Sei sulla pagina 1di 17

Candidate name: Sara Risti

TABLE OF CONTENTS
1. Introduction

..2
1.1.
Probability ....
....2
1.1.1.
Experimental ..........................................................
................................................2
1.1.2.
Theoretical ..............................................................
................................................3
1.2.
Gambler's
fallacy ....................................................................................
.........................5
1.3.
Personal
engagement ...........................................................................
.........................5
2. Rationale .....................................................................................
................................................6
3. Modeling .....................................................................................
................................................7
3.1.
Game .............................................................................
.......................................................7
3.1.1.
Rules .......................................................................
...................................................7
3.1.2.
Objective ................................................................
..................................................7
3.2.
Outcome ........................................................................
.....................................................7
3.2.1.
Raw
data ...................................................................................
................................7
3.2.2.
Proccessed
data ...................................................................................
...............10
4. Conclusion ..................................................................................
.............................................14
1 | Page

Candidate name: Sara Risti

5. References ..................................................................................
.............................................15

1.

INTRODUCTION

1.1 Probability
Probability is the extent to which something is probable; the likelihood of
something happening or being the case. 1 By using mathematics, one can
describe the chance of an event happening. 2 Probability in math is a number
between 0 and 1 which describes the odds of a certain event occurring. An
impossible event has 0% probability of happening and a certain event has 100%
probability of happening3.

1Probability - Definition and More from the Free Merriam-Webster Dictionary, Merriam-Webster, 2014,
<http://www.merriam-webster.com/dictionary/probability> (28.9.2014.)

2 Haese, Robert, Sandra Haese, Michael Haese, Marjut Maenpaa, and Mark Humphries. Mathematics for the
International Student: Mathematics SL. Adelaide: haese Mathematics, 2012. Print.

3 Ibid.
2 | Page

Candidate name: Sara Risti


One can calculate probability either by observing results of an experiment
(experimental probability) or by using arguments of symmetry (theoretical
probability)4.

1.1.1 Experimental probability


In experiments, there are 4 key terms that are used to calculate probability:

Number of trials: number of times the test has been conducted


Outcomes: the different results for each trial of the test
Frequency: of a specific outcome is the number of times that outcome
has been observed
Relative frequency: of an outcome is the frequency expressed as a
fraction of percentage for total number of trials

To make the terms clearer, an example experiment has ben conducted; a coin
has been thrown 200 times. The outcome can be either heads or tails. In the table
below is the recorded data.
Table 1.

OUTCOM

The relative frequency or probability is calculated by

Probability ( outcome A )=

frequency of outcome A
number of trials

P ( heads )=

109
=0.545 100 =54.5
200

P ( heads )=

91
=0.455 100 =45.5
200

4 Ibid.
3 | Page

Candidate name: Sara Risti


This ensues that the probability for flipping a coin and it landing on heads is
54.5% and for it landing on tails 45.5%. This is the relative frequency. From no
further testing, the only conclusion one can pull is that these are the odds of a coin
flip. But, if one were to have a bigger number of trials, they would observe chances
closer to 50% for both heads and tails5.

1.1.2 Theoretical probability


This probability is based on what we theoretically expect to occur. 6 The chance for
any outcome to happen is the equal. As an example, a die. A die has 6 sides ergo 6
outcomes and if we assume that the die is not loaded, every side (number) has a

1
6

chance. And so the same formula can be used:

Probability ( outcome A )=

frequency of outcome A
number of trials

In theoretical probability there are complementary and compound events.


Complementary events are those were one of the events must occur;

P ( A )+ P ( A ) =1
Considering if A is an event, A is the respective complementary event.
In the example of a coin;

P ( heads ) + P ( heads )=1


1 1
+ =1
2 2

Compound events are the probability of two or more things happening at once 7.
These kinds of experiments a conducted with two or more object for example a coin
and a die, or two coins or two dies. There are independent compound events and
dependent compound events.
5 Lawoflargenumbersanimation2.gif (100169), Wikimedia, unknown,
<http://upload.wikimedia.org/wikipedia/commons/4/49/Lawoflargenumbersanimation2.gif> (28.9.2014.)

6 Haese, Robert, Sandra Haese, Michael Haese, Marjut Maenpaa, and Mark Humphries. Mathematics for the
International Student: Mathematics SL. Adelaide: haese Mathematics, 2012. Print.

7 www.shmoop.com/basic-statistics-probability/compound-events.html, Basic Statistics and Probability, 2014


<http://www.shmoop.com/basic-statistics-probability/compound-events.html> (28.9.2014)

4 | Page

Candidate name: Sara Risti


Independent compound events are those where one event does not affect the
probability of the second, third, n th event.8 As an example, tossing a die does not
effect a coin in any way so the two events are independent of each other and each
have their own probability. To calculate the independent events, saying that A is one
event and B is another, one uses this formula;

P ( AB )=P( A) P( B)
For example, to calculate what is the probability of getting heads on a coin and
rolling an even number on a die;

1 3 3 1
P ( headsan even number )=P ( heads ) P ( even number )= = = =25
2 6 12 4
Dependent compound events are those were the outcome of one event affect the
second, third, nth event9. An example for this are playing cards. There are 52 cards,
minus jokers, in one deck of cards. If you pull out one card, and then another,
without replacing or putting the first card back, what are the chances to pull out an
ace and a 5? The general formula is;

P ( AB )=P( A) P( BA)

P ( an acea 5 )=P ( ace ) P ( a5|ace )=

4
4
4
=
=0,60
52 51 663

After pulling out one card, that reduces the number or cards in the deck so
the second card has a slightly bigger chance to be any other card.

1.2 Gamblers/Monte Carlo fallacy


The Gamblers fallacy is A fair gambling device has produced a "run"that is, a
series of similar results, such as a series of heads produced by flipping a coin.
Therefore, on the next trial of the device, it is less likely than chance to continue
the run.10 It is a human way of misinterpreting statistically independence and
understanding that each throw is independent of the last one. We believe that there
must be balance in the universe so for example, a 6 on a die will only come up after
8 Haese, Robert, Sandra Haese, Michael Haese, Marjut Maenpaa, and Mark Humphries. Mathematics for the
International Student: Mathematics SL. Adelaide: haese Mathematics, 2012. Print.

9 Dependent Events, Math Goodies, 2014, <http://www.mathgoodies.com/lessons/vol6/dependent_events.html>


(28.9.2014)

10 Logical Fallacy: The Gambler's Fallacy, Fallacy Files, 2014, <http://www.fallacyfiles.org/gamblers.html>


(28.9.2014.)

5 | Page

Candidate name: Sara Risti


several throws, as a way to balance out the smaller number. This is completely false

because a 6 has the same chance, 6 , to be thrown as 1, 2, 3, 4, and 5.


The most famous example of this fallacy was on August 13 th 1913, in the
Monte Carlo Casino at a roulette table. The ball landed on black 26 times in a row.
People lost a lot of money betting against black, thinking that the next spin will
even out and be red11. The statistical chance of having a roulette ball landing on
black 26 times if there are 18 red number, 18 black numbers and a zero is 12;

P ()=
26

[ ]
18
37

18
100 =48.64
37
=0.0000000073=0.0000073

The chances for a ball to fall on black is one time are 48.64%, roughly half of
the time. But for it to land on black 26 times in a row is

0.0000073 . It was a nearly

impossible event, but it happened.


Of course this is an extreme example of the fallacy, simple ones being in the
realm of after having 5 heads in a row, a tails is due, certain numbers (bigger
ones) have a high chance to be picked out in a lottery, or waiting for a 6 13.

1.3 Personal engagement


Growing up, I played the board game Ludo which consisted of racing four
tokens from start to finish and the amount of spaces you could move per a round
was determined by the number you rolled on the die. I was always cheering for six
and thought that a certain number of throws had to pass in order to get a six. The
logic was If I get low numbers for a long time, I will get higher numbers afterwards
which assumed that the Universe keeps score of all die rolls and wanted to equally
distribute them. This is exactly the Gamblers fallacy in action. The reason why I am
doing this modeling is to realize that there is no equality but just probability of an
event (getting a six) and that it is easier to calculate it the odds than make intricate
patterns.

11 Gambler's fallacy - Wikipedia, the free encyclopedia, Wikipedia, unknown,


<http://en.wikipedia.org/wiki/Gambler's_fallacy> (28.9.2014.)

12 ROULETTE WHEEL | Colors | Numbers | Layout | Sequence, Roulette Wheel, unknown,


<http://www.fouroulette.com/roulette-wheel.htm> (28.9.2014.)

13 Gamblers Fallacy, Logically Fallacious, 2013, <http://www.logicallyfallacious.com/index.php/logicalfallacies/98-gambler-s-fallacy> (28.9.2014.)


6 | Page

Candidate name: Sara Risti

2.

RATIONALE

The reason why the topic of probability and discussing the Gambler's fallacy
has been taken for this investigation is because we humans, as a species, think we
can control everything and that everything must have some sort of sense,
justification. Most of us, including me prior to this assignment, do not realize that
chance is just that, chance, a percentage of an event happening or not happening
on which we can have no effect on but can merely view its results.
The modeling I chose was to play a simple game wth one die, adding up the
score after 10 throws and in first round the player with the smaller number wins, but
in the second round the player with the higher score wins. The reason why I chose
such a simple game was to highlight the fact that the players have nothing to do
with the outcome and that both players have the same chances to win because
they are playing with the same 2 die. I expect the score to be similar to one other,
differentiating in a maximum of 20 numbers.

7 | Page

Candidate name: Sara Risti

3.

MODELING

3.1 Game
3.1.1. Rules
Two players alternate in throwing a die. They write down the faces, for
example 6. After 10 throws, the both add up their scores. The player with the
smaller score in the first round wins and the player with the bigger score in the
second round wins.

3.1.2. Objective
The point of the this simple game is to couneract the rule in many other
games, where one where it is better to get higher numbers than lower ones and so
the opinion is is that it is easier to get a lower number than a higher one.
The point of this game is for the participants to realize that the dice does not
know which game is played therefore the outcome is pure chance, as it actually
is. By alternating the conditions of the winner, the illusion of a smaller number
getting thrown when a goal of the game is to have a bigger score and vice versa,
will be shattered and so the participants will learn on their skin not to trust the
Gambler's fallacy.

3.2 Game
3.2.1. Raw data
8 | Page

Candidate name: Sara Risti


The data recorded is from me playing three of my friends.
Table 2.1 First round with Lejla; smaller score wins, I win
name/num
ber of
throws
1
2
3
4
5
6
7
8
9
10

Total

SARA

LEJLA

6
1
2
2
1
4
2
4
4
3

1
1
2
4
4
6
3
3
3
5

29

32

Table 2.2 Second round with Lejla; bigger score wins, Lejla wins
name/num
ber of
throws
1
2
3
4
5
6
7
8
9
10

Total

SARA

LEJLA

2
1
4
1
1
1
6
4
6
3

5
4
5
6
6
4
4
3
3
1

29

41

Table 2.3 First round with Bella; smaller score wins, I win
name/num
ber of
throws
1
2
3
9 | Page

SARA

BELLA

4
1
2

3
5
4

Candidate name: Sara Risti


4
5
6
7
8
9
10

5
1
1
4
3
6
4

4
5
4
4
4
5
4

Total

31

42

Table 2.4 Second round with Bella; bigger score wins, I win
name/num
ber of
throws
1
2
3
4
5
6
7
8
9
10

Total

SARA

BELLA

4
6
4
6
6
4
3
1
5
1

2
1
1
5
2
5
4
6
5
5

40

36

Table 2.5 First round with Selma; smaller score wins, I win
name/num
ber of
throws
1
2
10 | P a g e

SARA

SELMA

5
6

3
3

Candidate name: Sara Risti


3
4
5
6
7
8
9
10

4
5
5
2
1
2
2
5

5
4
2
4
6
2
6
5

Total

37

40

Table 2.6 Second round with Selma, bigger score wins, Selma wins
name/numb
er of throws
1
2
3
4
5
6
7
8
9
10

Total

SARA

SELMA

4
2
4
2
1
6
5
1
6
4

6
5
5
6
4
5
6
1
5
4

35

47

3.2.1. Proccessed data


Graph 1.1

11 | P a g e

Candidate name: Sara Risti

Smaller score winner

OTHERS

SARA
6

2
6
6

7
2

Number of trials: 60

P (1 ) =

8
100 =13.33
60

P (2 )=

10
100 =16.67
60

P (3 )=

8
100 =13.33
60

P ( 4 )=

16
100 =26.67
60

P (5 )=

11
100 =18.33
60

P (6 ) =

4
100 =6.67
60

12 | P a g e

2
2

Candidate name: Sara Risti

Bigger score winner

OTHERS
8

2
2
3

Number of trials: 60

P (1 ) =

12
100 =20
60

P (2 )=

5
100 =8.33
60

P (3 )=

4
100 =6.67
60

P ( 4 )=

13
100 =21.67
60

P (5 ) =

12
100 =20
60

P (6 )=

13
100 =21.67
60

13 | P a g e

SARA
10

7
2

Candidate name: Sara Risti

Number of times of faces of a die rolled in total

OTHERS

15

5
14

10

Number of trials: 120

P (1 ) =

20
100 =16.67
120

P (2 ) =

15
100 =12.5
120

P (3 )=

12
100 =10
120

P ( 4 )=

3
100 =25
120

P (5 )=

22
100 =18.33
120

P (6 )=

18
100 =15
120

14 | P a g e

8
4

SARA
15

15

Candidate name: Sara Risti


For the amount of trials, 120, the chances for rolling a die are not exactly

1
6 , 16.67%. The only one with the exact percentage was the face with the
number 1, all other faces have either smaller or larger chances compared to the
theoretical one. This can be explained due to a small number of trials; even though
120 does seem like a lot, much bigger numbers must be used to obtain the chances
closer to

1
6 .

4.

CONSLUSION

From the data of the experiment I made, it is obvious that the Gambler's fallacy is
indeed, a fallacy, that every face of the die has a

1
6

to appear and the order of

appearance is nothing but chance. By researching and proving to myself this fact, I
am sure not to fall for it again.
The significance of this result is that it shows unequal distribution of probability in
smaller scale, meaning that people could fail to see the fallacy and act upon their
misunderstandings. If a person throws a die 50 times, they will observe a preference
to one side while a person that throws a die 300 times will notice that most of the
die sides appear equally.
The limitation to this experiment is that the number of times rolled is far too small
for a proper check of theoretical probability. The results showed that only the die
face 1 had the exact theorethical probability while others had too high or too low of
a value. The actual probabity will never be exactly as much as the theoretical one,
but as he number of trials continue, it will apprach the theorethical one.
An extension of this experiment could be either to scaled it up to size, perhaps to
have 1000+ die rolls or to change the type of dice (10-sided, 20-sided, 12-sided)
and observe the distrubution of the die face. Another extension would be to transfer
the experiment to a roulette wheel and measure the probability of getting red/black,
of choosing a number inside of the first/middle/last 18 numbers, odd/even. More
spins would assure results higher to the theorethical probability of

15 | P a g e

1
=2.7 .
37

Candidate name: Sara Risti

5. BIBLIOGRAPHY
Haese, Robert, Sandra Haese, Michael Haese, Marjut Maenpaa, and Mark
Humphries. Mathematics for the International Student: Mathematics SL. Adelaide:
Haese Mathematics, 2012. Print.

Basic statistics, unknown


<www.shmoop.com/basic-statistics-probability/compound-events.html>
(28.9.2014)

Dependent Events, Math Goodies, 2014,


<http://www.mathgoodies.com/lessons/vol6/dependent_events.html>
(28.9.2014)

Gamblers Fallacy, Logically Fallacious, 2013,


<http://www.logicallyfallacious.com/index.php/logical-fallacies/98-gambler-sfallacy> (28.9.2014.)

Gambler's fallacy - Wikipedia, the free encyclopedia, Wikipedia, unknown,


<http://en.wikipedia.org/wiki/Gambler's_fallacy> (28.9.2014.)

Logical Fallacy: The Gambler's Fallacy, Fallacy Files, 2014,


<http://www.fallacyfiles.org/gamblers.html> (28.9.2014.)

Probability - Definition and More from the Free Merriam-Webster Dictionary,


Merriam-Webster, 2014, <http://www.merriamwebster.com/dictionary/probability> (28.9.2014.)

ROULETTE WHEEL | Colors | Numbers | Layout | Sequence, Roulette Wheel,


unknown, <http://www.fouroulette.com/roulette-wheel.htm> (28.9.2014.)

16 | P a g e

Candidate name: Sara Risti

17 | P a g e

Potrebbero piacerti anche