Economic and Game Theory
Forum

20 5 10 20 50 100 all
Ask a question about: Other Topics
Respond to the question: Nash Equilibrium?

10/14/2017 08:47 AM by Elene Margalitadze; Nash Equilibrium

It is written that computing pure Nash equilibria in congestion games remains PLS-complete also when we restrict to affine delay functions. That is for dr(x) = arx + brx with, ar, br >= 0. Why is that? [View full text and thread]

11/30/2004 02:58 PM by name withheld; Answer to Miguel, Nash Equilibrium

Sorry, but (Call, Not Call) and (Not Call, Call) are both NE in pure strategies, it is a special case of an anticoordination game. [View full text and thread]

11/01/2004 04:59 PM by Miguel; Nash Equilibrium

I have a problem and an incomplete answer to it. If there's anyone who can find a solution to it, please post it on the website. Here it is

“A boy was hit by a car and is lying on the road, hurt. The driver ran away and 5 people gathered around the boy. The boy needs immediate medical care but for that to happen it is necessary that someone calls 911. Simultaneously and independently, each of the individuals decides if he/she will or not call for help.
If no doctor comes to rescue the boy, each of the individuals will feel guilt which will represent to him/her a utility of -15. Someone who asks for help will pay a cost that represents a utility of -5.

a) This game has a Nash equilibrium with pure-strategies? In case of an affirmative answer, show it/them.
b) This game has any Nash equilibrium with mixed-strategies in which every person has the same probability of calling 911? In case of an affirmative answer, show it. (Suggestion: define p as the probability of someone NOT calling 911).”



Concerning the first question, I find it very easy to answer IF there were only 2 bystanders and not 5. I’ll present the normal-form representation of the game for 2 players:
Players: P=(1;2)
Strategy Spaces:
S1= (Call; NotCall)
S2= (Call; NotCall)
Payoff Function of Player 1:
U1 (Call,Call) = -5
U1 (Call, NotCall) = -5
U1 (NotCall, Call) = 0
U1 (NotCall, NotCall) = -15

Payoff Function of Player 2:
U2 (Call,Call) = -5
U2 (Call, NotCall) = 0
U2 (NotCall, Call) = -5
U2 (NotCall, NotCall) = -15

Please note that I considered the possibility of both players call 911 at the same time, therefore both pay the cost. I don’t know if that’s correct.


Is there a Nash equilibrium with pure strategies? I believe there isn’t (in this 2-players game). This how I reached this conclusion:
If Player 1 decides to Call, Player 2 decides to NotCall.
If Player 1 decides to NotCall, Player 2 decides to Call.
Therefore, there isn’t any best strategy from Player 2 to Player 1’s strategies. Which means that there isn’t a Nash equilibrium with pure strategies.

Of course, since I assumed there were only 2 players, the outcome of the game with 5 players might be completely opposite to what I believe.



Now the second question gets trickier. I’ve already tried to solve it by using the initial assumption (that there are only 2 players) but it doesn’t seem to work out:

(1-p, p) is the mixed strategy in which Player 1 chooses to Call with the probability of 1-p and (1-r, r) is the mixed strategy in which Player 2 chooses to Call with the probability of 1-r.
If Player 1 plays (1-p, p) then Player 2’s expected payoffs are (1-p)(-5) + p(-5) = -5

Now the problem is that the result should be a function of p. Shouldn’t it be?

[Manage messages]