GAME THEORY

Decision Science 302

Pure Strategy Game
Strategy Game
A) PURE STRATEGY GAME: In pure strategy game players stay with the one strategy
game throughout the game. These are solved as follows:
STEP 1: Write down the pay off matrix with the maximizing player’s strategies along the
rows and the minimizing player’s strategies along the column.
STEP 2: Write down minimum element in each row ( row minima) on its right. Identify the
largest element among these element and mark it with an arrow. This principle of
identifying the maximum element among the row minimax is called “ maximin”.
STEP 3: Write down maximum element in each column ( column maxima) at its bottom.
Identify the smallest element among these element and mark it with an arrow. This principle
of identifying the minimum element among the column minimax is called “ Minimax”.
A) PURE STRATEGY GAME:
STEP 4: If these two arrowed elements( ie. Maximin and minimax) are same then the
element lying at the intersection of the corresponding row and column is called as “ Saddle
Point”.
STEP 5: The strategy corresponding to the saddle point ( represented by the corresponding
row and column heading) are the optimal strategies for the two players while the saddle
point is the value of game.
STEP 6: If the arrowed elements ( maximin and minimax ) are not same then there is no
saddle point and the value of the game lies between these two values. This is the case of
mixed strategy games.
Q. The following is the pay of matrix of a game being played by A and B.
Determine the optimal strategies for the players and the value of the game.

Player A Player B

B1 B2 B3

A1 12 -8 -2

A2 6 7 3

A3 -10 -6 2
 Player A Player B

B1 B2 B3 Row Minima

A1 12 -8 -2 -8

A2 6 7 3 3 <= Maximin

A3 -10 -6 2 -10

Column Maxima 12 7 3 = Minimax

As Minimax= Maximin Hence it is Pure Strategy Game

Stgy. Of Player A= (0,A2,0)= (0,1,0)= (0%,100%,0% )
Stgy. Of Player B= (0,0,B3)= (0,0,1)= (0%,0%,100%, )
Value of the Game= V=3 .
Q. Solve the following Game.
PLAYER Y

1 2 3 4 5

PLAYER I 1 3 2 7 4
X
II 3 4 1 5 6

III 6 5 7 6 5

IV 2 0 6 3 1
 PLAYER Y

1 2 3 4 5 Row Minima

PLAYER X I 1 3 2 7 4 1

II 3 4 1 5 6 1

III 6 5 7 6 5 5<= Maximin

IV 2 0 6 3 1 0

Column Maxima 6 5 <= Minimax 7 7 6

As Minimax= Maximin Hence it is Pure Strategy Game

Stgy. Of Player X= (0,0,III,0)= (0,0,1,0)= (0%,0%, 100%,0%)
Stgy. Of Player Y= (0,2,0,0)= (0,1,0,0)= (0%,100%, 0%,0%)
Value of the Game= V= 5 .
B) MIXED STRATEGY GAME:

When for a given game , we do not get a saddle point ( ie. maximin ≠minimax ) it is a case of mixed strategy game.
Here the players mix the different strategies with certain probabilities or proportions so as to optimized the
expected pay offs. We use arithmetical method to find the solution of 2 x 2 game with mixed strategies ( ie. without
saddle point ) as follows: Consider the pay off matrix

B1 (q) B2 (1-q )

A A1(p)
a a
11 12
A2 (1-p)
a a
21 22
B) MIXED STRATEGY GAME:
STEP 1: Let p= Probability that A use strategy A1 , 1-p= Probability that A use strategy A2
q= Probability that B use strategy B1 , 1-q= Probability that B use strategy B2
STEP 2: For player A, p and 1-p should be such that he should have same pay offs irrespective of B using
B1 or B2. Thus,
Expected pay offs to A when B uses B1= Expected pay offs to A when B uses B2

P= ( a –a ) / ( a +a –a –a )
22 21 11 22 21 12
Similarly

q= ( a
–a ) / ( a +a –a –a )
22 12 11 22 21 12
STEP 3: The value of the game ( for A )

ie. v= (a a –a a ) / ( a +a –a –a )
11 22 21 12 11 22 21 12
The value of the game for B will be ( -ve) .
B) MIXED STRATEGY GAME:
NOTE:

Let D= ( a +a –a –a )
11 22 21 12
Now Probability of using first row A1 is

P= ( second row difference starting with a22 ) / D = ( a –a )/D

22 21
Similarly Probability of using first column B1 is

q= ( second column difference starting with a22 ) / D = ( a –a )/D

22 12
Value of the game for A is
V= ( Difference of cross products of diagonal elements ) / D

= (a a –a a )/D
11 22 21 12
Q. Solve the following Game. ⇒ Solution -

B1 B2 B1 B2 Row
Minima
A1 6 -3
A1 6 -3 -3 <=
A2 -3 0 Maximin

A2 -3 0 -3 <=
Maximin

Column 6 0
Maxima Minimax

As Minimax Maximin
≠

Hence it is mixed Strategy Game

Ans.
B1 B2
B1 B2 q 1-q
A1 6 -3
A1 P 6 -3
A2 -3 0
A2 1-p -3 0
p= ( a –a )/(a +a –a –a ) = (0+3)/ (6+0+3+3) = 3/12
22 21 11 22 21 12
1-p= 1-3/12 = 9/12

q= ( a –a )/(a +a –a –a ) = (0+3)/ (6+0+3+3) = 3/12

22 21 11 22 21 12
1-p= 1-3/12 = 9/12

v= (a a –a a ) / ( a +a –a –a ) = [6*0-(-3)(-3)]/ (6+0+3+3) = -9/12

11 22 21 12 11 22 21 12
p= 3/12 , 1-p= 9/12 , q = 3/12 1-p =9/12 v = - 9/12

Stgy. Of Player A= (A1,A2)= (3/12,9/12)= (25%,75%)

Stgy. Of Player B= (B1,B2)= (3/12,9/12)= (25%,75%)

Value of the Game= V= -9/12 . As value of game v= -9/12 hence player B win
the game.
Q. Solve the following Game.

Player B

Player A B1 B2

A1 1 7

A2 6 2
 Player B

Pl B1 B2 Row Minima
ay
er A1 1 7 1
A A2 6 2 2 <= Maximin

Column Maxima 6 7
Maximin

As Minimax Maximin
≠

Hence it is mixed Strategy Game

 Player B B1 B2
Player B1 B2 q 1-q
A
A1 1 7 A1 p 1 7
A2 6 2 A2 1-p 6 2

p= ( a22 –a21 ) / ( a11 +a22 –a21 –a12 ) = (2-6)/ (1+2-6-7) = -4/ -10 = 4/10=2/5
1-p= 1-2/5 = 3/5
q= ( a22 –a21 ) / ( a11 +a22 –a21 –a12 ) = (2-7)/ (1+2-6-7) = -5/-10 =5/10=1/2
1-p= 1-1/2 = 1/2
v= (a11 a22 –a21 a12) / ( a11 +a22 –a21 –a12 ) = [2*1-6*7]/ (1+2-6-7) = -40/-10 = 4
p= 2/5 , 1-p= 3/5 , q = 1/2 1-p =1/2 v = 4

Stgy. Of Player A= (A1,A2)= (2/5,3/5)= (40%,60%)

Stgy. Of Player B= (B1,B2)= (1/2,1/2)= (50%,50%)

Value of the Game= V= 4. As value of game v= 4hence player A win the

game
 Principle of Dominance
The principle of dominance state that if a strategy of player dominates over another strategy
in all conditions ( ie. for all counter strategies by the other player ) then the later strategy (
being dominated ) can be ignored. A strategy dominates over the other strategy only if it is
preferable over the other under all conditions.
This principle can be applied to any ( m x n) size problem. In case of pure strategy
problem, it directly gives the solution and for a mixed strategy problem it can help reduce
the problem to ( 2 x 2) size which can be solved further using the algebraic method ( ie.
using Mixed strategy method)
Thus to solve a given ( m x n ) problem we can proceed as follows:
STEP 1: Check if there is saddle point. If a saddle point is there the optimal solution is
directly obtained.
STEP 2:If the problem does not have a saddle point then reduce it to ( 2 x 2) size by using
the following rules of dominance:
th
i)If the element in row ( say i row ) are less than or equal to ( ≤ ) the corresponding
th th
elements of any other row ( say j row) then neglect the i row.
th
ii)If the element in column ( say r row ) are greater than or equal to ( ≤ ) the
th th
corresponding elements of any other column ( say s column) then neglect the r
column.
iii)A pure strategy may be dominated if it is inferior to the combination of two other
strategies in a certain fixed proportion ( k: 1-k).
STEP 3: On getting the reduced pay off matrix of size (2 x 2 ) [after using the above rules
successively] we solve it further using the Algebric Method to get the optimal solution
Q. Solve the following Game.
PLAYER B
I P II III IV
PLAYER a 2 -2 4 1
A
b 6 -1 12 3
c -3 2 0 6
d 2 -3 7 1
 PLAYER B

I II III IV Row
Minima

PLAYER a 2 -2 4 1 -2
A
b 6 -1 12 3 -1<=
Maximin

c -3 2 0 6 -3

d 2 -3 7 1 -3

Column Maxima 6 2 12 6
**

As Minimax ≠ Maximin Hence it is mixed Strategy Game

 PLAYER B

I II III IV

A a 2 -2 4 1

b 6 -1 12 3

c -3 2 0 6

d 2 -3 7 1

● All the elements in the row a ≤ All the elements in the row b. Hence
neglect row a
● All the elements in the row d ≤ All the elements in the row b. Hence
neglect row d
 PLAYER B

I II III IV

A b 6 -1 12 3

c -3 2 0 6

● All the elements in the Column III≥ All the elements in the Column I .
Hence neglect Column III
● All the elements in the Column IV≥ All the elements in the Column II.
Hence neglect Column IV
 PLAYER B PLAYER B

I II I II

A b 6 -1 q 1-q

A b P 6 -1
c -3 2
c 1-p -3 2

p= ( a22 –a21 ) / ( a11 +a22 –a21 –a12 ) = (2+3)/ (6+2+3+1) = 5/12

1-p= 1-5/12 = 7/12

q= ( a22 –a21 ) / ( a11 +a22 –a21 –a12 ) = (2+1)/ (6+2+3+1) = 3/12

1-p= 1-3/12 = 9/12

v= (a11 a22 –a21 a12) / ( a11 +a22 –a21 –a12 ) = [6*2-(-3)(-1)]/ (6+2+3+1) = 9/12
p= 5/12 , 1-p= 7/12 , q = 3/12 1-p =9/12 v = 9/12

Stgy. Of Player A= (0,b,c,0)= (0,5/12,7/12,0)= (0%,14.67%, 58.33%,0%)

Stgy. Of Player B= (I,II,0,0)= (3/12,9/12,0,0)= (25%,75%, 0%,0%)

Value of the Game= V= 9/12 . As value of game v=9/12 hence player A

win the game
Q. Solve the following Game.
PLAYER B

B1 B2 B3 B4

PLAYER A A1 3 2 4 0

A2 3 4 2 4

A3 4 2 4 0

A4 0 4 0 8
 PLAYER B

B1 B2 B3 B4 Row
Minima

PLAYER A A1 3 2 4 0 0

A2 3 4 2 4 2<=
Maximin

A3 4 2 4 0 0

A4 0 4 0 8 0

Column Maxima 4 4 4 8

As Minimax Maximin ,
≠

Hence it is mixed Strategy Game

 PLAYER B

B1 B2 B3 B4

PLAYER A A1 3 2 4 0
A2 3 4 2 4
A3 4 2 4 0
A4 0 4 0 8

All the elements in the row A1 ≤ All the elements in the row A3. Hence neglect row A1
 PLAYER B

B1 B2 B3 B4

PLAYER A A2 3 4 2 4

A3 4 2 4 0

A4 0 4 0 8
All the elements in the Column B1 ≥ All the elements in the Column B3. Hence neglect Col B1

PLAYER B

B2 B3 B4

PLAYER A A2 4 2 4

A3 2 4 0

A4 4 0 8
 PLAYER B
B2 B3 B4
A2 4 2 4
PLAYER A A3 2 4 0
A4 4 0 8

B2 +B3 B4 B2 +B4 B3 B2 B3+B4

A2 3 4 A2 A2 4 3
4 2
A3 3 0 A3 A3 2 2
1 4
A4 2 8 A4 A4 4 4
2 0

www.sandipuniversity.edu.in
 Example
PLAYER B
B3 B4
A2 2 4
PLAYER A A3 4 0
A4 0 8

B3 B4 B3 B4 B3 B4
A2+A3 3 2 A2 +A4 A2 2 4
1 6
A4 0 8 A3 A3 +A4 2 4
4 0

www.sandipuniversity.edu.in
 Example
PLAYER B PLAYER B
B3 B4 B3 B4
A3 4 0 q 1-q
PLAYER A
A4 0 8 A3 P 4 0
PLAYER A
A4 1-p 0 8

p= ( a22 –a21 ) / ( a11 +a22 –a21 –a12 ) = (8-0)/ (4+8-0-0) = 8/12 = 2/3
1-p= 1-2/3 = 1/3
q= ( a22 –a21 ) / ( a11 +a22 –a21 –a12 ) = (8-0)/ (4+8-0-0) = 8/12 = 2/3
1-p= 1-2/3 = 1/3
v= (a11 a22 –a21 a12) / ( a11 +a22 –a21 –a12 ) = (4*8-0*0)/(4+8-0-0)= 32/12
p= 2/3 , 1-p= 1/3 , q = 2/3 1-p =1/3 v = 32/12

Stgy. Of Player A= (0,0,A3,A4)= (0,0,2/3,1/3)= (0%,0%, 66.67%,33.33%)

Stgy. Of Player B= (0,0,B3,B4)= (0,0,2/3,1/3)= (0%,0%, 66.67%,33.33%)

Value of the Game= V= 32/15 . As value of game v=32/15 hence player

A win the game

www.sandipuniversity.edu.in
Q. Solve the following Game. By Principle of Dominance Rule

PLAYER Y
1 2 3 4 5
I 1 3 2 7 4
II 3 4 1 5 6
PLAYER X
III 6 5 7 6 5
IV 2 0 6 3 1
 PLAYER Y
1 2 3 4 5 Row Minima

I 1 3 2 7 4 1
II 3 4 1 5 6 1
III 6 5 7 6 5 5<= Maximin
PLAYER X

As Minimax= Maximin Hence it is Pure Strategy Game

IV 2 0 6 3 1 0
6 5 <= Minimax 7 7 6
Column Maxima
As Minimax= Maximin Hence it is Pure Strategy Game

Stgy. Of Player X= (0,0,III,0)= (0,0,1,0)= (0%,0%, 100%,0%)

Stgy. Of Player Y= (0,2,0,0)= (0,1,0,0)= (0%,100%, 0%,0%)

Value of the Game= V= 5 . As value of game v=5 hence player X win the game
 Example 03
6
•
PLAYER Y
1 2 3 4 5
I 1 3 2 7 4
II 3 4 1 5 6
PLAYER X
III 6 5 7 6 5
IV 2 0 6 3 1
•
PLAYER Y
1 2 3 4 5
I 1 3 2 7 4
PLAYER X III 6 5 7 6 5
IV 2 0 6 3 1
•
PLAYER Y
1 2
I 1 3
PLAYER X III 6 5
IV 2 0

DEPARTMENT OF MANAGEMENT STUDIES (MBA ), SITRC

•

PLAYER Y
1 2
PLAYER X III 6 5

PLAYER Y
2
PLAYER X III 5