Successful Nash Equilibrium Agent for a 3-Player Imperfect-Information Game
This addresses the open problem of building effective AI agents for multiplayer games, though it is incremental as it focuses on a specific game type.
The researchers tackled the problem of creating strong agents for multiplayer games by developing an agent that uses an exact Nash equilibrium strategy to defeat realistic opponents in a 3-player imperfect-information game, demonstrating success despite theoretical limitations.
Creating strong agents for games with more than two players is a major open problem in AI. Common approaches are based on approximating game-theoretic solution concepts such as Nash equilibrium, which have strong theoretical guarantees in two-player zero-sum games, but no guarantees in non-zero-sum games or in games with more than two players. We describe an agent that is able to defeat a variety of realistic opponents using an exact Nash equilibrium strategy in a 3-player imperfect-information game. This shows that, despite a lack of theoretical guarantees, agents based on Nash equilibrium strategies can be successful in multiplayer games after all.