You are on page 1of 29

LECTURE 2

GAME THEORY Prof. Sanjay Christian


JGIBA (Assistant Professor)
M.Phil (Statistics), M.Sc (Statistics) & B.Sc (Statistics)
Mixed Strategies: Games without Saddle Point
When there is no saddle point, both players must obtain an optimal
mixture of strategies to find a saddle point.

Optimal strategy mixture for each player may be obtained, By assigning


probabilities to each strategy of being chosen.

The value of the game using mixed strategies represents least payoff which
player A can expect to win and least payoff which player B can loose.

The Expected payoff to a player in a game with payoff matrix 𝒂𝒊𝒋 of


Order 𝒎 × 𝒏 is given by,
𝑬 𝒑. 𝒒 = σ𝒎 σ 𝒏
𝒊=𝟏 𝒋=𝟏 𝒑𝒊 𝒂𝒊𝒋 𝒒𝒋 𝒑𝟏 , 𝒑𝟐 , … … 𝒑𝒎 mixed strategy for player A
𝒒𝟏 , 𝒒𝟐 , … … 𝒒𝒏 mixed strategy for player B
1
MIXED
ST R AT EG Y

2
GAME
CAN BE
S O LV E D
BY,

3
Dominance Property

01 02 03
If all the elements If 01 & 02 points
If all the elements
of a column (kth are not satisfied
of a row (kth row)
column) are more then take average
are less than or
than or equal to of two or more
equal to the
the elements of pure strategies
elements of the
the other column and compare with
other row (rth
(rth column), the remaining
row), then kth
then kth column pure strategies &
row is dominated
is dominated by delete if it is
by rth row.
rth column. inferior.

Dominance Property is used when the payoff matrix is a profit matrix for the
player A and a loss matrix to player B. Otherwise the principle gets revered.
ALGEBRAIC METHOD
Is used to determine probability value by using
different strategies by player A and B

Consider 2 × 2 two person zero sum game without


saddle point having the payoff matrix for player A
𝐵1 𝐵2
𝐴1 𝑎11 𝑎12
𝐴2 𝑎21 𝑎22

Optimum mixed strategy for Player A is 𝑨𝟏 , 𝑨𝟐 Optimum mixed strategy for Player B is 𝑩𝟏 , 𝑩𝟐
𝑺𝑨 = 𝒑𝟏 , 𝒑𝟐 𝑺𝑩 = 𝒒𝟏 , 𝒒𝟐
𝒂𝟐𝟐 − 𝒂𝟐𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐
𝒑𝟏 = 𝒑𝟐 = 𝟏 − 𝒑𝟏 𝒒𝟐 = 𝒒𝟏 = 𝟏 − 𝒒𝟐
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐 𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐

𝒂𝟏𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐 𝒂𝟐𝟏


𝑽𝒂𝒍𝒖𝒆 𝒐𝒇 𝒕𝒉𝒆 𝑮𝒂𝒎𝒆: 𝑽 =
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
Player B
𝐁𝟏 𝐁𝟐 𝐁𝟑 𝐁𝟒

3 Solve the following game. 𝐀𝟏


𝐀𝟐
5
9
1
4
7
15
4
6
Player A
Answer: 𝐀𝟑 0 5 3 9
𝑨𝟒 5 0 10 4
Player B Row
Minimum
𝐁𝟏 𝐁𝟐 𝐁𝟑 𝐁𝟒 Here,

𝐀𝟏 5 1 7 4 1 From Row Minimum, we find Maximum Value i.e.,


Maxi-Min=4
𝐀𝟐 9 4 15 6 4
Player A and
𝐀𝟑 0 5 3 9 0
From Column Maximum, we find Minimum Value
𝑨𝟒 5 0 10 4 0 i.e., Mini-Max=5

Column 9 5 15 9
Maximum 𝑴𝒂𝒙𝒊. 𝒎𝒊𝒏 ≠ 𝑴𝒊𝒏𝒊. 𝒎𝒂𝒙
Saddle point does not exist.
As Saddle Point does not exist, the player will use Mixed Strategies.
5 1 7 4
9 4 15 6 Step : 3 Elements of 3rd Row is less than 1st Row and
hence 3rd Row is dominated by 1st Row.
0 5 3 9
5 0 10 4 9 4 6
0 5 9
Now we reduce the size of the matrix
by using Dominance Property. Step : 4 Elements of 3rd Column is more than 2nd
Column and hence 3rd Column is dominated by
Step : 1 Elements of 1st Row is less than 2nd Row and 2nd Column
hence 1st Row is dominated by 2nd Row.
9 4
9 4 15 6
0 5
0 5 3 9
5 0 10 4
Step : 2 Elements of 3rd Column is more than 1st 𝒂𝟏𝟏 𝒂𝟏𝟐
Column and hence 3rd Column is dominated by 1st 𝒂𝟐𝟏 𝒂𝟐𝟐
Column
9 4 6
0 5 9
5 0 4
5 1 7 4 𝒂𝟏𝟏 = 𝟗 𝒂𝟏𝟐 = 𝟒
𝒑𝟏 + 𝒑𝟐 + 𝒑𝟑 + 𝒑𝟒 = 𝟏 9 4 15 6 𝒒𝟏 + 𝒒𝟐 + 𝒒𝟑 + 𝒒𝟒 = 𝟏
0 5 3 9 𝒂𝟐𝟏 = 𝟎 𝒂𝟐𝟐 = 𝟓
𝒑𝟏 = 𝟎 & 𝒑𝟒 = 𝟎 5 0 10 4 𝒒𝟑 = 𝟎 & 𝒒𝟒 = 𝟎
𝒂𝟐𝟐 − 𝒂𝟐𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐
𝒑𝟐 = 𝒒𝟏 =
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐 𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
𝟓−𝟎 𝟓−𝟒
= =
𝟗+𝟓 − 𝟎+𝟒 𝟗+𝟓 − 𝟎+𝟒
𝟓 𝟏
= =
𝟏𝟒 − 𝟒 𝟏𝟒 − 𝟒
𝟓 𝟏
= 𝒒𝟏 =
𝟏𝟎 𝟏𝟎
𝟏 𝟏 𝟗
𝒑𝟐 = 𝒒𝟐 = 𝟏 − 𝒒𝟏 − 𝒒𝟑 − 𝒒𝟒 = 𝟏 − −𝟎−𝟎=
𝟐 𝟏𝟎 𝟏𝟎
𝟏 𝟏 Optimum mixed strategy for Player B is 𝑩𝟏 , 𝑩𝟐 , 𝑩𝟑 , 𝑩𝟒
𝒑𝟑 = 𝟏 − 𝒑𝟏 − 𝒑𝟐 − 𝒑𝟒 = 𝟏 − 𝟎 − − 𝟎 =
𝟐 𝟐
𝟏 𝟗
Optimum mixed strategy for Player A is 𝑨𝟏 , 𝑨𝟐 , 𝑨𝟑 , 𝑨𝟒 𝑺𝑩 = 𝒒𝟏 , 𝒒𝟐 , 𝒒𝟑 , 𝒒𝟒 = , , 𝟎, 𝟎
𝟏 𝟏 𝟏𝟎 𝟏𝟎
𝑺𝑨 = 𝒑𝟏 , 𝒑𝟐 , 𝒑𝟑 , 𝒑𝟒 = 𝟎, , , 𝟎
𝟐 𝟐
𝒂𝟏𝟏 = 𝟗 𝒂𝟏𝟐 = 𝟒
𝒂𝟐𝟏 = 𝟎 𝒂𝟐𝟐 = 𝟓

𝒂𝟏𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐 𝒂𝟐𝟏


𝑽=
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
𝟗 𝟓 − 𝟎 𝟒
=
𝟗+𝟓 − 𝟎+𝟒
𝟒𝟓 − 𝟎
=
𝟏𝟒 − 𝟒
𝟒𝟓
=
𝟏𝟎
= 𝟒. 𝟓
Two companies are competing for business under the conditions so that one
company’s gain is another company’s loss. The following payoff in lakhs is derived by

4 company A for different strategies used for advertisements.

Company A
Company B
No advertising Medium advertising Heavy advertising
No advertising 5 1 -7
Medium advertising 8 7 8
Heavy advertising 11 9 5
Suggest optimal strategies for the two companies.
Answer: Here,
Player B Row From Row Minimum we find Maximum Value
Minimum i.e., Maxi-Min=7
𝐁𝟏 𝐁𝟐 𝐁𝟑
-7 and
𝐀𝟏 5 1 -7
From Column Maximum we find Minimum Value
Player A 𝐀𝟐 8 7 8 7
i.e., Mini-Max=8
𝐀𝟑 11 9 5 5
11 9 8 𝑴𝒂𝒙𝒊. 𝒎𝒊𝒏 ≠ 𝑴𝒊𝒏𝒊. 𝒎𝒂𝒙
Column
Saddle point does not exist,
Maximum
As Saddle Point does not exist, the player will use Mixed Strategies.
5 1 -7
8 7 8
11 9 5

Now we reduce the size of the matrix


by using Dominance Property.

Step : 1 Elements of 1st Row is less than 2nd Row and Step : 2 Elements of 1st Column is more than 2nd
hence 1st Row is dominated by 2nd Row. Column and hence 1st Column is dominated by 2nd
Column
8 7 8
7 8
11 9 5
9 5

𝒂𝟏𝟏 𝒂𝟏𝟐
𝒂𝟐𝟏 𝒂𝟐𝟐
5 1 -7 𝒂𝟏𝟏 = 𝟕 𝒂𝟏𝟐 = 𝟖
𝒑𝟏 + 𝒑𝟐 + 𝒑𝟑 = 𝟏 𝒒𝟏 + 𝒒𝟐 + 𝒒𝟑 = 𝟏
8 7 8
𝒂𝟐𝟏 = 𝟗 𝒂𝟐𝟐 = 𝟓
𝒑𝟏 = 𝟎 11 9 5 𝒒𝟏 = 𝟎
𝒂𝟐𝟐 − 𝒂𝟐𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐
𝒑𝟐 = 𝒒𝟐 =
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐 𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
𝟓−𝟗 𝟓−𝟖
= =
𝟕 + 𝟓 − 𝟗 + 𝟖) 𝟕+𝟓 − 𝟗+𝟖
−𝟒 −𝟑
= =
𝟏𝟐 − 𝟏𝟕 𝟏𝟐 − 𝟏𝟕
−𝟒 −𝟑
= =
−𝟓 −𝟓
𝟒 𝟒 𝟏 𝟑 𝟑 𝟐
𝒑𝟐 = 𝒑𝟑 = 𝟏 − 𝒑𝟏 − 𝒑𝟐 = 𝟏 − 𝟎 − = 𝒒𝟐 = 𝒒𝟑 = 𝟏 − 𝒒𝟏 − 𝒒𝟐 = 𝟏 − 𝟎 − =
𝟓 𝟓 𝟓 𝟓 𝟓 𝟓
Optimum mixed strategy for Player B is 𝑩𝟏 , 𝑩𝟐 , 𝑩𝟑
Optimum mixed strategy for Player A is 𝑨𝟏 , 𝑨𝟐 , 𝑨𝟑
𝟒 𝟏 𝟑 𝟐
𝑺𝑨 = 𝒑𝟏 , 𝒑𝟐 , 𝒑𝟑 = 𝟎, , 𝑺𝑩 = 𝒒𝟏 , 𝒒𝟐 , 𝒒𝟑 = 𝟎, ,
𝟓 𝟓 𝟓 𝟓
𝒂𝟏𝟏 = 𝟕 𝒂𝟏𝟐 = 𝟖

𝒂𝟐𝟏 = 𝟗 𝒂𝟐𝟐 = 𝟓

𝒂𝟏𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐 𝒂𝟐𝟏


𝑽=
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
𝟕 𝟓 − 𝟗 𝟖
=
𝟕+𝟓 − 𝟗+𝟖
𝟑𝟓 − 𝟕𝟐
=
𝟏𝟐 − 𝟏𝟕
−𝟑𝟕
=
−𝟓
= 𝟕. 𝟒
Solve the following Game.

5 𝟗
𝟑
𝟖
−𝟔
−𝟕
𝟒
𝟔 𝟕 −𝟕

Answer: Here,
Player B Row From Row Minimum we find Maximum Value
Minimum i.e., Maxi-Min=-6
𝐁𝟏 𝐁𝟐 𝐁𝟑
-7 and
𝐀𝟏 9 8 -7
From Column Maximum we find Minimum Value
Player A 𝐀𝟐 3 -6 4 -6
i.e., Mini-Max=4
𝐀𝟑 6 7 -7 -7

Column 9 8 4 𝑴𝒂𝒙𝒊. 𝒎𝒊𝒏 ≠ 𝑴𝒊𝒏𝒊. 𝒎𝒂𝒙


Maximum Saddle point does not exist,
As Saddle Point does not exist, the player will use Mixed Strategies.
9 8 -7
3 -6 4
6 7 -7

Now we reduce the size of the matrix


by using Dominance Property.

Step : 1 Elements of 3rd Row is less than 1st Row and Step : 2 Elements of 1st Column is more than 2nd
hence 3rd Row is dominated by 1st Row. Column and hence 1st Column is dominated by 2nd
Column
9 8 -7
8 -7
3 -6 4
-6 4

𝒂𝟏𝟏 𝒂𝟏𝟐
𝒂𝟐𝟏 𝒂𝟐𝟐
9 8 -7 𝒂𝟏𝟏 = 𝟖 𝒂𝟏𝟐 = −𝟕
𝒑𝟏 + 𝒑𝟐 + 𝒑𝟑 = 𝟏 𝒒𝟏 + 𝒒𝟐 + 𝒒𝟑 = 𝟏
3 -6 4
𝒂𝟐𝟏 = −𝟔 𝒂𝟐𝟐 = 𝟒
𝒑𝟑 = 𝟎 6 7 -7 𝒒𝟏 = 𝟎
𝒂𝟐𝟐 − 𝒂𝟐𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐
𝒑𝟏 = 𝒒𝟐 =
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐 𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
𝟒 − (−𝟔) 𝟒 − (−𝟕)
= =
𝟖 + 𝟒 − −𝟔 + (−𝟕) 𝟖 + 𝟒 − −𝟔 + (−𝟕)
𝟏𝟎 𝟏𝟏
= =
𝟏𝟐 − −𝟏𝟑 𝟏𝟐 − −𝟏𝟑
𝟏𝟎 𝟏𝟏
= =
𝟐𝟓 𝟐𝟓
𝟐 𝟐 𝟑 𝟏𝟏 𝟏𝟏 𝟏𝟒
𝒑𝟏 = 𝒑𝟐 = 𝟏 − 𝒑𝟏 − 𝒑𝟑 = 𝟏 − − 𝟎 = 𝒒𝟐 = 𝒒𝟑 = 𝟏 − 𝒒𝟏 − 𝒒𝟐 = 𝟏 − 𝟎 − =
𝟓 𝟓 𝟓 𝟐𝟓 𝟐𝟓 𝟐𝟓
Optimum mixed strategy for Player A is 𝑨𝟏 , 𝑨𝟐 , 𝑨𝟑 Optimum mixed strategy for Player B is 𝑩𝟏 , 𝑩𝟐 , 𝑩𝟑
𝟐 𝟑 𝟏𝟏 𝟏𝟒
𝑺𝑨 = 𝒑𝟏 , 𝒑𝟐 , 𝒑𝟑 = , ,𝟎 𝑺𝑩 = 𝒒𝟏 , 𝒒𝟐 , 𝒒𝟑 = 𝟎, ,
𝟓 𝟓 𝟐𝟓 𝟐𝟓
𝒂𝟏𝟏 = 𝟖 𝒂𝟏𝟐 = −𝟕

𝒂𝟐𝟏 = −𝟔 𝒂𝟐𝟐 = 𝟒

𝒂𝟏𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐 𝒂𝟐𝟏


𝑽=
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
𝟖 𝟒 − −𝟔 −𝟕
=
𝟖 + 𝟒 − −𝟔 + (−𝟕)
𝟑𝟐 − 𝟒𝟐
=
𝟏𝟐 − −𝟏𝟑
−𝟏𝟎
=
𝟐𝟓
−𝟐
=
𝟓
Solve the following Game. (Extra Example for Practice).

6 𝟏
𝟔
𝟕 𝟐
𝟐 𝟕
𝟔 𝟏 𝟔

Answer: Here,
Player B Row From Row Minimum we find Maximum Value
Minimum i.e., Maxi-Min=2
𝐁𝟏 𝐁𝟐 𝐁𝟑
1 and
𝐀𝟏 1 7 2
From Column Maximum we find Minimum Value
Player A 𝐀𝟐 6 2 7 2
i.e., Mini-Max=6
𝐀𝟑 6 1 6 1

Column 6 7 7 𝑴𝒂𝒙𝒊. 𝒎𝒊𝒏 ≠ 𝑴𝒊𝒏𝒊. 𝒎𝒂𝒙


Maximum Saddle point does not exist,
As Saddle Point does not exist, the player will use Mixed Strategies.
1 7 2
6 2 7
6 1 6

Now we reduce the size of the matrix


by using Dominance Property.

Step : 1 Elements of 3rd Row is less than 2nd Row and Step : 2 Elements of 3rd Column is more than 1st
hence 3rd Row is dominated by 2nd Row. Column and hence 3rd Column is dominated by 1st
Column
1 7 2
1 7
6 2 7
6 2

𝒂𝟏𝟏 𝒂𝟏𝟐
𝒂𝟐𝟏 𝒂𝟐𝟐
1 7 2 𝒂𝟏𝟏 = 𝟏 𝒂𝟏𝟐 = 𝟕
𝒑𝟏 + 𝒑𝟐 + 𝒑𝟑 = 𝟏 𝒒𝟏 + 𝒒𝟐 + 𝒒𝟑 = 𝟏
6 2 7
𝒂𝟐𝟏 = 𝟔 𝒂𝟐𝟐 = 𝟐
𝒑𝟑 = 𝟎 6 1 6 𝒒𝟑 = 𝟎
𝒂𝟐𝟐 − 𝒂𝟐𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐
𝒑𝟏 = 𝒒𝟏 =
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐 𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
𝟐−𝟔 𝟐−𝟕
= =
𝟏+𝟐 − 𝟔+𝟕 𝟏+𝟐 − 𝟔+𝟕
−𝟒 −𝟓
= =
𝟑 − 𝟏𝟑 𝟑 − 𝟏𝟑
−𝟒 −𝟓
= =
−𝟏𝟎 −𝟏𝟎
𝟐 𝟐 𝟑 𝟏 𝟏 𝟏
𝒑𝟏 = 𝒑𝟐 = 𝟏 − 𝒑 𝟏 − 𝒑𝟑 = 𝟏 − − 𝟎 = 𝒒𝟏 = 𝒒𝟐 = 𝟏 − 𝒒𝟏 − 𝒒𝟑 = 𝟏 − − 𝟎 =
𝟓 𝟓 𝟓 𝟐 𝟐 𝟐
Optimum mixed strategy for Player A is 𝑨𝟏 , 𝑨𝟐 , 𝑨𝟑 Optimum mixed strategy for Player B is 𝑩𝟏 , 𝑩𝟐 , 𝑩𝟑
𝟐 𝟑 𝟏 𝟏
𝑺𝑨 = 𝒑𝟏 , 𝒑𝟐 , 𝒑𝟑 = , ,𝟎 𝑺𝑩 = 𝒒𝟏 , 𝒒𝟐 , 𝒒𝟑 = , ,𝟎
𝟓 𝟓 𝟐 𝟐
𝒂𝟏𝟏 = 𝟏 𝒂𝟏𝟐 = 𝟕
𝒂𝟐𝟏 = 𝟔 𝒂𝟐𝟐 = 𝟐

𝒂𝟏𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐 𝒂𝟐𝟏


𝑽=
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
𝟏 𝟐 − 𝟔 𝟕
=
𝟏+𝟐 − 𝟔+𝟕
𝟐 − 𝟒𝟐
=
𝟑 − 𝟏𝟑
−𝟒𝟎
=
−𝟏𝟎
=𝟒
Player B
𝐁𝟏 𝐁𝟐 𝐁𝟑 𝐁𝟒

7 Solve the following game.


(Extra Example for Practice)
𝐀𝟏
𝐀𝟐
1
1
0
2
2
0
-2
2
Player A
Answer: 𝐀𝟑 2 0 2 -2
𝑨𝟒 -2 2 -2 1
Player B Row
Minimum
𝐁𝟏 𝐁𝟐 𝐁𝟑 𝐁𝟒 Here,

𝐀𝟏 1 0 2 -2 -2 From Row Minimum, we find Maximum Value i.e.,


Maxi-Min=0
𝐀𝟐 1 2 0 2 0
Player A and
𝐀𝟑 2 0 2 -2 -2
From Column Maximum, we find Minimum Value
𝑨𝟒 -2 2 -2 1 -2 i.e., Mini-Max=2

Column 2 2 2 2
Maximum 𝑴𝒂𝒙𝒊. 𝒎𝒊𝒏 ≠ 𝑴𝒊𝒏𝒊. 𝒎𝒂𝒙
Saddle point does not exist.
As Saddle Point does not exist, the player will use Mixed Strategies.
1 0 2 -2
1 2 0 2 Step : 3 Elements of 3rd Row is less than 1st Row and
hence 3rd Row is dominated by 1st Row.
2 0 2 -2
-2 2 -2 1 2 0 2
0 2 -2
Now we reduce the size of the matrix
by using Dominance Property. Step : 4 Elements of 1st Column is more than 3rd
Column and hence 1st Column is dominated by 3rd
Step : 1 Elements of 1st Row is less than 3rd Row and Column
hence 1st Row is dominated by 3rd Row.
0 2
1 2 0 2
2 0 2 -2 2 -2
-2 2 -2 1

Step : 2 Elements of 1st Column is more than 3rd 𝒂𝟏𝟏 𝒂𝟏𝟐


Column and hence 1st Column is dominated by 3rd 𝒂𝟐𝟏 𝒂𝟐𝟐
Column
2 0 2
0 2 -2
2 -2 1
1 0 2 -2 𝒂𝟏𝟏 = 𝟎 𝒂𝟏𝟐 = 𝟐
𝒑𝟏 + 𝒑𝟐 + 𝒑𝟑 + 𝒑𝟒 = 𝟏 1 2 0 2 𝒒𝟏 + 𝒒𝟐 + 𝒒𝟑 + 𝒒𝟒 = 𝟏
2 0 2 -2 𝒂𝟐𝟏 = 𝟐 𝒂𝟐𝟐 = −𝟐
𝒑𝟏 = 𝟎 & 𝒑𝟒 = 𝟎 -2 2 -2 1 𝒒𝟏 = 𝟎 & 𝒒𝟐 = 𝟎
𝒂𝟐𝟐 − 𝒂𝟐𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐
𝒑𝟐 = 𝒒𝟑 =
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐 𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
−𝟐 − 𝟐 −𝟐 − 𝟐
= =
𝟎 + (−𝟐) − 𝟐 + 𝟐 𝟎 + (−𝟐) − 𝟐 + 𝟐
−𝟒 −𝟒
= =
−𝟐 − 𝟒 −𝟐 − 𝟒
−𝟒 −𝟒
= =
−𝟔 −𝟔
𝟐 𝟐
𝒑𝟐 = 𝒒𝟑 =
𝟑 𝟑
𝟐 𝟏 𝟐 𝟏
𝒑𝟑 = 𝟏 − 𝒑𝟏 − 𝒑𝟐 − 𝒑𝟒 = 𝟏 − 𝟎 − − 𝟎 = 𝒒𝟒 = 𝟏 − 𝒒𝟏 − 𝒒𝟐 − 𝒒𝟑 = 𝟏 − 𝟎 − 𝟎 − =
𝟑 𝟑 𝟑 𝟑
Optimum mixed strategy for Player A is 𝑨𝟏 , 𝑨𝟐 , 𝑨𝟑 , 𝑨𝟒 Optimum mixed strategy for Player B is 𝑩𝟏 , 𝑩𝟐 , 𝑩𝟑 , 𝑩𝟒
𝟐 𝟏 𝟐 𝟏
𝑺𝑨 = 𝒑𝟏 , 𝒑𝟐 , 𝒑𝟑 , 𝒑𝟒 = 𝟎, , , 𝟎 𝑺𝑩 = 𝒒𝟏 , 𝒒𝟐 , 𝒒𝟑 , 𝒒𝟒 = 𝟎, 𝟎, ,
𝟑 𝟑 𝟑 𝟑
𝒂𝟏𝟏 = 𝟎 𝒂𝟏𝟐 = 𝟐
𝒂𝟐𝟏 = 𝟐 𝒂𝟐𝟐 = −𝟐

𝒂𝟏𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐 𝒂𝟐𝟏


𝑽=
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
𝟎 −𝟐 − 𝟐 𝟐
=
𝟎 + (−𝟐) − 𝟐 + 𝟐
𝟎−𝟒
=
−𝟐 − 𝟒
−𝟒
=
−𝟔
𝟐
=
𝟑
Player B
𝐁𝟏 𝐁𝟐 𝐁𝟑 𝐁𝟒

8 Solve the following game.


(Extra Example for Practice)
𝐀𝟏
𝐀𝟐
6
0
8
3
-16
-4
-10
2
Player A
Answer: 𝐀𝟑 18 4 10 51
𝑨𝟒 10 2 5 16
Player B Row
Minimum
𝐁𝟏 𝐁𝟐 𝐁𝟑 𝐁𝟒 Here,

𝐀𝟏 6 8 -16 -10 -16 From Row Minimum, we find Maximum Value i.e.,
Maxi-Min=4
𝐀𝟐 0 3 -4 2 -4
Player A and
𝐀𝟑 18 4 10 51 4
From Column Maximum, we find Minimum Value
𝑨𝟒 10 2 5 16 2 i.e., Mini-Max=8

Column 18 8 10 51
Maximum 𝑴𝒂𝒙𝒊. 𝒎𝒊𝒏 ≠ 𝑴𝒊𝒏𝒊. 𝒎𝒂𝒙
Saddle point does not exist.
As Saddle Point does not exist, the player will use Mixed Strategies.
6 8 -16 -10
0 3 -4 2 Step : 3 Elements of 3rd Row is less than 2nd Row
and hence 3rd Row is dominated by 2nd Row.
18 4 10 51
10 2 5 16 8 -16 -10
4 10 51
Now we reduce the size of the matrix
by using Dominance Property. Step : 4 Elements of 3rd Column is more than 2nd
Column and hence 3rd Column is dominated by
Step : 1 Elements of 2nd Row is less than 3rd Row and 2nd Column
hence 2nd Row is dominated by 3rd Row.
8 -16
6 8 -16 -10
4 10
18 4 10 51
10 2 5 16
Step : 2 Elements of 1st Column is more than 3rd 𝒂𝟏𝟏 𝒂𝟏𝟐
Column and hence 1st Column is dominated by 3rd 𝒂𝟐𝟏 𝒂𝟐𝟐
Column
8 -16 -10
4 10 51
2 5 16
6 8 -16 -10 𝒂𝟏𝟏 = 𝟖 𝒂𝟏𝟐 = −𝟏𝟔
𝒑𝟏 + 𝒑𝟐 + 𝒑𝟑 + 𝒑𝟒 = 𝟏 0 3 -4 2 𝒒𝟏 + 𝒒𝟐 + 𝒒𝟑 + 𝒒𝟒 = 𝟏
18 4 10 51 𝒂𝟐𝟏 = 𝟒 𝒂𝟐𝟐 = 𝟏𝟎
𝒑𝟐 = 𝟎 & 𝒑𝟒 = 𝟎 10 2 5 16 𝒒𝟏 = 𝟎 & 𝒒𝟒 = 𝟎
𝒂𝟐𝟐 − 𝒂𝟐𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐
𝒑𝟏 = 𝒒𝟐 =
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐 𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
𝟏𝟎 − 𝟒
= 𝟏𝟎 − (−𝟏𝟔)
𝟖 + 𝟏𝟎 − 𝟒 + (−𝟏𝟔) =
𝟖 + 𝟏𝟎 − 𝟒 + (−𝟏𝟔)
𝟔 𝟐𝟔
= =
𝟏𝟖 − −𝟏𝟐 𝟏𝟖 − −𝟏𝟐
𝟔 𝟐𝟔
= =
𝟑𝟎 𝟑𝟎
𝟏 𝟏𝟑
𝒑𝟏 = 𝒒𝟐 =
𝟓 𝟏𝟓
𝟏 𝟒 𝟏𝟑 𝟐
𝒑𝟑 = 𝟏 − 𝒑𝟏 − 𝒑𝟐 − 𝒑𝟒 = 𝟏 − − 𝟎 − 𝟎 = 𝒒𝟑 = 𝟏 − 𝒒𝟏 − 𝒒𝟐 − 𝒒𝟒 = 𝟏 − 𝟎 − −𝟎=
𝟓 𝟓 𝟏𝟓 𝟏𝟓
Optimum mixed strategy for Player A is 𝑨𝟏 , 𝑨𝟐 , 𝑨𝟑 , 𝑨𝟒 Optimum mixed strategy for Player B is 𝑩𝟏 , 𝑩𝟐 , 𝑩𝟑 , 𝑩𝟒
𝟏 𝟒 𝟏𝟑 𝟐
𝑺𝑨 = 𝒑𝟏 , 𝒑𝟐 , 𝒑𝟑 , 𝒑𝟒 = , , 𝟎, 𝟎 𝑺𝑩 = 𝒒𝟏 , 𝒒𝟐 , 𝒒𝟑 , 𝒒𝟒 = 𝟎, , ,𝟎
𝟓 𝟓 𝟏𝟓 𝟏𝟓
𝒂𝟏𝟏 = 𝟖 𝒂𝟏𝟐 = −𝟏𝟔
𝒂𝟐𝟏 = 𝟒 𝒂𝟐𝟐 = 𝟏𝟎

𝒂𝟏𝟏 𝒂𝟐𝟐 − 𝒂𝟏𝟐 𝒂𝟐𝟏


𝑽=
𝒂𝟏𝟏 + 𝒂𝟐𝟐 − 𝒂𝟐𝟏 + 𝒂𝟏𝟐
𝟖 𝟏𝟎 − 𝟒 −𝟏𝟔
=
𝟖 + 𝟏𝟎 − 𝟒 + (−𝟏𝟔)
𝟖𝟎 + 𝟔𝟒
=
𝟏𝟖 − −𝟏𝟐
𝟏𝟒𝟒
=
𝟑𝟎
𝟐𝟒
=
𝟓
= 𝟒. 𝟖

You might also like