Econ 402, Homework 1 Solutions

Kaustav Das

January,24th, 2011

1. Find the maximin strategies and the value of the following ZSG.

L R
U 2 1
D 5 −2
Strategies: U , R . Value= 1.

2. Find the maximin strategies and the value of the following ZSG.

L R
U −2 3
D 5 0
Let p denote the probability with which Player 1 (row player) chooses U, and let
q denote the probability with which Player 2 (column player) chooses L. Then the
expected payoff to Player 1 is

π(p, q) = −2pq + 3p(1 − q) + 5(1 − p)q

= 3p + (5 − 10p)q.

1
Then
3
if p < 12 ;

 3p


< 2

3
min π(p, q) = 3p = 2
if p = 12 ; and
q 

3
if p > 12 ,


5 − 7p < 2

since if p < 21 , π(p, ·) is minimized at q = 0, and if p > 12 then π(p, ·) is minimized at

q = 1. Thus Player 1’s maximin strategy is p∗ = 21 . Since π(p∗ , q) = 32 for all q, the
value of the game is 32 .
Player 2’s expected payoff is

−π(p, q) = (10q − 3)p − 5q.

Therefore
− 3 < − 32 3


 5q if q < 10
;


min
p
−π(p, q) =  −5q = − 32 if q = 3
10
; and

−5q < − 32 3


if q > 10
,
so q∗ = 3
10
is Player 2’s maximin strategy.

3. The the possible outcomes of the game together with their respective proba-
bilities and Player 1 payoffs are as follows: We use the fact that if the card is High,
Player 1 Bets with probability 1.
1
• Outcome: High,Bet,Call. Probability: 2
× 1 × q. Player 1 payoff: 6.
1
• Outcome: High,Bet,Fold. Probability: 2
× 1 × (1 − q). Player 1 payoff:
1.
1
• Outcome: High,Fold. Probability: 2
× 0. Player 1 payoff: −1.
1
• Outcome: Low,Bet,Call. Probability: 2
× p × q. Player 1 payoff: −6.
1
• Outcome: Low,Bet,Fold. Probability: 2
× p × (1 − q). Player 1 payoff:
1.
1
• Outcome: Low,Fold. Probability: 2
× (1 − p). Player 1 payoff: −1.

2
Player 1’s expected payoff, π(p, q), is the sum of the payoffs multiplied by their
respective probabilities. After collecting terms, this yields

1
π(p, q) = [(5 − 7p)q + 2p].
2

If 5 − 7p < 0, i.e., p > 57 , then π(p, ·) is minimized at q = 1. If 5 − 7p > 0, i.e., p < 57 ,

then π(p, ·) is minimized at q = 0. If 5 − 7p = 0, i.e., p = 57 , then π(p, q) = 12 × 2p = 57
for all values of q. Hence
1 5
(1 − p) < if p > 75 ;
2 7



1 5
min π(p, q) = 2
× 2p = 7
if p = 75 ; and
q 

1 5
if p < 75 .

2
× 2p < 7

Thus Player 1’s maxmin strategy is p∗ = 57 . Since

∗ 5
min π(p , q) = ,
q 7

the value of the game is 57 .

To derive Player 2’s maxmin strategy, first observe that

1
−π(p, q) = [(7q − 2)p − 5q].
2

If 7q − 2 < 0, i.e., q < 27 , then −π(·, q) is minimized at p = 1. If 7q − 2 > 0,

i.e., q > 27 , then −π(·, q) is minimized at p = 0. If 7q − 2 = 0, i.e., q = 27 , then
−π(p, q) = 12 × −5q = − 75 for all values of q. Hence
1
(2q − 2) < − 57 if q < 27 ;
2



1
min π(p, q) = 2
× −5q = − 75 if q = 27 ; and
p 

1
× −5q < − 75 if q > 27 .

2

Thus Player 2’s maxmin strategy is q∗ = 27 .