Narrow Results

Cellular automata are used to study two-dimensional local hawk-dove games. Both Nash and Pareto optimal concepts are used. The systems have fixed point, cyclic and chaotic-like regimes. The results differ significantly from that of the normal one.

Prisoner's Dilemma games with two and three strategies are studied. The corresponding replicator equations, their steady states and their asymptotic stability are discussed. Local Prisoner's Dilemma games are studied using Pareto optimality. As in the case with Nash updating rule, the existence of tit for tat strategy is crucial to imply cooperation even in one dimension. Pareto updating implies less erratic behavior since the steady state configurations are mostly fixed points or at most 2-cycle. Finally, Prisoner's Dilemma game is simulated on small-world networks which are closer to real systems than regular lattices. There are no significant changes compared to the results of the regular lattice.