This is a version of prisoners dilemma played by each cell in a cellular automata.
The payoff described is a little different than most PD descriptions I've seen.
Here the mutual defection payoff is just as bad as the 'sucker' payoff - 0.
The 'temptation' payoff is 1.85, and the mutual cooperation payoff is 1.
Each cell plays an iteration of prisoner's dilema with each of its 8 neighbor cells.
The starting state of each cell (cooperate/defect) determines the result for each of these
pairwise games. The sum of payoffs for these 8 games is the cell's score for that round.
Each cell adopts the play of its highest scoring neighbor for its own next play.

--

DavidIlstrup? - 06 Jan 2007