Title: Convergence to Equilibrium: Potential Games
Abstract:A lingering question from Chapter 3 regarding the Nash equilibrium is why one should ever expect players to play a Nash equilibrium. Perhaps if there exists only one Nash equilibrium, and players are ...A lingering question from Chapter 3 regarding the Nash equilibrium is why one should ever expect players to play a Nash equilibrium. Perhaps if there exists only one Nash equilibrium, and players are introspective and know the entire game model (including other players' payoffs), an expectation of Nash equilibrium play is warranted. But what about other, more typical, cases? What if there are multiple equilibria? What if players know only their own payoffs and have no awareness of the payoffs of other players? Can we really expect Nash equilibrium play in these cases? Unfortunately, in general, the answer seems to be "no." The Nash equilibrium may still be valuable to help us understand the stable states of the system, and it may give us some clue as to how to design systems that will converge to more desirable equilibria. But, in arbitrary games there is no known "learning" algorithm that guarantees convergence of play to a Nash equilibrium when players have limited information.Read More
Publication Year: 2006
Publication Date: 2006-01-01
Language: en
Type: book-chapter
Indexed In: ['crossref']
Access and Citation
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot