CONTENT

Certificate ii

Declaration iii

Acknowledgement iv

Chapter 1: Introduction
1.1 introduction to OR 4
1.2 introduction to game theory 5
1.3 Some basic definition 5-8

Chapter 2: solution of game with Saddle Point

2.1 the maximum and minimax principle 9-10
2.2 Some definition 11
2.2 a )saddle point 11
2.2 b) value of game 11
2.2 c) fair game 11
2.2 d) strictly determinable 11
2.3 procedure to find the saddle point 11
2.4 solution of game with Saddle Point 12-16
Chapter 3. Solution of game without saddle point
3.1 mixed strategy 17-19
3.2 solution of 2 X 2 games without saddle point 20
3.3 dominance property and its example 20-26
3.4Graphical method for the solution of 2 X n and m X 2 games 27-32
3.5 Equivalence of the rectangular Matrix Game and Linear Programming 33-35
3.6 Solution of Rectangular Game by Simplex Method
Chapter 4. Application and limitation of game theory
4.1 application 40-43
4.2 limitation of game theory 43
Chapter 5. Conclusion 44
Reference. 45

1
2
CHAPTER 1: INTRODUCTION

1.1 Introduction to OR
Operation research, often shortened to initialism OR, is a discipline that deals with
the development and application of advance analytical methods to improve decision
making.
It is sometimes considered to be a subfield of mathematical sciences or management
sciences.
Operation research arrives at optimal or near-optimal solutions to complex decision-
making problems.
OR is often concerned with determining the extreme values of some real-world
objective: the maximum (profit, performance or yield) or minimum (loss, risk or
cost).
OR also has strong ties to computer science and analytics.
The major sub-disciplines in modern operation research as identified are: -
● Computing and Information technology.
● Financial engineering.
● Manufacturing and supply chain management.
● Policy modeling and public sector world.
● Revenue management.
● Stochastic models.
● Transportation.
In 17th century , mathematicians Blaise Pascal and Christiaan Huygens solved
problems involving complex decisions by using game- theoretic ideas & expected
values.

3
1.2 Introduction to Game theory
Game theory is the study of mathematical models of strategic interactions among
rational agents. Game theory was developed extensively in the 1950s by many
scholars. It has been recognized as an important tool in many fields, of social science,
as well as in logic, system science, and computer science.
The theory is helpful when two or more individuals or organizations with conflicting
objectives try to make decisions. In such situations, a decision made by one decision-
maker affects the decisions made by one or more of the remaining decision-makers
and the final outcome depends upon the decision of all parties. Such situations often
arise in the field of business, industry, economics sociology, and military training.
This theory is applicable to a wide variety of situations such as two players
struggling to win at chess, two enemies planning war tactics, candidates fighting an
election, marketing competing products, firm struggling to maintain their market
shares, negotiations between organizations unions etc.
The theory of game is based on the minimax principle put forwarded by J. Van
Neumann which implies that each competitor will act so as to minimize his
maximum loss or maximize his minimum gain, i.e., achieve the best of the worst.
The theory of game was developed by Von Neumann (called father of game theory)
in 1928. His theory was followed by the 1944 book Theory of Games and Economic
Behavior, co-written with Oskar Morgenstern, which considered cooperative games
of several players.
The primary contribution of game theory has been its concept rather than its formal
application to the solution of real problems.

1.3 Some basic definitions:

1.3.1 Player: The competitors in the game are known as players. A player may be
an individual or group of individuals or an organization.
1.3.2 Strategy: The strategy of a player is the predetermined rule by which a player
decides his course of action.
a) Pure strategy: If the player selects the same strategy each time , then it is
referred to as pure strategy. In this case, each player knows exactly what the
other player is going to do.

4
 b) Mixed strategy: When the players use a combination of strategies and each
player always kept guessing as to which course of action is to be selected by
other player at a particular occasion then it is known as mixed strategy.
1.3.3 Competitive Game: A competitive situation is called a competitive Game if
it has the following properties:
● There are finite number of competitors called players, if the no. of competitor
is 2 then it is called a two-person game & if the no. of competitors is n>2 then
it is called n-persons game.
● Each player has a list of finite number of possible activities (the list may not
be the same for each player).
● A play is said to be played when each of the players (competitors) choose a
single course of action available to him. Here, it is assumed that the choice
are made simultaneously so that no player knows his opponent’s choice until
he has decided his own course of action.
● Every play i.e., combination of courses of action determines an outcome
(which may be many) which determines a set of payments (+ve, -ve or 0) one
to each player.
1.3.4 Finite or Infinite Games: A game said to be a finite game if it has a finite no.
of moves (choice), each of which involves only a finite number of alternatives.
A game which is not a finite game is called an infinite game.
1.3.5 Zero-sum game: Consider a game where there are n competitors, and
competitor i has Ni courses of action available to him. Then the total no. of possible
outcomes to a play of the game will be N1, N2,….,Nn. Let a particular outcome
𝜃 result in a payment p(i,𝜃 ) to competitor i. Then the game is called a zero-sum game
if for every possible outcome 𝜃 , we have,
∑𝑛𝑖=1 𝑝(𝑖, 𝜃 ) = 0
In other words, a game is said to be zero-sum game if the sum of payments to all
competitors after a play of the game is restricted to zero.
1.3.6 Two-person zero-sum ( or Rectangular ) Game: A game with only two
players in which the gains of one player are the losses of another player, is called a
two-person zero-sum game. In other words, the game in which the algebraic sum of
gains and losses of all the players is called zero-sum game.
Two persons zero-sum game is also called as rectangular games as these are usually
represented by a pay-off matrix in rectangular form.

5
1.3.7 Pay-off matrix: In two-person zero-sum game, the resulting gain can easily
be represented in the form of a matrix called the pay-off matrix or gain matrix.
Thus, Pay-off matrix is a table which shows how payments should be made at the
end of a play or game.
Let us consider a game with only two players (competitors) A & B in which player
A has m- courses of action & player B has n courses of action. The game can be
described by means of a pair of matrices such that
1. Row designation for each matrix are the courses of action available to A.
2. Column designations for each matrix are the courses of action available to B.
3. The cell entries are the payments to A for one matrix and to B for other matrix.
The cell entry aij is payment to A, in A’s pay-off matrix when A chooses the course
of action j.
4. In a zero-sum two-person game, the cell entry in B’s pay-off matrix will be
negative of the corresponding cell entry in A’s pay-off matrix.
The two pay-off matrices are as follows:

A’s Pay-off matrix

The pay-off matrix to player B is (-aij).
e.g.:- A & B play a game called “Two finger match”. Both simultaneously show
either 1 or 2 fingers.
A wins and gets ₹1 from B if :- A’s & B’s matches i.e., 1,1 or 2,2.

6
B wins and gets ₹1 from A if:- finger does not match i.e., 1,2 or 2,1.
Thus, the game is two person zero-sum game, since the winning of 1 player as taken
the losses of other.

The pay-off matrix to A is as follows:

A
Fig.1 Fig.2

Fig.1
+1 -1
B
Fig.2 -1 +1

B’s pay- off matrix:

B
Fig.1 Fig.2

Fig.1
-1 +1
A
Fig.2 +1 -1

i.e., B’s pay-off matrix, the cell entries are just the –ve of the corresponding cell
entries of A’s.

7
CHAPTER 2: SOLUTION OF GAME WITH SADDLE
POINT

2.1 The maximin and minimax principle:

Minimax strategy: Minimizing one’s own maximum loss.
Maximin strategy: Maximizing one’s own minimum gain.

The minimax criterion of optimality states that if a player lists the worst possible

outcomes of all his potential strategies, he will choose that strategy to be most

suitable for him which corresponds to the best of these worst outcomes. Such a

strategy is called optimal strategy.

The maximin- minimax is used for the selection of optimal strategies by two players.

This method is known as the calculus method.

To understand the principles of maximin & minimax, consider a game with two

players A and B. Player A has m strategies (moves) and player B has n

strategies(moves). The game can be described in the form of pay-off matrix such

that the cell entry aij is the payment to A in A’s pay-off matrix when A chooses the

ith strategies and B chooses the jth strategies.

8
Pay-off matrix for the player A is as follows:

A\B 1 2 .… j …. n
1 a11 a12 …. a1j …. a1n
2 a21 a22 …. a2j …. a2n
. . . . . . .
. . . . . . .
. . . . . . .
I ai1 ai2 …. ai3 …. ain
. . . . . . .
. . . . . . .
. . . . . . .
M am1 am2 …. amj …. amn

Here player A is called the maximizing player & B is called the minimizing player
i.e., B wishes to minimize his maximum loss.
Step 1: Select the minimum element of each row of the pay-off matrix.
i.e., max 𝑎ij i=1,2,...m.
Step 2 : Select the maximum element of each column of the pay-off matrix.
i.e., 𝑚𝑖𝑛𝑗 𝑎 j=1,2,...n.
Step 3: Obtain the maximum value of each row minimum,
i.e., 𝑚𝑖𝑛𝑗 𝑎 𝑉
Step 4: Obtain the minimum value of each column maximum,
i.e., 𝑎ij = 𝑉
If 𝑎ij = 𝑎ij = V
i.e., max(min) = min (max) , then the position of that element is a saddle point of the
pay-off matrix. And the corresponding strategy is called optimal strategy.

9
2.2 Some definitions: -
2.2a) Saddle point
A saddle point of a pay-off matrix is a position of such an element which lie both
minimum in the row & maximum in the column.

2.2b) Value of Game

The cell entry of pay-off matrix at the saddle point is the value of the game.

2.2c) Fair Game

A game is said to be fair if
Max.min = min.max = 0

2.2d) Strictly Determinable

A game is said to be strictly determinable game if
Maximin = minimax = V

2.3 Procedure to find the saddle points

1) We may now summarize the procedure of locating the saddle point of a pay-off
matrix as follows:
~Step 1: Select the minimum element of each row of the payoff matrix and mark
them with circle.
~Step 2: Select the maximum element of each column of the payoff matrix and
mark them with squares.
~Step 3: If there appears an element in the pay-off matrix with a circle & square
together, then that position is called saddle point and the element is the value of the
game.

10
2.4 Solution of the game with saddle point
To obtain a solution of game with a saddle point, It is feasible to find out
● Best strategy for player A.
● Best strategy for player B.
● The value of the game.
The best strategy for player A and B will be those which corresponds to the row and
the column respectively through the saddle point.
Example 1:
Solve the pay-off matrix:

Solution:
Column max 5 3 1 5 6
Strategy of player A: II
Strategy of player B : III
11
Value of the game = 1

Example 2: Solve the following game:

Solution:

Thus, the matrix has two saddle points at (1,1) and (1,3), thus the solution of the
game is given by

i) the best strategy for player A is I.

ii) The best strategy for player B is either I or III.

iii) The value of the game is 6 for A and –6 for B.

12
Example 3: Solve the game whose pay-off matrix is given by

Find the saddle point, strategy and value of game.

Solution: B

Hence, saddle point (1,1)

Strategies A I (row)
And for B I (column)
The value of game is +1 for A and -1 for B.

Example 4:
Find the ranges of value p and q which will render the entry (2,2) a saddle point for
the game.

13
Solution:
First ignoring the value of p and q, we determine the maximin & minimax values of
pay-off matrix as follows:

Since the entry (2,2) is the saddle point of the game.

Therefore, maximin value = 7
And minimax value = 7
This impose the condition on p as p≤7 and q as q≥7.
Hence the range of p and q will be p≤7 and q≥7.

Example 5:
The pay-off matrix of a game is given below. Find the best strategy for each player
and the value of a pay of the game to A and B.

14
To find the saddle point, putting squares around column maximum and circles
around row minimum, we get,

Column maximum 9 6 4 8 8
Obviously, the matrix has a saddle point at position (2,3).
Thus, the solution of the matrix is given by
a) The best strategy for player A is II.
b) The best strategy for player B is III.
c) The value of the game is 4 to A and -4 to B.

15
CHAPTER 3: SOLUTION OF GAME WITHOUT
SADDLE POINT

3.1 Mixed Strategy

A mixed strategy exists in a strategic game, when the player does not choose one
definite action, but rather, choose according to probability distribution over course
of his actions.

There are cases where a pure strategy equilibrium does not exist.

Therefore, we have to find the probability for which the player would be willing to
randomize between his actions. This probability will depend on the expected pay-
off of the player for each of his action.

* Important properties of optimal mixed strategies:

1) If one of the players adheres to his optimal strategy and other deviates from
his optimal strategy, then the deviating players can only decrease his yield and
cannot increase in any case (at most may be equal).
2) If one of the players adhere to his optimal strategy then the value of the game
does not alter if the opponent uses his supporting strategies any either singly
or mixed.
3) If a fixed number is added to each element of the pay-off matrix, the optimal
strategies remain unchanged while the value of the game increases by c.
4) If every element of the pay-off matrix is multiplied by a constant c, then the
optimal strategies remain unchanged while the value of the game becomes c
times the value of the original game.

16
3.2 Solution of 2x2 Games without saddle point.
We consider a 2x2 game which do not have a saddle point. So, in this case the best
strategies are the mixed strategies. i.e., here we shall determine the probabilities with
which each strategy should be selected.
For this, we have an important theorem.
Theorem: for any zero-sum two-person game where the optimal strategies are not
pure strategies and for which A’s pay-off matrix is

The optimal strategies are (x1,x2) and (y1,y2) given by

𝑥1 𝑎22 −𝑎21 𝑦1 𝑎22 −𝑎21
𝑥2 = 𝑎11 −𝑎12 and 𝑦2
=
𝑎11 −𝑎12

And the value of the game to A is given by

𝑎11 𝑎22 −𝑎12 𝑎21
ν = (𝑎
11 +𝑎22 )−(𝑎12 +𝑎21 )

Proof: If (x1,x2) and (y1,y2) are the mixed strategies for players A and B
respectively.
x1+x2=1 …............................................(1)
y1+y2=1 ................................................(2)
x1≥ 0, x2≥0, y1 ≥0, y2 ≥ 0

Now, the expected gain to A is a11x1 +a21x2 when B uses strategy I

and the expected gain to A is a12x1 +a22x2 when the B uses strategy II
Similarly, the expected loss to B is a11y1 +a12y2 when A uses strategy I
And similarly, the expected loss to B is a21y1 + a22y2 when the A uses strategy II

17
If v is the value of the rectangular game has a solution), then since A expects to get
at least v.

 ∴ a11x1 +a21x2 ≥ v .............................................(3a)

a12x1 + a22x2≥ ν ..............................................(3b)

Also, B expects to loose at most v ( value of the game).

 ∴ a12x1 + a22x2≥ ν .............................................(4a)

a12y1 + a22y2 ≤ ν ............................................(4b)

Now the problem is to find the value of x1,x2,y1,y2 satisfying (1) ,(2),(3)
and (4)

Regarding (3) and (4) strictly equations, we have

a11x1 +a21x2 = ν ....................................(5)
a12x1 + a22x2 = ν .....................................(6)
a11y1 + a21y2 = ν ......................................(7)
a12y1 + a22y2 = ν ......................................(8)
subtracting (6) and (5), we have,

(a11 – a12)x1 = (a22-a21 )x2

𝑥1 𝑎22 −𝑎21
  = …............(9)
𝑥2 𝑎11 −𝑎12

Similarly subtracting (8) from (7) and simplifying, we have

𝑦 𝑎 −𝑎
  1 = 22 12 .......................(10)
𝑦 2𝑎 −𝑎 11 21

From (9), we have,

𝑎11 −𝑎12
x2 =   x1
𝑎22 −𝑎21
∴ Putting in (1), we have,

18
 It can be verified that for all the ordering (16) of the elements of the pay-off
matrix, x1, x2, y1, y2 given by (11), (12), (13), (14) are all non-negative. Hence,
these values of x1, x2, y1, y2 given by (11),(12), (13), (14) from the solution of
the problem and the value of the game is given by (15).

3.3 Dominance Property and its examples

The principle of Dominance in Game theory also known as (dominance strategy or
dominance method) states that if one strategy of a player dominates over the other
strategy in all conditions then the later strategy can be ignored.
Clearly, a player would have no intensive to use inferior strategies which are
dominated by the superior ones.

★ Dominance Rule:
♦ Rule 1: If all the elements of the row (say ith row) of given matrix is less or
equal to the corresponding elements of other row (jth row), then ith row be
dominated (removed) by jth row.
♦ Rule 2: If all the elements of the column (say ith column) of given matrix is
greater or equal to the corresponding elements of the other column then it is
removed.
♦ Rule 3: If the elements of any row is less than equal to the average of rest
rows, then this row is dominated (removed).
♦ Rule 4: If the elements of any column is greater than or equal to the average
of rest column, then the column is dominated (removed).
In this way, we get 2x2 matrix.

19
 ★ Dominance examples

Player B
I II III IV
I
3 5 4 2
Player A II
5 6 2 4
III
2 1 4 0
IV
3 3 5 2

Use the principle of dominance to solve this problem.

Solution:

There is no saddle point in this game.

Using dominance property.
Here, we see that all the elements in the row III are less than or equal to the
corresponding elements in row I.
So, row III is dominated by row I
B
I II III IV

20
 I 3 5 4 2
A II 5 6 2 4
IV 3 3 5 2

Since the elements of column III are greater than or equal to the elements of column
I. So, II is dominated by column I & we get,

Again, we see that, all the element in row I is less than or equal to the corresponding
elements of row III. So, now I is dominated row III.

B
I II IV
II 5 2 4
IV 3 5 2

Also, the element the element of column I is greater than the corresponding elements
of column IV.
So, column I is dominated by column IV.
B
III IV
II
IV 2 4
5 2

We have: a11 = 2 ; a12= 4

21
 A21 = 5 ; a22 = 2
Now, this 2x2 game has no saddle point and cannot be reduced further.
Therefore, the optimal strategies will be mixed strategies with probabilities x 1 and
x2 respectively and B chooses his II and II strategies with probabilities y1 and y2
respectively, then by formula for 2x2, we get,
𝑎22 −𝑎21
x1=   (𝑎
11 +𝑎22 )−(𝑎12 +𝑎21 )

2−5
= 
(2+2)−(4+5)

−3
= 
4−9

−3
= 
−5

3
= 
5
𝑎11 −𝑎12
x2=   (𝑎
11 +𝑎22 )−(𝑎12 +𝑎21 )

2−4 −2
=   (2+2)−(4+5) =  
4−9

−2
=  
−5

2
=  
5

And
𝑎22 −𝑎12
y1=   (𝑎
11 +𝑎22 )−(𝑎12 +𝑎21 )

22
 2−5
=   (2+2)−(4+5)

3
= 
5

𝑎11 −𝑎21
y2=   (𝑎
11 +𝑎22 )−(𝑎12 +𝑎21 )
2−5
=  
−5
3
=  
5

And the value of the game

𝑎11 𝑎22 −𝑎12 𝑎21
v= (𝑎
11 +𝑎22 )−(𝑎12 +𝑎21 )

2×2−4×5
= (2+2)−(4+5)

4−20 −16 16
= = =
4−9 −5 5

Here, the optimal solution of the given game is

3 2
optimal strategy for player A is (0, 0, , )
5 5
3 3
Optimal strategy for player B is (0, 0, , )
5 5
16
And the value of the game is .
5

Q: Solve the game whose pay-off matrix is

−1 −2 8
7 5 −1
6 0 12

23
First of all , we find the saddle point of the given payoff matrix, by squaring greatest
element in each column & circling lowest element in each row.

As there is no element lying between circle & square.

So, there is no saddle point.
Using dominance rule,

Row I is dominated by row III & we get,

Again, column I is dominated by II & III as all the elements in column I is greater
than the corresponding element of column II & III. So, we have,

24
Comparing this table with [𝑎11 𝑎12 𝑎21 𝑎22 ] we have,
𝑎11 = 5, 𝑎12 = -1, 𝑎21 = 0, 𝑎22 = 12
Let x1 & x2 be the pure strategy for A and y1 & y2 be the pure strategy for B, then
𝑎22 −𝑎21 12−0
x1 = 〖(𝑎〗 = (5+12)
11 +𝑎22 )−(𝑎12 +𝑎21 )
12 12
= =
17+1 18
2
=
3
𝑎11 −𝑎12
x2 = 〖(𝑎〗
11 +𝑎22 )−(𝑎12 +𝑎21 )
5—1 6 1
= (5 + 12)— 1 + 0 = =
18 3
𝑎22 −𝑎12
y1 = 〖(𝑎〗
11 +𝑎22 )−(𝑎12 +𝑎21 )
12—1 5
= (5 + 12)— 1 + 0 =
18
𝑎11 𝑎12 −𝑎12 𝑎21
The value of the game =
18
60−0 10
= =
18 3
Thus,
2 1
Pure strategy for A is (0, , )
3 3
13 5
Pure strategy for B is (0, , )
18 18

3.4 Graphical method for the solution of (2×n) and (m×2) games

3.4a) Introduction
Graphical method is used to solve 2×n or m×2 games i.e. , a game with mixed
strategies which has only two pure strategies (undominated ) for one of the players.
Since the optimal strategies for both players assign non-zero probabilities to the
same number of pure strategies, therefore if one player has only two pure strategies,
the other player will also use two strategies only. Graphical method helps us to find
which two strategies should be used. Consider a 2×n game which has no saddle point
whose pay-off matrix is as follow :

25
If x1 and x2 are the probabilities with which the player A uses his pure strategies,
then
x1 +x2= 1; x1 ≥0, x2 ≥ 0
x2 =1-x1

The expected pay-off to player A for different pure strategies used by player B are
tabulated as follows :

Pure strategies E(v), expected pay-off to player A

Used by player B
1 a11x1 + a21x2 = a11x1 + a21(1-x1) = (a11-a21)x1 + a21
2 a21x1 + a22x2 = a12x1 + a22(1-x1) = (a12 –a22)x1 +a22
…. …. ….. …..

…. …. ….. …..

N a1nx1 + a2nx2 = a1nx1 + a21(1- x1) = (a1n –a2n)x1 + a2n

Thus, it is clear that A’s expected pay-off varies linearly with x1. From the minimax
criterion for mixed strategies game the player A will select that value of x1 which
will maximize his minimum expected pay-off. To find this value we plot the
following straight line as function of x1 :
E(v) = 〖(𝑎〗11 − 𝑎21 )x1 + a21
E(v) = (a12 –a22)x1 + a22
… … … …
E(v) = (a1n- a2n)x1 + a2n

26
3.4b) Method to plot lines
To plot the expected pay off lines, we draw two parallel lines one unit apart and
mark a scale on each as shown in fig.

These two lines represent the two strategies available to A. then we draw lines to
represent each to B’s strategies. To represent B’s 1st strategy (see pay-off matrix)
we join a on scale 1 to a21 on scale 2. This line will represent the expected pay-off
line E(v)=(a11-a21)x1 +a21 with x1 as x-axis and E(v) as y axis.
Similarly, we may draw other pay off lines.
The lower boundary to these lines will give the minimum expected pay-off as
function of x1. The highest point ‘P’ on this lower boundary will give the maximum
expected pay-off to A and hence optimum value of x1. Then we determine only two
strategies for player B corresponding to those two lines which pass through this
maximum point P. if more than two lines pass through this point then any two of
them having opposite signs for their slopes will be alternative optimum solutions.
27
In this way the game is reduced to 2x2 game which can be solved easily.

3.4 c) Examples
We can solve m ×2 games in the same manner except that mini9max point P is the
lowest point on the upper boundary.
Example 1: Solve the game whose pay-off matrix is

Solution: This 2×4 game has no saddle point. Therefore, we shall use graphical
method to reduce this game to 2×2 game.
If x1 and x2 are the probabilities with which the player A uses his pure strategies
then
x1+ x2 =1, x1 ≥ 0, x2 ≥ 0
x2 = 1-x1
the expected pay-off to player A for different pure strategies used by player B may
be tabulated as follows:
Pure strategy Used by E(V), A’s Expected pay-off
player B

I E(v) =1x1+2x2 =x1+2(1-x1) = -x1+2

II E(V)=3x1+5x2=3x1+5(1-x1) = -2x1+5
III E(V)= -3x1+4x2=-3x2+4(1-x1) = -7x1+4
IV E(V)=7x1-6x2=7x1-6(1-x1) =13x1

Now we draw the four pay-off lines in the graph.

E(v) = -x1 + 2
28
 E(v) = -2x1 + 5
E(v) = -7x1 + 4
E(v) = 13x1 – 6
To draw these lines first we draw two parallel lines one unit apart and mark a scale
on each as shown in fig. these two lines represent two strategies available to A. Now
join on scale 1 and 2 on scale 2. This line represents B’s 1st strategies i.e., the pay-
off line
E(v) = - x1 +2

Similarly, joining 3 on scale 1 to 5 on scale 2, -3 on scale 1 to 4 on scale 2 and 7 on

scale 1 and -6 on scale 2, we get the other pay-off lines respectively. Now lowest
boundary APB (thick line in fig) to these lines give the minimum expected pay off
to A. the highest point P on this lower boundary will give the maximum expected
pay and hence the expected value o x1. Thus the best strategies for player B are III
and IV pure strategies passing through point P.

29
Therefore, the game is reduced to 2 ×2 game given by the following pay-off matrix:

If y1, y2 are the probabilities with which player B chooses his best strategies III and
IV respectively. By using the formulae. we have (here a11 = -3, a12 =7, a21 = 4, a22 =
-6)

1 1 13 7 1
x1= , x1= , y1= , y2= , v= .
2 2 20 20 2

Solution of the game is

1 1
(i)for player A optimal mixed strategies are ( , )
2 2

13 7
(ii) for player B optimal mixed strategies are ( 0, 0 , , )
20 20

1 1
And (iii) The value of the game is to A and - - to B .
2 2

30
3.5) Equivalence of the rectangular Matrix Game and Linear
Programming:

Probabilities y1 y2 ..... yj ….. yn

Pure strategies 1 2 ….. j ….. n

x1 1 a11 a12 …. aij ….. a1n

x2 2 a21 a22 …. a2j ..... a2n
: : …. …. …. …. …. ….
: : …. …. …. …. …. ….
A xi J ai1 ai2 …. aij …. ain
: : …. …. …. …. …. ….
: : …. …. …. …. …. ….
xm M Am1 am2 ..... amj …. amn

Consider a rectangular game played by two players A (maximizing player) and B

(minimizing player) with pay-off matrix [aij]m×n.
Let X = [x1, x2 , …. xm], Y= [y, y2, ….. yn] be the mixed strategies of the two players
A and B respectively by which A and B select their pure strategies.
We have shown that A selects his optimal mixed strategies which will
Min.[ Max. { ∑𝑚 𝑚 𝑚
1=1 𝑎𝑖1 𝑥𝑖 }, ∑1=1 𝑎𝑖2 𝑥𝑖 ], ….. ∑1=1 𝑎𝑖𝑛 𝑥𝑖 }]\
Such that x1 + x2 + …+ xm =1
and xi ≥0, Ɐ i = 1, 2, …. m.

also B select his optimal mixed strategies which will

Min.[ Max. { ∑𝑛𝑗=1 𝑎1𝑗 𝑦𝑗 }, ∑𝑛𝑗=1 𝑎2𝑗 𝑦𝑗 ], ….. ∑𝑛𝑗=1 𝑎𝑚𝑗 𝑦𝑗 }]
Such that y1 + y2 + …+ yn =1
and yj ≥0, Ɐ j = 1, 2, …. n.

31
if Min.{ ∑𝑚 𝑚 𝑚
1=1 𝑎𝑖1 𝑥𝑖 }, ∑1=1 𝑎𝑖2 𝑥𝑖 ], ….. ∑1=1 𝑎𝑖𝑛 𝑥𝑖 }= v
Then A expects to gain at least v. The criterion for A to choose his strategy is to
maximize his least gain v.
Since v is minimum of all expected gains therefore, we have
{ ∑𝑚 𝑚 𝑚
1=1 𝑎𝑖1 𝑥𝑖 }, ∑1=1 𝑎𝑖2 𝑥𝑖 ], ….. ∑1=1 𝑎𝑖𝑛 𝑥𝑖 }≥ v
Thus, A’s problem is to determine x1 , x2 , …xn ,
To Maximize Z= v
Subject to the constraint

a11x1 + a21 x2 + … + am1 xm ≥ v

a12 x1 + a22 x2 + … + am2 xm ≥ v
……………………….............. ……………….(1)
………………………......…....
a1nx1 + a2n x2 + … + amn xm ≥ v
x1 + x2 + …+ xm=1
and x1 , x2 , ……xm ≥ 0

We take v to be positive. If it is not positive then we make it simply by adding a

large quantity to all terms of payoff matrix. Again dividing equation by 1 by v we
take x1/v = X1, x2/v = x2,……………………….,x2/v = Xm.
We have
a11x1 + a21 x2 + … + am1 xm ≥ 1
a12 x1 + a22 x2 + … + am2 xm ≥ 1
………………………......…....
………………………......…....
a1nx1 + a2n x2 + … + amn xm ≥ 1
and X1 + X2 + ….. + XM = 1/v
X1, X2, ….. , XM ≥1/v
Since Max. v = min. (1/v)
𝑥1 +𝑥2 +⋯+𝑥𝑚
= Min{ 𝑣 }

Min. (x1+ x2 +…… xm)

32
Thus, the given rectangular game reduces to the following L.P.P
Min. x* = 1/v = X1 + X2 + …. Xm
Subject to the constraints
a11x1 + a21 x2 + … + am1 xm ≥ 1
a12 x1 + a22 x2 + … + am2 xm ≥ 1
………………………......….... ………(2)
………………………......…....
a1nx1 + a2n x2 + … + amn xm ≥ 1
and X1 ≥ 0, X2 ≥ 0, …..,Xm ≥ 0

Now considering the problem from B’s point of view who want to minimize v
(because the gain of A is the loss of B), proceeding in the similar manner we get
following L.P.P:
Maximize y* =1/v = Y1 + Y2 + … + Yn
Subject to the constraints
a11Y1 + a12 Y2 + … + a1n Yn ≤ 1
a21 Y1 + a22 Y2 + … + a2n Yn ≤1
………………………......….... ………(3)
………………………......…....
am1Y1 + am2 Y2 + … + amn Yn ≤1
and Y1 ≥ 0, Y2 ≥ 0, …..,Yn≥ 0
where y* =1/v, Y1 =y1/v, Y2= y2 /v , ….. Yn = yn/v
It can be seen that the L.P.P given in (2) and (3) are the duals of each other i.e., B’s
problem is the dual of A’s problem and vice-versa. Therefore, if one problem is
solved then the other is solved automatically.

After getting the values of Xi, Yj and min. of ∑ Xi which is equal to Max. of of ∑ 𝑌j,
we can find the value of xi and yj from
xi = v Xi and yj = v Y j

33
3.6 Solution of Rectangular Game by Simplex Method:
If a problem has no saddle point, then we convert it to a L.P. problem for player B
by the method discussed above. This L.P. problem in solved by simplex method and
from final simplex table of the solution we can read the solution for player A by the
dual method.
For clear understanding of the method see the following example:
Example: find the best strategies and the value of the following game:

Solution: first we construct the table for row minimums and column maximums
as follows:

It is clear from the above table that the value v of the game lies between -1 and 3.
i.e. -1 ≤ v ≤3
⸫ It is possible that the value of the game may be negative or zero. Thus, we add a
constant K =4 to all the element of the matrix so that all these elements of the matrix
34
become positive assuming that the value of the game represented by this new matrix
is non- negative and non-zero. The transformed (reduced) matrix is as follows:

Let X = (x1, x2, x3) and Y = (y1, y2, y3) be the mixed strategies of the two players
respectively, where x1 , x2 , x3 ; y1, y2 , y3 are the probabilities with which they choose
their pure strategies.
Therefore, B’s problem is
To find (y1, y2, y3) which minimize v’
Subject to
5y1 + 3y2 + 7y3 ≤ v’
7y1 + 9y2 + y3 ≤ v’
10y1 + 6y2 + 2y3 ≤ v’
And y1+y2+y3= 1
y1, y2, y3 ≥ 0
Here v’ > 0, dividing above equations by v’ an putting y 1/v’ =Y1 , y2/v’ = Y2, y3/v’
=Y3 the B’s problem reduces to the following L.P. problem :
𝑦1 +𝑦2 +𝑦3
Max. Z = = Y1+Y2+Y3:
𝑣′
Such that 5Y1 + 3Y2 + 7Y3 ≤ 1
7Y1 + 9Y2 + Y3 ≤ 1
10Y1 + 6Y2 + 2Y3 ≤ 1
And Y1, Y2, Y3 ≥ 0
Introducing the slack variables, the constraints reduce to the following equations:
5Y1 + 3Y2 + 7Y3+Y4 = 1
7Y1 + 9Y2 + Y3+ Y5 = 1
10Y1 + 6Y2 + 2Y3+Y6 =1
35
Taking Y1=0, Y2=0, Y3=0, we have Y4=0, Y5=1, Y6=1, which is the starting B.F.S.
computation work by simplex method is shown in the following table:

⸫Y1 =0, Y2 = 1/10, Y3 =1/10 and Max. Z = 1/5

From Z =1/v’ we have Min. v’ =5
⸫y1 =Y1v’=0, y2=Y2v’ = 1/2, y3 =Y3v’=1/2
And value of the original game v = v’-4 = 1.
Since A’s strategy are the dual of the above problem

36
 𝑥1 2 𝑥2 1 𝑥3
⸫ X1= = , X2= = , X3= =0
𝑣′ 15 𝑣′ 15 𝑣′
(value of Δ4, Δ5, Δ6 with sign changed)
2 1
⸫x1= , x2= , x3=0.
3 3
Hence, the optimal solution is
(i)Best strategy for A (2/3, 1/3 , 0)
(ii)Best strategy for B (0, 1/2 , 1/2)
(iii) Value of the game v = 1

37
Chapter 4: Applications and Limitations of Game Theory

4.1) Application
Game theory is not just theory; it’s also applied to economics, business, biology,
computer science, political science, psychology and philosophy. Game theory can
describe a number of specific phenomena: interpersonal relations, competition, war
and political affairs. From historical aspects game theory can be identified in the
works of ancient philosophers. It is applied to develop theories of ethical or
normative behavior. Economists and philosophers have applied game theory to
understand ration behavior. There are some merits of game theory in the given
below:
⁎ Game theory possesses the following merits: [2]
1.) Game theory shows the importance to duopolists of finding some way to agree.
It helps to explain why duopoly prices tend to be administered in a rigid way. If
prices were to change often, tacit agreements would not be found and would be
difficult to enforce.
2.) Game theory tries to explain how the duopoly problem cannot be determined.
For this, it uses the solution without saddle point under constant-sum-two-person
game. At the same time, the duopoly problem without a saddle point is solved by
allowing each firm to adopt mixed strategies on a probability basis. In this way, the
duopoly problem is shown to be always determined.
3.) “Prisoner’s Dilemma” in game theory points towards collective decision making
and the need for cooperation and common rules of road.
4.) Again, game theory is helpful in solving the problems of business, labour and
management. As a matter of fact, a businessman always tries to guess the strategy
of his opponents so as to implement his plans more efficiently. Similar is the case of
management in trying to solve the problem of union’s bargaining for higher wages.
Management might adopt the most profitable counter- strategy to tackle such
problems. Further, producers might make decisions in which estimation of profits
were to be balanced against the cost of production.
5.) There are certain economic problems which involve risk and technical relations.
They can be handled with the help of the mathematical theory of games. Problems

38
of linear programming and activity analysis can provide the main basis for economic
application of the theory of games.

⁎ Application and applied uses:

1.) Economic and business

Game theory is a major method used in mathematical economics and business
for modeling competing behavior of interacting agents. Applications include
a wide array of economic phenomena and approaches, such as auctions,
bargaining, mergers, and acquisitions pricing fair division, duopolies,
oligopolies, social network formation, agent based computational economics,
general equilibrium, mechanism design and voting system, and across such
broad areas as- experimental economics, industrial organization and political
economy.[3]

2.) Project management

Sensible decision-making is critical for the success of projects. In project
management, game theory is used to model the decision making process of
players, such as investors, project manager, contractors, subcontractors,
governments and customers. Quite after, these players have competing
interests and sometimes their interests are directly deterimental to other
players making project management boundaries well- suited to be modeled by
game theory. [4]

3.) Political science

The applications of game theory in political science is focused in the
overlapping areas of fair decisions, political economy, public choice, war
bargaining, positive political theory, and social choice theory. In each of these
areas, researchers have developed game: - theoretic models in which the
players are often voters, states, special interest groups and politicians.

39
4.) Biology
In biology game theory has been used as a model to understand many different
phenomena. It was first used to explain the evolution (any stability) of the
approximate 1:1 sex ratios. Fisher (1930) suggested that the 1:1 sex ratios are
a result of evolutionary forces acting on individuals who could be seen as
trying to maximize their number of grandchildren.

Additionally, biologists have used “evolutionary game theory” and ESS

(Evolutionary Stable Strategy) to explain the emergence of animal
communication [5].

Biologists have used the game of chicken to analyze fighting behavior and
territoriality [6].
According to Maynard Smith, in the preface to Evolution and the Theory of
Games, “paradoxically, it has turned out that game theory is more readily
applied to biology than to the field of economic behavior for which it was
originally designed”.
Evolutionary game theory has been used to explain many seemingly
incongruous phenomena in nature [7]. One such phenomena is known as
biological altruism. This is a situation in which an organism appears to act in
a way that benefits other organisms and is deterimental to itself.
Evolutionary game theory explains this altruism with the idea of kin selection.

5.) Computer science and logic

Game theory has come to play an increasingly important role in logic and
computer science. Several logical theories have a basis in game semantics.
In addition, computer scientists have used games to model interactive
computations. Also, game theory provides a theoretical basis to the field of
multi-agent systems[8].
Separately, game theory has played a role in online algorithm; in particular,
k-server problem, which has in the past been referred to as games with
moving cost & request answer games.

40
 6.) Retail and Consumer Product Pricing
Game theory applications are often used in pricing strategies of retail and
consumer markets particularly for the sale of inelastic goods. With retailers
constantly competing against one-another for consumer market share, it has
become a fairly common practice for retailers to discount certain goods,
intermittently, in the hopes of increasing fast-traffic in brinck and mortar
locations, or increasing sales of ancillary or complimentary products [9].

4.2 Limitations of Game theory [10]

a) Game theory assumes that each firm has knowledge of the strategies of the
other as against its own strategies and is able to construct the pay-off for a
possible solution.
This is highly unrealistic assumption & has little practicability.
e.g., an entrepreneur is not fully aware of the strategies available to him, much
less those to his rival. He can only have a guess of his & his rival’s strategies.
b) The theory of games assumes the both player to be prudent men. Each rival
moves on this presumption that his opponent will always make a wise move
& then he adopts a countermove, which is an unrealistic assumption.
c) Game theory makes it easy to understand a two-person constant-sum game.
But as the analysis is elaborated to three or four person games, it becomes
complex and difficult.
However, the theory of game has not been developed for games with more
than four players.
d) Further, the minimax principal which provides a solution to the constant-sum
game assumes that each player makes the best of worst possible situation.
How can be the best situation be known if the worst does not arise?
Moreover, most entrepreneurs act on the presumptions of the existence of
favorable market conditions and the question of making the best of the worst
does not arise at all.
e) The various strategies followed by a rival against the other lead to an endless
chain of thought which is impracticiable. There is no end to chain of thought
when A choose one strategy & B adopts a counter-strategy and vice-versa.

41
 CHAPTER 5: CONCLUSION

Game theory is a classic theory which is applicable to most of the field.

The main significant of the game theory is to formulate the alternative strategy
to compete with one another and in the same sense it is an essential tool for
decision making process according to fluctuations in relevant contents.
Game theory is a mathematical framework for strategy analysis and design as
well as for optimal decision moving under conflict & behavioral uncertainty.
On the one hand, game theory plays a key role for modern economics; on the
other, it suggests possible approach & solution for complex strategic problems
in various fields of human activity.
Strategic interaction has proved to be a powerful idea & although its
application, especially beyond economics, remains controversial, it has proven
fruitful in suggesting new perspectives and new ways of formalizing older
insights.

42
 References:

1.) R.K. Gupta, Operation Research, Krishna Publication, 2020.

2.) Dominance:GameTheory
http://www.universalteacherpublicsation.comuniv/ebooks/orch9/domin.htm

3.) Praveen, Mahendra,-“Application of Game theory in Project management:

A structural review & Analysis”.

4.) Game Theory in economics:

https://www.yourarticlelibrary.com/economics/game-theory-in-economics-
importance-limitations-and-other-details/28930

5.) Harper & Maynard Smith (2003)

6.) Maynard smith, John(1974)-“ The theory of games and the evolution of
animal conflicts”. P.No:209-221

7.) Alexander, J. Mckenzie –“ Evolutionary Game Theory” ,Stanford

encyclopedia of philosophy.

8.) Shoham, Yoa v; Leyton- Brown, Kelvin (15 Dec2008)-“Multiagent

systems: Algorithmic, Game- theoretic and logical functions”. Cambridge
University Press

9.) Kopalle; Shumstry- “Game theory models of pricing”. (PDF)

10.) Game Theory in economics:

https://www.yourarticlelibrary.com/economics/game-theory-in-economics-
importance-limitations-and-other-details/28930

43