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 -