3:00–4:00 pm Eckhart Hall, Room 202
Title: "Game Dynamics As The Meaning Of The Game"
Abstract: The modern era of game theory started with Nash's theorem in 1950, establishing that all finite games have a stable solution from which players will not deviate. When computer scientists embraced game theory four decades later, computational flaws of this concept came under scrutiny: The Nash equilibrium is not unique, and it is intractable to find one. I will recount how three theorems, serendipitously proved during this past year, suggest an alternative meaning of the game: A game can be seen as a mapping from a prior distribution of the players' behavior to the limit distribution under the dynamics of repeated play, and reasonable variants of this mapping can be computed efficiently.