When strategies spread not by rationality, but by who reproduces—the biology of games.
Evolutionary game theory applies game theory to populations where strategies spread by reproduction and selection rather than by deliberate rational choice. Pioneered by John Maynard Smith, it explains how behaviors like cooperation, altruism, and aggression can persist. It explains why certain strategies survive in a population, even when individual players are not consciously optimizing.
Definition: An evolutionarily stable strategy (ESS) is one that, once established in a population, cannot be invaded by a rare mutant playing a different strategy—because the incumbent performs at least as well against the mutant as the mutant performs against it.
No rationality required: Strategies spread because they are successful, not because players calculate. A successful behavior reproduces and displaces weaker ones.
Hawks and doves: A classic model where aggressive hawks and pacifist doves interact. The stable outcome is often a mix, balancing the gains of aggression against its costs.
The puzzle: Why would altruism and cooperation evolve when selfish behavior seems more fit? Evolutionary game theory shows circumstances under which cooperation thrives.
Reciprocity: In repeated interactions, cooperative strategies like Tit-for-Tat outperform and spread because they mimic the behavior of partners (see Repeated Games).
Kin selection: Helping relatives passes on shared genes, so apparent selflessness can be individually adaptive.
Group effects: Cooperative groups may outcompete selfish ones, sustaining norms even when individuals would prefer to free-ride.
Biology: Explains animal behavior—fighting, mating, and territoriality—as the outcome of population dynamics rather than conscious strategy.
Human norms and culture: Cooperation, fairness, and punishment can be modeled as evolving behaviors, shedding light on why these instincts are widespread.
Economics and markets: Behaviors that persist in markets—trust, reciprocity, reputation—can be understood as stable evolutionary outcomes.
Network dynamics: Evolutionary modeling shows how strategies spread through connected populations, useful for understanding the diffusion of behavior and technology.