Prison cells

Cellular Automata (CAs) represent a powerful and versatile platform for simulating complex systems. Here we focus on an abstraction of a social system in which pairs of neighbouring agents interact according to a famous example of game theoretical dynamic: the so-called "Prisoner's Dilemma".

There are two possible strategies, cooperation (C) and defection (D), and no memory. The players change their strategy each turn to emulate the neighbour(s) achieving the greatest profit in the previous iteration; the full payoff is the sum of the rewards of each one-to-one game.

For simplicity, the choice of values is the following

$$ S = P = 0 \quad R = 1 \quad T = b $$

so there is just one parameter. This does not prevent the model from exhibiting the full complex dynamics, with chaotic behaviour and fractal patterns.

Output example (250x250)