playing for bri l l iancy in favour of 1 8
lZ'lhS+ ..txhS Maxim was intending to play 1 8
xhS! (instead of 1 8 'ii'x d5+ ? ! ) 1 8 . . . ..teS 1 9
'ii'xd 5+ ..teS ( 1 8 . . . 'ii? g 7 1 9 .l:.f7+ 'ili'xf7 20
..tc4 lZ'lc7 20 'ii'f3 (20 ':f8+ 'it'xf8 2 1 ..txeS+
..txhS+ and 1 8 .. :ii'e S 1 9 ..txhS are both bad
lZ'lxeS 22 ..txf8 is also possible) 20 . . . lZ'ld7 2 1
for Black) 1 9 'ii'x eS+ ! ! 'ii'x eS 20 ..txhS.
':ae 1 , a n d Black h a s no defence.
17 'ii'x d4! cxd4
1 8 lZ'lh6+
Black resigned.

Tepl itsky ( 1 4) - Paruti n


Tashkent 1 989
GrOnfeld Defence
1 lZ'lf3 d5
2 c4 c6
3 e3 lZ'lf6
4 lZ'lc3 g6
5 d4 ..tg7
Creative Ach ievements of Pupils from the School 4:J 1 93

6 e2 0-0
7 0-0 li:)bd7
A transposition of moves has led to the
Sch lechter Variation of the Gru nfeld De
fence . Black's last move is considered
inaccurate, since now Wh ite can exchange
on d5, not fea ring the development of the
black knight at c6 - the optimal square in
this variation. I n this way Wh ite gained a
clear advantage in the game Botvinnik-Blau
(Olympiad , Tel Aviv 1 964) after 8 cxd5 cxd5
9 'it'b3 e6 1 0 a4 b6 1 1 .i.d2.
However, White's move in the game is also
not bad .
1 6 eSI
8 b3 e6
An imaginative decision . Such moves are
I n a game with Boris Kantsler, Van Teplitsky very easy to overlook. Now the win of a
found the antidote to another scheme of piece by 1 7 g4 is threatened . The 'auto
development for Black: after 8 . . . b6 there matic' 1 6 bxc4 would have left Black more
followed 9 a4 a5 1 0 cxd5 li:)xd5 1 1 li:)xd5 opportun ities for counterplay after 1 6 . . . e5
cxd5 1 2 a3 l:le8 1 3 l:[c1 a6 1 4 b5! with 1 7 d5 f8 ! ? (but not 1 7 . . . li:)c5? 1 8 xc5
the better game. bxc5 1 9 d6 l:tb8 20 li:)a4 with a decisive
9 'it'e2 l:te8 advantage for White).
1 0 b2 as 16 . . . fS
1 1 l:.ad1 li:)hS If 1 6 . . . cxb3 there is the unpleasant reply 1 7
12 a3! li:)e4 ! ( 1 7 g4 is weaker because of 1 7 . . . c5!
1 8 gxh5 xf3 1 9 xf3 cxd4 with fine
U p to here Wh ite has simply deployed his
cou nterplay, fully compensating for the
pieces sensibly. But now he reacts to the
sacrificed piece). Wh ite responds in the
opponent's plans and takes prophylactic
same way to 1 6 . . . b5!? ( 1 7 li:)e4 ! b4 1 8
measures agai nst . . . f7-f5 , on wh ich there
li:)d6).
follows 1 3 d6, controlling the dark squares.
17 exf6 ! ?
12 . . . b6
I nteresting play, although t h e qu iet 1 7 bxc4
Black changes pla n , but his knight proves
would also have ensured Wh ite the better
badly placed on the edge of the board
chances.
(remember the famous axiom of Dr. Tar
17 . . . li:)hxf6
rasch ! ) . White obtains good play by simple
means: he prepares an offensive in the 1 8 xe4
centre. This move leads to g reat compl ications.
1 3 l:tfe1 .i.b7 Wh ite exchanges two bishops for a rook and
pawn . The consequences of such an ex
1 4 e4 l:te8
change are usually d ifficult to assess cor
1 S 'it'd2 dxe4 rectly. I n many cases, especially in the
middlegame, the two pieces prove stronger,
since they can create more th reats to the
1 94 Creative Achievements of Pupils from the School

opponent. In the g iven position Tepl itsky 'ilVxd6 (but not 26 'ilVf7+? ..ti>h6 27 ttJe6
correctly reckoned that the activity of his because of 27 . . . 'ilVg8) Black proves helpless
heavy pieces, after seizing control of the against the un ited onslaught of the white
only open file, together with Black's weak pieces : there is no satisfactory defence
ened castled position, would prove more against the th reats of 27 ttJe6+ or 27 ':'e7+.
sign ificant factors than the potential power If 24 . . . 'ilVc7 Wh ite has the decisive 25 'iVe6+
of the h itherto dormant black bishops. ..ti>h8 26 ttJce4! .i.xe 1 27 ttJxf6 . Finally, in the
18 . . . b5 variation 24 . . . .i.xc3 25 'ilVe6+ ..ti>g7 26 'iVf7+
1 9 .i.xe6+ lhe6 ..ti>h6 27 ':xd7! 'ilVxd7 (27 . . . ttJxd7 28 ttJe6)
20 l:txe6 b4 28 'ilVxf6 .i.xe 1 29 ttJf7+ 'ilVxf7 30 'ilVxf7 Black
has two bishops and a rook for the q ueen ,
21 'ii'e 2 ! ? bxa3
but one of the bishops is inevitably lost.
If 2 1 . . . bxc3, then 22 l:Ie7 c2 23 l:tc1 with an
The attempt by Black to gain counterplay on
attack (but not 23 'ii'xc2 .i.f8).
the e-file proves unsuccessful .
22 11e1 .i.f8
25 'ilVe6+ ..ti>h8
22 . . . ttJf8 23 lIe7 lIc7 was bad in view of 24
26 ':'xd7 ]::t e 8
'ii'c4+ ..ti>h8 25 'iff7 .
27 ':xh7+
23 ttJg5 .i.b4
This d ispels the opponent's last illusions.
Black overlooks a spectacular stroke by his
Now Wh ite gains a decisive material advan
opponent. However, 23 . . .lIc7 could also
tage, which Teplitsky confidently converts
have been answered by 24 lId6 ! ! .i.xd6
into a win .
(24 . . . 'ifb8 25 'ife6+ ..ti>h8 26 ttJce4 ! ) 25
'ii'e 6+ ..ti>g7 26 'ifxd6 ..ti>g8 27 'ife6+ ..ti>g7 28 [27 'iVxf6+! 'ilVxf6 2 8 l:txe8+ .i. f8 29 ttJxh 7
'iff7+ ..ti>h6 29 ttJe6 'ife8 30 'ilVg7+ with a would have been more quickly decisive -
decisive attack. Dvoretsky.]
27 . . . ttJxh7
28 'ilVxe8 ttJxg5
29 'ilVxg6 ttJf7
30 :e3 'ilVg7
31 'ilVe6 .i.d6
32 :th3+ ..ti>g8
33 ttJe4 'ilVxd4
34 ttJf6+ ..ti>f8
35 'ilVe8+ ..ti>g7
36 l:.h7+ ..ti>xf6
37 'ilVxf7+
Black resigned

24 11d6 ! ! Boguslavsky (1 5) - Morozevich


Vacati ng t h e e6-sq uare for t h e decisive Moscow J u n ior Championship 1 990
i nvasion of the wh ite queen. King 's Indian Defence
24 . . . 'iVf8 1 d4 ttJf6
I n the event of 24 . . . .i.xd6 25 'iVe6+ ..ti>g7 26 2 c4 g6
Creative Achievements of Pupils from the School ltJ 1 95

3 lLlc3 g7 Prematu re activity. It is extremely danger


4 e4 0-0 ous to open up the game with your
development incomplete .
S lLlf3 d6
1 2 exfS gxfS
6 e2 eS
1 3 h6 xh6
7 0-0 'ife8
1 4 'iVxh6 'iVg6
A rare continuation , the main virtue of which
is that the play now departs from familiar 1 4 . . . l:tf6 came into consideration.
paths. 1 S 'ife3 as
8 dxeS dxeS 16 a3 l:te8
9 e3 17 cS 'iVg7?
After 9 lLld5 Boguslavsky g ives the variation And this is already a serious mistake. Black,
9 . . . lLla6 1 0 lLlxf6+ xf6 1 1 e3 'fie7 1 2 who is behind in development, makes
lLld2 lLlc5 1 3 b4 lLle6 1 4 c5 l:td8 1 5 'iVc2 another move with an already developed
g5 with an equal game. The latest piece - of course, this is a impermissible
brai nwave in this open ing line is 9 b4! ? , luxury!
employed b y Ivanchuk against Judit Polar i n I nteresti ng variations would have arisen
Novgorod in 1 996. After 9 . . . c 6 1 0 b5 'iVe7 1 1 after 1 7 .. .f4 . If 1 8 'iVe4 , then 1 8 . . . lLlf6 1 9
a4 :d8 1 2 a3 'iVe8 1 3 'fib3 g4 1 4 a5 a6 'iVxg6+ hxg6 20 c4+ g7 2 1 lLlg5 f5
1 6 bxa6 lLlxa6 1 6 'ifxb7 Wh ite clearly stood (preparing . . . lLla6) 22 lLlf7 d3! 23 xd3
better. xf7 with a tenable position. However, as
9. . . lLlfd7?! Dvoretsky pointed out, Wh ite can play more
energetically: 1 8 'iVd2 !? e4 1 9 c4+ g7
Theory, not without reason , g ives prefer
20 lLlxe4 ! , for example: 20 . . . 'iVxe4 21 l:tae 1
ence to 9 . . . b6 . After 1 0 lLld5 lLla6 Wh ite has
'iVg6 22 'iVxf4 lLlf6 23 ':xe8 'ifxe8 24 l:t e 1 ,
tried various continuations, but nowhere
or 20 . . . l:txe4 2 1 d3 axb4 22 xe4 'ifxe4
has he gained an advantage:
23 l:tae 1 - i n both cases with a powerfu l
1 1 'ifc2 lLlg4 1 2 d2 c6 1 3 lLle3 lLlxe3 1 4 attack.
xe3 lLl b4 with unclear play (Gostisa
Probably Black should fi rst have weakened
Kupreich ik, Belgrade 1 988);
the opponent's onslaught by exchanging
1 1 g5 lLlxe4! ? 12 e7 c6 1 3 xf8 'iVxf8 one pair of rooks: 1 7 . . . axb4!? 1 8 axb4
with good compensation for the sacrificed l:txa 1 .
exchange (Vucicevic-Kupreich ik, Belgrade
1 8 c4+ h8
1 988);
1 9 lLlgS l:te7
1 1 lLld2 !? lLld7 12 'ifa4 b7, and the
position is unclear (Dreev-Gelfand, Arn hem Now 1 9 .. .f4? is too late in view of 20 lLlf7+.
1 989). 20 lLle6 'iVf6
I n stead of 1 0 lLld5 the restrai ned 1 0 h3 is I n the event of 20 . . . 'iVf7 Boguslavsky was
more promising . intending to play 21 lLlc7 'iVxc4 22 lLlxa8
1 0 'ifd2 axb4 23 axb4 'fixb4 24 lLlb6 (24 lLla4 is also
possible) with the possible continuation
The immediate 1 0 b4 !? followed by c4-c5
24 . . . lLlxb6 25 cxb6 f4 26 'iVd2 'iVxb6 27 ':a8
and lLld2-c4 was also not bad .
l:td7 (27 . . . 'iVc7 28 l:txb8) 28 'iVc2 Wc7 29
10 . . . c6 l:tc1 b5 (29 . . . lLla6 30 lLlb5 'iVd8 3 1 lLla7 ':c7
1 1 b4 fS?! 32 l:td 1 'iVe8 33 'iVd2 also fails to save
1 96 <;t> Creative Achievements of Pupils from the School

Black) 30 ltJxb5 'ii' b 7 31 l:txb8 'ii'x b8 32 Black is unable to d isentangle his clump of
'ii'xc6, and Black's position collapses. pieces on the q ueenside. His pawn-grab
21 ltJc7 1:Ia7 bing operation is explained by the well
known argument: ' if I'm going to suffer, then
22 bS!
at least I ' l l have something to show for it' . By
Switching to decisive action. energetic play Boguslavsky qu ickly con
22 . . . a4 cludes the game.
27 . . . 'ifxcs
28 l:1ac1 'ifxa3
If 28 . . . 'ii'f8 , then simply 29 e6 ltJb6 30
xc8 ltJxc8 31 1:Ixe5 and wins.
29 e6 'iff8
30 'ifc3
30 l:.xe5 was also possible.
30 . . . b6
30 . . .'ifc5 would not have saved Black in
view of 3 1 'ifa 1 'iifa 32 1:Ixe5.
3 1 1:IxeS 'iif6
32 ':'gSI
A simple, but attractive stroke . Since the
captu re of the q ueen leads immed iately to
23 ltJcdSI?
mate , Black can only try to postpone this
Of course, Wh ite has an undisputed advan inevitable finish by a couple of moves.
tage. Very often it is not easy to decide
32 . . . h6
whether the moment for concrete measures
has arrived . In the g iven instance Wh ite had 33 :g8+ h7
a good opportun ity to further strengthen his 34 'ii'd 3+
position by 23 1:Iad 1 . However, the ex Black resigned.
changing combination undertaken by Bo
guslavsky, wh ich thematically resembles ZViagintsev (1 3) - Nachev
the previous game, is also good : Wh ite Voskresensk 1 990
sharpens the situation at a moment when Slav Defence
the opponent's pieces are least well pre 1 d4 dS
pared for coord inated action.
2 c4 c6
23 . . . cxdS
3 ltJf3 ltJf6
24 ltJxdS f4 4 ltJc3 dxc4
This intermediate move does noth ing to S a4 g4
change the evaluation of the position : the
A rather risky variation , which has been
black pawns i n the centre remain under fire
upheld in a number of games by Predrag
by the heavy pieces.
N i kolic. True, to cast doubts on it Wh ite has
2S 'ii' d 2 'ii'g S to play very energetically.
26 ltJxe7 'ii'x e7 6 ltJeS hS
27 :fe1 7 h3!
C reative Achievements of Pupils from the School ltJ 1 97

The most u npleasant conti nuation for Black. The game Yusupov-G retarsson , G roningen
At a favourable opportun ity Wh ite wants to 1 997, went 1 3 . . . tDd5 14 :b1 e6 1 5 tDxc4!
occupy the centre with his pawns, and for xc4?! 1 6 'iVxc2 xc5 1 7 dxc5 a6 1 8 b4
this he unpins his e-pawn. with advantage to Wh ite - however, 1 5 . . .
7 . . . tD a6 xc5 1 6 dxc5 0-0 demands further testing.
What can happen if Black does not fight for 14 tDxb7 tDd5
the in itiative is illustrated by a game of Not 1 4 . . . 'iVxd2? because of 1 5 xc6+! 'it'd7
Vadim Zviag intsev against l Iya Frog (Mos 1 6 tDxd7 tDxd7 1 7 :d 1 f5 1 8 :xd7 tDxa 1 1 9
cow 1 989), in wh ich after 7 . . . tDbd7 8 g4 :d8+ f7 20 l::t x a8 (Khenki n-Sapis, Lenin
g6 9 tDxc4 e6 1 0 g2 b4 1 1 0-0 0-0 1 2 g rad 1 989).
a5! tDd5 1 3 'it'b3 b5 1 4 axb6 tD7xb6 1 5 e4
1 5 'iVg5!
tDxc3 1 6 bxc3 e7 1 7 f4 ! Wh ite success
fully carried out his plan of seizing the This home preparation by Zviagi ntsev sets
centre. Black un pleasant problems. 1 5 tDxc6 'iVxd2
1 6 xd2 is less dangerous in view of
8 g4 g6
1 6 . . . tDxa 1 1 7 xd5 e6 (Ftacn ik) or 1 6 . . . e6
9 g2 tD b4
(Gelfand, Kapengut) with roughly equal
1 0 0-0 play. Now both 1 6 tDxc6 and 1 6 'ii'f5 are
If 1 0 e4 there would have followed 1 0 . . :iVxd4. th reatened .
10 . . . c2 ! ? 15 . . . e6! ?
Weaker is 1 0 . . .tD d 7 ? ! 1 1 tDxc4 with advan After t h e critical 1 5 . . .f6!? Zviag i ntsev was
tage to Wh ite (Gelfand-Khuzman , USSR intending to conti nue 1 6 'ii' h 5+ g6 1 7 tDxg6
1 987). hxg6 1 8 'iVxh8, and if 1 8 . . . g5, then White
11 'iVd2 b3 retains the in itiative by playing 1 9 e4! . And in
1 2 tDe4! the event of 1 5 ... 'ii' b 6 , according to his
1 2 'ii'f4 ! ? h6 1 3 e3 leads to unclear analysis, strong is 16 'iVf5 'iVxb7 1 7 'iVxf7+
complications ( Levitt-Flear, British Champi d8 1 8 ':d 1 c3 ( 1 8 . . . tDxa 1 ? 1 9 xd5 xd 1
onship 1 989). If 1 2 a5, with the unequ ivocal 20 xc6 'iVxc6 2 1 tDxc6+ riit c7 22 f4+
idea of advancing the pawn fu rther, modern xc6 23 'iVe6+ with mate in th ree moves;
theory recommends 1 2 . . . e6, not paying any 1 8 . . . tD b4 1 9 f4) 1 9 tDxc6+ 'ii' xc6 20 xd5
attention to the opponent's threat. After 1 3 xd5 21 ':xd5+ c8 22 :d3 ! .
a6 'iVc7 1 4 axb7 'iVxb7 (Campos Moreno
Rogers, Olympiad , Manila 1 992) Wh ite still
has to demonstrate that he has sufficient
compensation for the sacrificed pawn .
12 . . . tDc2
1 2 . . . tDxe4? 1 3 'iVxb4 tDd6 is bad in view of
the spectacu lar rejoinder poi nted out by
Gelfand and Kapengut: 1 4 'iVxb7! ! , and
Wh ite wins.
13 tDc5
If 1 3 tDxc6 , then 1 3 . . . 'iVb6! (Gelfa n d ,
Kapengut).
13 . . . 'iVxd4
1 98 Creative Achievements of Pupils from the School

1 6 .lte3 ! ! 2 0 lLlxe6 ! 'iVa3! 2 1 .ltxd5 cxd5 2 2 lLlc7+


A very unpleasant move for Black. I t 'iii> e 7 (or 22 . . . 'iii> d 8 23 'iVf6+ 'iii> c8 24 'iVc6) 23
transpires that i t is unfavourable to captu re 'iVg5+ f6 24 'iVg7+ Black would have lost
the bishop. 1 6 . . . lLldxe3?? loses immedi i mmed iately.
ately to 1 7 .ltxc6+, wh ile after 1 6 . . . lLlcxe3 20 l:tab1 ! 'iVd4
the f-file is opened and Wh ite obtains a very 20 . . . 'iVa3 was bad in view of 2 1 lLlxb3 cxb3
dangerous attack: 1 7 fxe3 f6 1 8 .l:txf6 gxf6 22 lLlxc6 , but 20 . . . 'iVc3!? came into consid
1 9 'iiV h 5+ (C. Horvath ), or 1 7 . . . 'iVxb2 1 8 eration. Zviag i ntsev was planning to reply
l:tab 1 'ilfc3 1 9 l:[xf7 , and if 1 9 . . . .ltc2 there 21 lLle4 , but then Black is able to simpl ify
follows 20 'iVxg7! .ltxg7 2 1 lLld6+ 'iii> d 8 22 the game by 2 1 . . . .l:tg8! 22 'iVxg8 (forced)
lLlxc6 mate. 22 . . . :xg8 23 lLlxc3 lLlxc3 with an unclear
16 . . . 'iiVx b2 endgame. The prophylactic 21 'iii> h 1 ! ? is
1 7 .ltc5! stronger, or the more direct 21 e4 !? lLlc7
Wh ite intensifies the pressure. Now 1 7 . . . (after 21 . . . lLlf4 the reply 22 'iVg5! is ex
lLlxa 1 ? loses to 1 8 .ltxd5 cxd5 1 9 lLld6+ . If tremely u npleasant) 22 lLlcd7! l:[fe8 23
1 7 . . . h6!? Wh ite had prepared 1 8 'ilff4! ! f6 1 9 'iVxf7 (poi nted out by Dvoretsky).
'iiVe4 'ii'x e5 2 0 'ii'g 6+ 'iii> d 7 2 1 'ilff7+ 'iii> c8 22 2 1 lLlcd7?!
lLla5 l:[b8 23 .i.xf8 with a very strong attack. When the opponent is defending tena
However, as Dvoretsky poi nted out, instead ciously, it ca n be very d ifficult to conduct an
of 21 . . . 'iii> c8?! Black can play 21 . . . .lte7 ! , not attack faultlessly. I n stead of the move in the
fearing 22 .ltxe7 lLlxe7 23 l:tad 1 + lLld4 . game, 21 lLlxb3! cxb3 22 l:txb3 was simpler,
Possibly then Wh ite would have had to force when the difference in the placing of the
a draw by 22 xd5 exd5 23 !iL.xe7 'iVxe7 24 kings should be bound to tell . The operation
lLlc5+ 'iii> d 6 25 lLlb7+. plan ned by Wh ite allows the opponent
17 . . . .ltxc5 u nexpected saving chances .
1 8 'iVxg7!
This i ntermediate move, threatening mate ,
is t h e point o f Wh ite's idea .
18 . . . lIf8
Comparatively best. The queen sacrifice
1 8 . . . 'ilfxe5 1 9 'ilfxe5 .ltd4 would not have
brought any relief, since the th reats created
by the wh ite queen together with the knight
would have been too dangerous. For exam
ple, after 20 'iVd6 !iL.xa 1 21 'iVxc6+ 'iii> f8 22
lLld6 'iii> g 7 23 .ltxd5 exd5 24 lLlf5+ th ings
end i n mate.
I nterposing 1 8 . . . .ltxf2+? would merely open
a n additional line for the attack: 1 9 l:[xf2
'iVxa 1 + 20 .ltf1 ! (weaker is 20 'iii> h 2 .l:tf8 2 1 21 . . . lLlc3?
.l:txf7 'iVxe5+) 2 0 . . . .l:tf8 2 1 .l:txf7. Black th i n ks that he is forced to go in for an
1 9 lLlxc5 0-0-0 1 exchange of blows. With his king exposed ,
I n such a position it is easy to overlook this leads to a rapid finish. Of course,
someth ing. Thus i n the event of 1 9 . . . lLlxa 1 ? 21 .. J:bd7? was bad : 22 'iVxf8+ lId8 23
Creative Achievements of Pupils from the School lD 1 99

'ii'x d8+! xd8 24 ttJxc6+ d7 25 ttJxd4 5 f3 0-0


ttJxd4 26 e3. The only possibil ity of a 6 .lte3 ttJc6
defence was 2 1 . . J:tfe8! ! 22 e3!? (22 .ltxd5 7 ttJge2 a6
exd 5 ! ) 22 . . . 'ii' c3 (22 . . . ttJcxe3!? 23 fxe3
8 a3
ttJxe3 24 'ii'xf7 l:1e7 ! is possibly stronger) 23
ttJc5 ! ? l:te7 ! , and if 24 ttJe4, then 24 .. .f5 1 . The main continuation is the natu ral 8 'iif d 2,
True, even in this case Wh ite h a s good but the move i n the game, preparing play on
chances of success. He continues 25 'iifx e7 the queenside, is also not without venom.
'ii'x e5 26 'iifc 5!? (26 'iifg 5 also comes into However, l Iya Makariev is well prepared for
consideration) 26 . . . fxe4 27 'iifx c6+ ttJc7 28 such a tu rn of events and he acts in
.ltxe4 with the better game. For example, accordance with the recommendations of
28 .. J::t d 5 29 .ltxd5 'iifx d5 30 'ii'x d5 ttJxd5 31 theory.
l:txb3 cxb3 32 lIb1 ttJc3 33 l:txb3 ttJxa4 34 8. . . .ltd7
l::t b 5. 9 b4 'iWb8 ! ?
22 ttJxf8 1 ttJxb1 Black u n h u rriedly prepares . . . b7-b5, since
23 .ltxc6 'ii' b 6 in the given variation it is simpler for Wh ite
23 . . . 'iiff4? is bad in view of 24 ttJxe6. to respond to the opponent's actions than to
carry out his own pla n . Thus after the
24 'iifxfl c3
immed iate 9 . . . b5 there would have followed
After the comparatively best 24 . . . ttJd4 there 1 0 cxb5 axb5 1 1 d5 ttJe5 1 2 ttJd4, wh ile if
is the adequate reply 25 .ltd7+ b8 26 9 . . . e5, then 1 0 d5 ttJe7 1 1 g4 ttJe8 1 2 ttJc1 f5
l:txb 1 . The move in the game loses even 1 3 ttJb3 (Korch noi-Stein, 3 1 st USSR Cham
more material. pionship, Leningrad 1 963).
25 .ltd7+ 10 'iif d 2
Black resig ned . If now 1 0 ttJc1 , then 1 0 . . . b5 would be very
Of cou rse, when we talk about an attack, we timely, since in the event of 1 1 cxb5 axb5 the
have a mental pictu re of brill iant combina b5-pawn is indirectly defended ( 1 2 i.xb5
tions and subtle sacrifices. However, in ttJxb4). If 1 0 d5 ttJe5 1 1 ttJd4 , then 1 1 . . . c6
many cases one can manage without 1 2 dxc6 ( 1 2 f4 ttJg4 ; 1 2 i.e2 cxd5 1 3 cxd5
brill iance. Thus in the following game, nc8) 1 2 . . . bxc6 1 3 i.e2 a5 with cou nterplay.
victory was gai ned without any outward 10 . . . b5
effects, but by apparently very simple Now in the event of the pawn exchange on
means. But such simpl icity is deceptive : b5 Black will be th reatening to captu re on
very often it can be more d ifficult to fi nd a b4 . Also possible was Boleslavsky's recom
q u iet move , strengthening the position , than mendation 1 0 . . . :e8!? (this move is usefu l if
to land a tactical blow. Wh ite should move his knight from e2, since
then . . . e7-e5 gains i n strength ) 1 1 g3 b5 1 2
N i kiti n - Makariev (1 5) c5 a5 1 3 l:[ b 1 e6 1 4 f2 'ii' b 7 with equal
chances.
CIS J u n ior Championsh ip, J u rmala 1 992
King 's Indian Defence 11 g4
A bold decision - Wh ite mounts an offensive
1 d4 ttJf6
over an excessively wide front. On the other
2 c4 g6
hand , contin uations such as 1 1 d5? ttJe5 1 2
3 ttJc3 .ltg7 cxb5 ttJc4 , 1 1 cxb5 axb5 1 2 d 5 ttJxb4 o r 1 1
4 e4 d6 ttJc1 e5 are not too impressive. A possible
200 C reative Achievements of Pupils from the School

alternative is 1 1 c5!? a5 1 2 b 1 axb4 1 3


axb4, maintaining his position i n the centre.
It would have been much harder for Black to
create counterplay ( 1 3 . . . dxc5 1 4 bxc5), and
for the moment the a-file, which has been
conceded to h i m , does not play any role.
11 . . . bxc4
1 2 h4
If 1 2 1:[ b 1 Black was planning 1 2 . . . e5 1 3 d5
ltJa7 ( 1 3 ... ltJe7 with the idea of . . . ltJe8 and
. . . f7-f5 is also possible) 14 a4 c6 with
counterplay.
1 2 g5!? deserved serious consideration. If
12 . . . ltJe8 there would have followed 1 3 h4! 14 . . . 'ili'd8 !
e5 14 d5 ( 1 4 h5!?) 14 . . . ltJe7 1 5 h5 with
This modest retu rn of the queen prepares
u npleasant threats . Black would probably
counterplay in the centre.
have had to decide on 1 2 . . . ltJh5! 1 3 ltJg3!
1 5 f4
e5 14 ltJxh5 gxh5 ( 1 4 . . . exd4?! 1 5 ltJxg7
dxe3 1 6 'ili'b2 ) 1 5 ltJd5 ( 1 5 dxe5!?) 1 5 . . . exd4 If 1 5 d5 Black would have conti nued
( 1 5 . . . ltJxd4? 1 6 .ltxd4 exd4 1 7 ltJf6+), and if 1 5 . . . ltJe5 1 6 ltJd4 e6! 1 7 f4 ( 1 7 dxe6 c5 ! )
1 6 ltJf6+, then either 1 6 . . . .ltxf6 1 7 gxf6 dxe3 1 7 . . . exd5 1 8 ltJxd5 ( 1 8 fxe5 dxe5) 1 8 . . . ltJg4
1 8 'ili'g2+ .ltg4, or 1 6 . . . 'it h8 1 7 .ltxd4 ( 1 7 with nu merous threats.
ltJxd7? dxe3) 1 7 . . . ltJxd4 1 8 'ii'x d4 .lte6 with 15 . . . e5
possibil ities of counterplay on the queenside Beginning a counterattack in the centre.
by . . . c7-c5 or . . . a6-a5. 16 fxe5?
12 . . . h5
1 6 d5? was also incorrect in view of
Wh ite's offensive on the kingside must be 1 6 . . . exf4 . But now Black's position i n the
halted . The unexpected 1 2 . . . a5 1 3 b5 centre is reinforced even more. It was time
ltJb4?! is too pretty to be true. Wh ite can to think about development and to prepare
choose between the restrained 14 axb4 castl ing by 1 6 .i.g2.
axb4 1 5 1:[xa8 bxc3 ( 1 5 . . . 'ili'xa8?! 1 6 ltJa2)
16 . . . dxe5
16 ltJxc3 'ili'xa8 1 7 .ltxc4 with the better
1 7 d5 ltJa7
game, and the more refined 1 4 ltJc1 ! ? with
the threat of ': b 1 . 18 a4
1 3 g5 ltJh7 In parrying the obvious threat of . . . ltJa7-b5,
1 4 1:[b1 ?! Wh ite allows a more unpleasant plan for the
opponent. However, it is not easy to suggest
This move would appear to hand the
anything sensible. Thus 1 8 .ltxa7 1:[xa7
in itiative to the opponent. 1 4 f4 ! with the
(with the threat of . . . f7-f6) is hopeless, as is
threat of 1 5 d5 was more consistent. If
1 8 .ltc5 1:[e8, or 1 8 d6 cxd6 1 9 'ili'xd6 ltJb5.
14 . . . a5, then according to Makariev's analy
Meanwhile, the pawn captu red on c4 is
sis Wh ite should continue 1 5 b5 ltJa7 1 6 1:[b 1
increasingly beg i n n ing to resemble a sound
( 1 6 a4 c6! ) 1 6 . . . 'ili'e8 1 7 a4 ltJc8 1 8 ltJg3 ltJb6
extra pawn . . .
1 9 f5. It probably makes sense for Black to
sacrifice a pawn by 1 4 . . . e5!? 1 5 dxe5 .ltg4 18 . . . ltJc8!
1 6 exd6 .ltf3 with double-edged play. The knight goes to d6, where it not only
Creative Achievements of P u p i l s from the School ctJ 201

defends the c4-pawn , but also presses on I should like to conclude this account of
the white centre. Black's advantage in examples of the pupils' play with one more,
creases. later game by Vadim Zviagi ntsev, which was
1 9 ltJg3 ltJd6 judged to be the best game in Informator
20 i.e2 No.62 and was included in a collection,
published in England , of the 1 00 best
games ever played .

Cifuentes - Zviagi ntsev (1 8)


Wijk aan Zee 1 995
Slav Defence
1 d4 d5
2 c4 e6
3 ltJf3 ltJf6
4 ltJc3 c6
5 e3 ltJbd7
6 "c2 b6! 1
7 i.e2
Black would like to play . . . f7-f6 , after which 7 i.d3! i.b7 8 0-0 is more energetic, and if
Wh ite's position on the kingside must 8 . . . dxc4?!, then 9 i.xc4 c5 1 0 'iVe2 .
crumble. However, the immed iate 20 . . .f6? 7... i. b7
allows 21 i.xh5! or 21 ltJxh5! with com 8 0-0 i.e7
pletely u n necessary complications. A subtle 9 ':d1
prophylactic move prepares a decisive
I n Zviagi ntsev's opin ion , 9 b3 was more
offensive with gain of tempo.
accu rate.
20 . . . "e8 1
9. . . 0-0
21 'iVa2
1 0 e4 dxe4
Parrying the opponent's most obvious
1 1 ltJxe4 'iVc7 1
(2 1 . . . i.xa4) and least dangerous threat.
It is useful to prevent i.f4 .
21 . . . f6
1 2 ltJc31 1
22 :g 1 11 fxg5
23 hxg5 'iVe7 I n this way Wh ite can not hope for an
advantage. If he wanted to obtain a comfort
24 'iVd2 ':f41
able game, he should not have avoided
The decisive stroke! After Black's 'qu iet' exchanges. Both 1 2 ltJxf6+ ltJxf6 and 1 2
20th move his attack has ru n like clockwork. i.g5 c5 wou ld have led to approximate
25 i.xh5 gxh5 equal ity.
26 ltJxh5 :g4 12 . . . c5
27 ':xg4 i.xg4 1 3 d511
28 ltJg3 ]::t f8 A risky conception. 1 3 ltJb5 'ifb8 14 g3 cxd4
29 g6 ltJf6 1 5 ltJbxd4 :e8 would also not have promised
30 i.c5 ltJh5 White anything. Possibly he should have
Wh ite resigned . completed his development with 1 3 i.g5.
202 W C reative Achievements of Pupils from the School

13 . . . exd5 In the event of 23 .i.f4 Black was intending


1 4 cxd5 a6 to avoid the exchange of bishops by
1 5 lLlh4 23 . . . .i.f8 ! , with somewhat the better chances.
All the same White does not achieve his aim 23 . . . J::[ a d8
- the exchange of the blockading piece . The Zviagi ntsev includes his last reserves in the
immediate 1 5 a4 was better. The simple 1 5 battle. 23 . . . lLlxf2? 24 'it'xf2 'iii' h 3 25 .i.f4
.i.g5 or even the ultra-active 1 5 'iii'f5 !? also .i.xf4 was prematu re in view of 26 lLlxf4 (but
came into consideration . not 26 gxf4 because of 26 .. .l:he3! 27 'it'xe3
15 . . . g6 lLlg4+). Also noth ing was given by 23 . . . iVh3
The immed iate 1 5 . . . .i.d6 would have led to 24 .i.f4 l:tad8 25 lLlc4 .
roughly the same situation , but Black did not 24 .i.g2?
wa nt to allow his opponent any active Wh ite takes control of the h3-sq uare and
possibil ities after 1 6 lLlf5 .i.xh2+ 1 7 'it' h 1 creates the possible th reat of f2-f3 . This
.i. d 6 1 8 lLle4, although the pawn sacrifice very natu ral operation nevertheless has a
looks questionable. tactical flaw. Little was also promised by 24
1 6 .i.h6 J::[ fe8 .i.f4 .i.f8 , since if 25 lLlc4 there is the simple
1 7 d2?! reply 25 . . . .i.xd5. Fol lowi ng the example of
his opponent, Wh ite should have brought
Now Wh ite should defin itely have restricted
his rook into play, si nce for the moment the
Black's play on the queenside by 1 7 a4! .
sacrifice on f2 does not work: 24 l:tac1
17 . . . .i.d6
lLlxf2? 25 'it>xf2 h3 26 .i.f4 'iVxh2+ 27 lLlg2
With this thematic move Black parries the .i.xf4 28 lLlxf4.
opponent's simple th reats of 1 8 d6 and 1 8
lLlf5 .
1 8 g3 b5
1 9 .i.f3
If 1 9 lLlf5, then 1 9 . . . b4 is un pleasant.
19 . . . b4
1 9 . . . lLle5 also came into consideration .
2 0 lLle2
Hoping to create counterplay after a2-a3 or
lLlf5 .
20 . . . lLle4
Here also it was not too late for 20 . . . lLle5.
The move in the game allows Wh ite to bring
his king's knight into play and consolidate
his position. 24 . . . lLlxf2!
2 1 'ifc2 lLldf6 This sacrifice is merely the prelude to a
22 lLlg2! genuine combinative firework display. The
The knight heads for e3 and, given the wh ite king is l u red into the centre , under the
opportun ity, to c4 . fi re of the opponent's pieces.
22 . . . d7 25 'it'xf2 l:txe3 !
23 lLle3 26 .i.xe3?!
C reative Achievements of Pupils from the School ctJ 203

If Cifuentes had anticipated the fate await


ing h i m , he would undoubtedly have pre
ferred 26 c.ti>xe3 liJg4+ 27 c.ti>d2 liJxh6 28
c.ti>c1 'iVe7, although in this case too the
chances are with Black, who has a pawn for
the exchange with good attacking chances.
26 . . . liJg4+
27 c.ti>f3 liJxh2+
28 c.ti>f2 liJg4+
29 c.ti>f3
Wh ite was probably hoping that his oppo
nent would repeat moves: after a l l , Black is
a rook down ! But Zviagi ntsev continues to
find new attacking resources. 31 . . . 'iVe3+ 1 1
29 . . . 'iVe6! The most elegant solution , lead ing by force
30 .i.f4 to mate . However, the prosaic 3 1 . . . Sl.xf4
would also have won :
Although this move loses, it should not
really be criticised . Other continuations 3 2 gxf4 'iVe3+ 3 3 c.ti>xg4 Sl.c8+ 34 f5 (34 c.ti>h4
would most probably have led to the same 'iVf2+ 35 liJg3 :e3) 34 . . . Sl.xf5+ 35 c.ti>h4
result: .f2+ 36 liJg3 :e3 ;

30 'iVe4 'iVxe4+ 3 1 c.ti>xe4 :e8+; 32 liJxf4 liJh2+ 33 c.ti>f2 'iVe3 mate.


30 Sl.c1 c4 ! 3 1 'iVe4 (or 3 1 liJf4 liJh2+ 32 I n my view, this dual solution does l ittle to
c.ti>f2 Sl.c5+ with mate) 3 1 . . . 'iVxe4+ 32 c.ti>xe4 spoil the aesthetic impression made by
liJf2+ 33 c.ti>d4 liJxd 1 with a material and Zviagintsev's wonderful attack.
positional advantage; 32 Sl.xe3
30 .i.g5 .i.e7 31 liJf4 'iVe3+ 32 c.ti>xg4 .i.c8+ Black gives mate even more quickly after 32
33 liJe6 'iVxg5+ 34 c.ti>f3 fxe6 with a decisive c.ti>xg4 Sl.c8+ 33 c.ti>g5 (33 c.ti>h4 Sl.e7+)
attack. 33 . . . :e5+ .
30 . . . e8 ! 32 . . . :xe3+
Evidently the most energetic continuation of 33 c.ti>xg4 .i.c8+
the attack. Clearly weaker was 30 . . . .i.xd5+ 34 c.ti>g5
3 1 l:txd5 'iVxd5+ 32 'ii'e4 , but both 30 . . . .i.xf4
Or 34 c.ti>h4 Sl.e7 mate .
3 1 'iVe4 ! ? 'iVxe4+ 32 c.ti>xe4 .i.g5 and
30 .. .f5!? were perfectly possible. 34 . . . h6+!

31 'iVc4 The final touch . 34 . . . c.ti>g7 would have


prolonged the game in view of 35 lt h 1 (or 35
There is no other defence against 3 1 . . .
'iVxc5) .
.i.xd5+.
35 c.ti>xh6 .:te5
Against mate on the next move (36 . . . .i.f8 or
36 . . . l:th5) there is no defence .
White resigned .
204

I ndex of P layers and Analysts

Adams 1 29 Fori ntos 1 49


Ahues 80 Franzoni 6 1
Alexander 9, 31
Anand 1 23, 1 78 , 1 79, 1 80 Georgadze 1 38
Grigoriev 8 1
Bagirov 42, 44 Gulko 1 26
Bareev 1 77 Gurgen idze 53
Belavenets 53
Beliavsky 53, 1 04, 1 49 Hort 27
Belov 1 53, 1 55, 1 56, 1 58 , 1 60 H O bner 1 2 1
Berg 27
Blumenfeld 35, 36, 83 loh lesen 53
Bobrov 28 Ivanchuk 1 1 8
Boguslavsky 1 9 1 , 1 94 Ivkov 9 1
Bondarenko 1 5 , 25
Janowski 82
Bronstein 1 4 1
Jansa 28
Butnoryus 1 2

Kamshonkov 1 55
Capablanca 48
Karpov 46, 55
Chekhov 20
Kasparian 23
Chern i n 53
Kasparov 46, 55, 1 77, 1 78 , 1 79, 1 80
Chernosvitov 1 82
Kholmov 1 49
Cifuentes 201
Kh ramtsov 1 45
Ciocaltea 1 6
Kmoch 78, 79
Denisov 1 82 Kotkov 1 7
Dolmatov 59, 6 1 , 63, 66, 68, 73 Kotov 1 0
Dvoretsky 1 2 , 1 7 , 20, 44 , 47, 99, 1 38 , Krasen kow 33
1 4 1 , 1 43, 1 45 Kuznetsov 1 5

Engel 1 00 , 1 1 2 Lag unov 33


Euwe 9, 82 Larsen 53, 66
Evans 1 34 Lasker, Em. 82
Lempert 1 56
Fahrni 84 Lepin 1 9 1
Fischer 1 34 Lerner 1 6 , 59, 73
Flesch 63 Liberzon 1 6
I ndex of Players and Analysts ttJ 205

Liburkin 25 Sanakoev 1 00 , 1 0 1 , 1 05, 1 07 , 1 1 2 , 1 65


Lukin 1 6 Sax 97
Lungdal 1 07 Sergeev 36, 8 1
Lyublinsky 1 49 Shamkovich 1 33
Shevechek 1 05
Maeder 1 65 Simagin 1 33, 1 49, 1 5 1
Makariev 1 99 Smyslov 53
Makarychev 53 Sokolov, A. 28
Marshall 3 1 Suetin 42
Maryasin 1 43
Mestel 1 75 Tal 47, 85, 89
Miles 53 Tepl itsky 1 92
M ityaev 1 60 Timoshchenko 29
Morozevich 1 94 Tolonen 1 50
Tsariov 94
Nachev 1 96
Najdorf 1 0
Vaganian 29
Naumkin 1 49
Van der Sterren 68
N i kitin 1 99
Vasyu kov 85
N imzowitsch 48, 78, 82
Vaulin 1 53
Ochoa 1 50 Vera 1 50
Orlov 83 Vulfson 94

Paruti n 1 92 Wotawa 1 4
Pch iolkin 1 50
Peev 99 Xie J u n 1 1 6
Petrosia n , A. 1 49
Petrosian, T. 9 1 , 97, 1 5 1 Yach mennik 1 58
Pinter 53 Yates 79, 80
Platonov 28 Yusupov 1 1 5, 1 1 6 , 1 1 8 , 1 2 1 , 1 23, 1 26 ,
1 29
Razuvaev 1 04
Rebel 8 1 1 5 Zaitsev, A. 1 0 1
Ribli 89 Zedek 1 49
Rivas Pastor 1 75 Zhivtsov 35
Rotlewi 84 Zviagintsev 1 96 , 201
206

I ndex of Ope n i ngs

Alekh ine Defence 44

Caro-Kan n Defence 63, 66

French Defence 1 38 , 1 4 1

Grunfeld Defence 1 33, 1 92

King's I ndian Attack 85

King's I ndian Defence 1 94 , 1 99

Modern Benoni 1 9 1

N imzo-Indian Defence 9 1 , 1 26

P h ilidor Defence 59

Pi rc-Ufimtsev Defence 97, 1 29, 1 43

Queen's Gambit Accepted 1 1 8 , 1 82

Queen's I ndian Defence 55

Queen's Pawn Open ing 1 1 5, 1 23

Reti Opening 20, 68, 89

Ruy Lopez 1 1 6

Sicilian Defence 6 1 , 73, 94 , 1 65

S i magi n-Larsen Open ing 1 45

Slav Defence 1 2 1 , 1 96, 201

Potrebbero piacerti anche