Computer Science – Computer Science and Game Theory
Scientific paper
2005-09-20
Computer Science
Computer Science and Game Theory
Appeared in: Proc. of the 10th conference on Theoretical Aspects of Rationality and Knowledge (TARK X), pp. 22-38 (2005)
Scientific paper
Two natural strategy elimination procedures have been studied for strategic games. The first one involves the notion of (strict, weak, etc) dominance and the second the notion of rationalizability. In the case of dominance the criterion of order independence allowed us to clarify which notions and under what circumstances are robust. In the case of rationalizability this criterion has not been considered. In this paper we investigate the problem of order independence for rationalizability by focusing on three naturally entailed reduction relations on games. These reduction relations are distinguished by the adopted reference point for the notion of a better response. Additionally, they are parametrized by the adopted system of beliefs. We show that for one reduction relation the outcome of its (possibly transfinite) iterations does not depend on the order of elimination of the strategies. This result does not hold for the other two reduction relations. However, under a natural assumption the iterations of all three reduction relations yield the same outcome. The obtained order independence results apply to the frameworks considered in Bernheim 84 and Pearce 84. For finite games the iterations of all three reduction relations coincide and the order independence holds for three natural systems of beliefs considered in the literature.
No associations
LandOfFree
Order Independence and Rationalizability does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Order Independence and Rationalizability, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Order Independence and Rationalizability will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1387