Testing game theory

Experimentalists frequently claim that human subjects in the laboratory violate such game-theoretic solutions as Nash equilibrium and subgame perfect equilibrium. It is here argued that this claim is usually premature. What have been rejected are certain joint hypotheses concerning subjects’ preferences, rationality and knowledge. This note discusses conceptual and methodological aspects of non-cooperative game theory in its epistemic interpretation. An alternative “empirical” interpretation is outlined, and an associated empirical equilibrium hypothesis is formulated.