Game Theory:

We have considered many problems in LPP for finding optimal solutions. In

all these problems, we had only one decision maker and he was free to
decide what to produce and what not to produce. However, this is a very
idealistic situation and in many real life problems, there are more than one
decision makers whose decisions are conflicting with each other. In such a
situation, one competitor called a Player has many options called strategies
and another competitor called player B also has many counter options to
the strategies of player B. This is called a game problemof two players.
`
Pay off Matrix: Each combination of alternatives of Players A and B is
associated with an outcome aij. If aij is positive, then it represents a gain to
player A and loss to player B. If it is negative, it represents a loss to A and
gain to B. A sample pay off matrix of player is given below:

Player B
I II … n
I a11 a12 … a1n
Player A II a21 a22 … a2n
. . . . .
. . . . .
. . . . .
m am1 am2 … amn

MaxMin Principle: The principle maximizes the minimum

guaranteed gains of Player A. The minimum gains with respect to
different alternatives of A, irrespective of B’s alternatives are
obtained first. The maximum of these minimum gains is known as
the maximum value and the corresponding alternative is called as
maxmin strategy.
MinMax Principle. The principle minimizes the maximum losses of
Player B. The maximum losses with respect to different
alternatives of Player B, irrespective of Player A’s alternatives are
obtained first. The minimum of these maximum losses is known
as the minmax value and the corresponding alternative is called
as minmax strategy.

Saddle Point: If in a game, Maxmin value is equal to minmax

value, then the game is said to have a saddle point and the game
is said to be stable.In this case each player will choose his pure
strategy.

Value of the game: If the game has a saddle point, then the value
of the cell at the saddle point in the payoff matrix is called the
value of the game.`

Two-person Zero-sum game: In a game with two players, if the

gain of one is equal to loss of the other, then that game is called
two-person zero sum game.

Exampe1

Player B Min of max

each
row
I II III
Player A I 1 3 6 1
II 2 1 3 1 1
III 6 2 1 1
Max of 6 3 6
each
colum
n
min 3

Saddle Point: In our case Minimax is not equal to MaxMin. Hence, no

Saddle Point exist.
Ex.2

Player B Min of max

each
row
I II III
Player A I 20 15 22 15
II 35 45 40 35 35
III 18 20 25 18
Max of 35 45 40
each
colum
n
min 35

Saddle Point: In this case, Minmax is equal to MaxMin. Hence, Saddle

Point exist. Game is stable.
The common cell (A2,B1) in the game matrix has the maxmin=minmax.
Therefore the optimal pure strategies of both the players are II and I and
the value of the is 35.The player A has a gain of Rs.35 and player B has a
loss of Rs.35.The strategies being used by both the players are pure
strategies.

Player B Min of max

each
row
I II III IV
Player A I 1 7 3 4 1
II 5 6 4 5 4 4
III 7 2 0 3 0
Max of 7 7 4 5
each
column
min 4
Saddle point at A2,B3 and optimal value of the game is 4. The player A has
a gain of Rs.4 and Player B has a loss of Rs.4.
Graphical method
The graphical method is used to solve the games whose payoff matrix
has two rows and n columns (2 x n) or m rows and two columns (m x 2)
Algorithm for solving 2 x n matrix games

Draw two vertical axes 1 unit apart. The two lines are x1 = 0, x1 = 1
Take the points of the first row in the payoff matrix on the vertical
line x1 = 1 and the points of the second row in the payoff matrix on
the vertical line x1 = 0.
The point a1j on axis x1 = 1 is then joined to the point a 2j on the axis
x1 = 0 to give a straight line. Draw ‘n’ straight lines for j=1, 2… n
and determine the highest point of the lower envelope obtained.
This will be the maximin point.
The two or more lines passing through the maximin point
determines the required 2 x 2 payoff matrix. This in turn gives the
optimum solution by making use of analytical method.

Ex.3

Player B Min of max

each
row
I II III IV
Player A I (x1) 2 2 3 -1 -1
II (x2) 4 3 2 6 2 2
Max of 4 3 3 6
each
column
min 3

Thus MaxMin=2 and MinMax=3. Thus the game does not have saddle
point. Let the player A choose his strategies I and II with probabilities x1
and x2 where x2=1-x1 as x1+x2=1.
We can solve a game problem if we have a game of the dimension 2xn or
mx2.
A’s expected payoffs corresponding to B’s strategies are given as follows.

B’s pure strategies Expected payoff x1=0 x1=1

1 2x1+4x2 or 2x1+4(1-x1)= -2x1+4 4 2
2 2x1+3x2 or 2x1+3(1-x1)= -x1+3 3 2
3 3x1+2x2=x1+2 2 3
4 –x1+6x2= -7x1+6 6 -1
To determine B’s optimal strategies, it should be noticed that three lines
pass through the Maxmin point. This is an indication that B can mix all
three strategies. Any two lines having opposite signs for their slopes, define
an alternative optimum. Thus, the three combinations (II,III), (II,IV) and
(III,IV) are optimal whereas (I,IV) is non optimal.

Player B
II III
Player A I 2 3
II 3 2
or

Player B
II IV
Player A I 2 3
II 3 2

or

Player B
III IV
Player A I 3 -1
II 2 6

We will use analytical method, an alternative approach to find the

probabilities associated with the strategies using any of the above
combinations.

Using the first option, we have

Player B
II (y2) III (y3)
Player A I (x1) 2 3
II (x2) 3 2
We write as

2x1+3x2= ν
3x1+2x2= ν

Or 2x1+3(1-x1)= ν
3x1+2(1-x1)= ν

Or -x1+3= ν
x1 +2= ν
or -2x1+1=0 or x1=1/2 and x2 =1-x1=1-1/2=1/2
For finding probabilities associated with Player B’ strategies,
We write as
2y2+3y3= ν
3y2+2y3= ν
Or 2y2+3(1-y2)= ν
3y2+2(1-y2)= ν

Or -y2 +3= ν
y2 +2= ν

or –2y2+1=0

or y2 = 1/2 and y3=1-y2=1-1/2=1/2

Thus player A will play his strategies I and II with probabilities 1/2 each and
player B plays his strategies II and III with probabilities 1/2 each.

Thus, We writes as ( A1, A2) with (x1*,x2*) =(1/2, 1/2)

and (B2,B3) = (y2*,y3*)=(1/2, 1/2)
We also calculate the value of the game as
V*= a12.x1*.y2*+a13.x1*.y3*+a22.x2*.y2*+a23.x2*.y3*
=2.1/2.1/2+3.1/2.1/2+3.1/2.1/2+2.1/2.1/2
= 2/4+3/4+3/4+2/4
= 10/4
= 5/2 (The same result as obtained earlier)
Algorithm for solving m x 2 matrix games

Draw two vertical axes 1 unit apart. The two lines are x1 =0, x1 = 1

Take the points of the first row in the payoff matrix on the vertical line x 1 =
1 and the points of the second row in the payoff matrix on the vertical line x 1
= 0.
The point a1j on axis x1 = 1 is then joined to the point a 2j on the axis x1 = 0 to
give a straight line. Draw ‘n’ straight lines for j=1, 2… n and determine the
lowest point of the upper envelope obtained. This will be the minimax
point.
The two or more lines passing through the minimax point determines the
required 2 x 2 payoff matrix. This in turn gives the optimum solution by
making use of analytical method.

Example 4.

Player B
I (y) II (1-y) Min Max
of
each
row
Player A I 1 3 1
II 3 1 1
III 5 -1 -1 1
IV 6 -6 -6
Max of 6 3 1
each
column
min 3
Solution:

Here the saddle point does not exist. Since we have only two strategies for
player B, We can apply Graphical Method.

Expected Payoff B

A’s Alternative B’s expected Payoff y=0 y=1

I y+3(1-y)= -2y+3 3 1
II 3y+(1-y) = 2y+1 1 3
III 5y-(1-y) = 6y-1 -1 5
IV 6y-6)1-y)= 12y-6 -6 6

We plot the above four lines as:

Thus we select any two lines with opposite slopes passing through the
minmax point P namely lines I and IIIwith following payoff matrix
 Player B
I(y1) II(y2)
Player A I (x1) 1 3
III (x3) 5 -1

For optimal strategies of player A, we have

x1+5x3= ν
3x1-x3= ν

Or x1+5(1-x1)= ν
3x1-(1-x1)= ν

Or -4x1+5= ν
4x1 -1= ν

or -8x1+6=0 or x1=3/4 and x3 =1-x1=1-3/4=1/4

We write as
y1+3y2= ν
5y1-y2= ν
Or y1+3(1-y1)= ν
5y1-(1-y1)= ν

Or -2y1 +3= ν
6y1 -1= ν

or –8y1+4=0

or y1 = 1/2 and y2=1-y1=1-1/2=1/2

Thus player A will play his strategies I and III with probabilities 3/4 and
1/4respectively while player B plays his strategies I and II with probabilities
1/2 each.
Thus, We writeoptimal strategies as ( A1, A3) with (x1*,x3*) =(3/4, 1/4)
and (B2,B3) = (y1*,y2*)=(1/2, 1/2)
We also calculate the value of the game as
V*= a11.x1*.y1*+a12x1*.y2*+a31.x3*.y1*+a32.x3*.y2*
=1.3/4.1/2+3.3/4.1/2+5.1/4.1/2+(-1)1/4.1/2
= 3/8+9/8+5/4-1/8
= 2 (The same result as obtained earlier)
Dominance Principle)

A strategy dominates over the other only if it is preferable over

other in all conditions. The concept of dominance is especially
useful for the evaluation of two-person zero-sum games where
a saddle point does not exist.

Generally, the dominance property is used to reduce the size of a

large payoff matrix.

Dominant Strategy Rules:If all the elements of a column

(say ith column) are greater than or equal to the corresponding
elements of any other column (say jth column), then the ith column
is dominated by the jth column and can be deleted from the matrix.

 If all the elements of a row (say ith row) are less than or equal
to the corresponding elements of any other row (say jth row),
then the ith row is dominated by the jth row and can be
deleted from the matrix.

Dominance Example: Game Theory

Use the principle of dominance to solve this problem.

Solution:

Player B

I II III IV Minimum

I 3 5 4 2 2

Player A II 5 6 2 4 2

III 2 1 4 0 0

IV 3 3 5 2 2

Maximum 5 6 5 4

There is no saddle point in this game.

Using Dominance Property In Game Theory

If a column is greater than another column (compare corresponding

elements), then delete that column.
Here, I and II column are greater than the IV column. So, player B has no
incentive in using his I and II course of action.

Player B

III IV

I 4 2

Player A II 2 4

III 4 0

IV 5 2

If a row is smaller than another row (compare corresponding elements),

then delete that row.Here, I and III row are smaller than IV row. So, player
A has no incentive in using his I and III course of action.
 Player B

III IV

Player A II 2 4

IV 5 2