In this paper, we first obtain some fixed point theorems in locally convex topological vector spaces which in turn imply some existence theorems of maximal elements for βΏ-condensing correspondences. Then by employing approximation methods, we prove some existence theorems of equilibria for generaliz
Beliefs correspondences and equilibria in ambiguous games
β Scribed by Giuseppe De Marco; Maria Romaniello
- Publisher
- John Wiley and Sons
- Year
- 2011
- Tongue
- English
- Weight
- 187 KB
- Volume
- 27
- Category
- Article
- ISSN
- 0884-8173
No coin nor oath required. For personal study only.
β¦ Synopsis
The Nash equilibrium concept combines two fundamental ideas. First, rational players choose the most preferred strategy given their beliefs about what other players will do. Second, it imposes the consistency condition that all players' beliefs are correct. This consistency condition has often been considered too strong, and different solution concepts have been introduced in the literature to take into account ambiguous beliefs. In this paper, we show, by means of examples, that in some situation beliefs might be dependent on the strategy profile and that this kind of contingent ambiguity affects equilibrium behavior differently with respect to the existing models of ambiguous games. Hence, we consider a multiple prior approach and subjective beliefs correspondences, which represent an exogenous ability of each player to put restrictions on beliefs over outcomes consistently with the strategy profile; we investigate existence of the equilibrium concepts corresponding to different attitudes toward ambiguity (namely, optimism and pessimism). Finally, we analyze particular beliefs correspondences: beliefs given by correlated equilibria and by ambiguity levels on events.
π SIMILAR VOLUMES
## Abstract We define a new concept of constrained strategic equilibrium (CSE) for Bayesian games. We show that a sequence of CSEs approximates an equilibrium under standard conditions. We also provide an algorithm to implement the CSE approximation method numerically in a broad class of Bayesian g