TY - GEN

T1 - When the players are not expectation maximizers

AU - Fiat, Amos

AU - Papadimitriou, Christos

PY - 2010

Y1 - 2010

N2 - Much of Game Theory, including the Nash equilibrium concept, is based on the assumption that players are expectation maximizers. It is known that if players are risk averse, games may no longer have Nash equilibria [11,6]. We show that 1 Under risk aversion (convex risk valuations), and for almost all games, there are no mixed Nash equilibria, and thus either there is a pure equilibrium or there are no equilibria at all, and, 1 For a variety of important valuations other than expectation, it is NP-complete to determine if games between such players have a Nash equilibrium.

AB - Much of Game Theory, including the Nash equilibrium concept, is based on the assumption that players are expectation maximizers. It is known that if players are risk averse, games may no longer have Nash equilibria [11,6]. We show that 1 Under risk aversion (convex risk valuations), and for almost all games, there are no mixed Nash equilibria, and thus either there is a pure equilibrium or there are no equilibria at all, and, 1 For a variety of important valuations other than expectation, it is NP-complete to determine if games between such players have a Nash equilibrium.

UR - http://www.scopus.com/inward/record.url?scp=78649611612&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-16170-4_1

DO - 10.1007/978-3-642-16170-4_1

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:78649611612

SN - 3642161693

SN - 9783642161698

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 1

EP - 14

BT - Algorithmic Game Theory - Third International Symposium, SAGT 2010, Proceedings

T2 - 3rd International Symposium on Algorithmic Game Theory, SAGT 2010

Y2 - 18 October 2010 through 20 October 2010

ER -