On Self-Enforcement in Extensive-Form Games
1990 (English)Report (Other academic)
This paper discusses the issue of self-enforcement in finite extensive form games. A new criterion is developed which essentially requires that, for any given degree (є) of accuracy, every player's strategy should be locally optimal at each of his information sets under some probabilistic hypothesis about all player's behavior which at least approximately (within є) agrees with the tested profile at all points in the game where his information set provides no evidence of a deviation. Like sequential equilibrium, this criterion is more stringent than subgame perfection but less stringent than ("trembling hand" extensive-form) perfect equilibrium. In contrast to sequential equilibrium, the approach builds in an individualistic and set-theoretic consistency condition which can be viewed as a version of Occam's razor.
Place, publisher, year, edition, pages
Stockholm: IIES , 1990. , 41 p.
Seminar Paper / Institute for International Economic Studies, Stockholm University, ISSN 0347-8769 ; 470
IdentifiersURN: urn:nbn:se:su:diva-41774OAI: oai:DiVA.org:su-41774DiVA: diva2:337720