File talk:Prisoner's Dilemma.jpg

This image does not represent the game as described. The game as described gives a more negative utility value to the Betray/Betray than Cooperate/Cooperate. (2,2)>(1,1) does not describe a more negative utility value for B/B vs. C/C. The positive numbers should be negative.

In the picture above, both players would want to unilaterally deviate from (2,2) to (0,3) or (3,0), and then to (1,1) whereby neither player would unilaterally deviate, making (1,1) or (Cooperate, Cooperate) the Pure Action Nash Equilibrium. However, if the numbers are negative, then the players will unilaterally deviate from -1,-1, to 0,-3 or -3,0 and finally to -2,-2 whereby they would not want to unilaterally deviate. This is my first suggested edit to Wikipedia and it won't let me update literally the picture we have already posted just with negative numbers.

Source: "A Course in Game Theory" Martin J. Osborne, Ariel Rubenstein 1994, MIT Press (pg 17)