The games agents play--in markets, conflicts, or most other contexts--often defy strict game-theoretic analysis. Games may be unmanageably large (combinatorial or infinite state or action spaces), and present severely imperfect information, which could be further complicated by partial dynamic revelation. Moreover, the game may not even be specified in the precise form required for game-theoretic reasoning. For example, we may have at best a simulator or limited access to the real world for experimentation, or some other form of experiential data or knowledge.

With colleagues and students over the past few years, I have been developing a body of techniques for strategic analysis, adopting the game-theoretic framework but employing it in domains where direct "model-and-solve" cannot apply. This empirical game-theoretic methodology embraces simulation, approximation, statistics and learning, and search. Through examples of such techniques and illustrative application to auction games, supply chains, 4-player chess, and other scenarios, I argue that such a toolkit can support practical automation of routine strategic reasoning.