Predicting How People Play Games: Reinforcement Learning in Experimental Games with Unique, Mixed Strategy Equilibria

TLDR

The study investigates learning dynamics in long‑run experimental games with unique mixed‑strategy equilibria. Learning models’ descriptive and predictive accuracy were assessed by simulating each experiment with parameters estimated from other experiments. A simple one‑parameter reinforcement learning model outperforms equilibrium predictions, and adding forgetting, experimentation, or higher rationality further improves predictive power, supporting low‑rationality cognitive game theory. JEL classification: C72, C92.

Abstract

We examine learning in all experiments we could locate involving 100 periods or more of games with a unique equilibrium in mixed strategies, and in a new experiment. We study both the ex post ( best fit) descriptive power of learning models, and their ex ante predictive power, by simulating each experiment using parameters estimated from the other experiments. Even a one-parameter reinforcement learning model robustly outperforms the equilibrium predictions. Predictive power is improved by adding forgetting and experimentation, or by allowing greater rationality as in probabilistic fictitious play. Implications for developing a low-rationality, cognitive game theory are discussed. (JEL C72, C92)