Games Real Actors Could Play

The use of game-theoretical explanations and predictions in empirical social science research is often thought to be precluded by the unrealism of the cognitive and computational capabilities that mathematical game theory imputes to its idealized “players†as well as by the prohibitive information costs that an attempt to reconstruct these cognitions and computations would impose on researchers. The article tries to show that these misgivings are exaggerated. Under realistic conditions actors will often be able to pragmatically approximate complete-information conditions regarding each other's strategy options and payoffs. Moreover, empirical research will, in many situations, be able to reconstruct actors' relevant opportunities, perceptions and preferences from socially constructed institutions, norms, and expectations that have always been the subject of mainstream social science research.