An Investigation of Competitive Preference Structures and Posterior Performance Through a Bayesian Decision-Theoretic Approach

In this paper we analyze competitive decision-making situations in terms of their preference structures and posterior performance, through a Bayesian decision-theoretic framework. The setting is that of a two-by-two, two-person, non-zero-sum and noncooperative game which is repeated over time. The dynamic behavior of the competitors for different classes of games, as identified by their preference structures, is examined and a classification scheme is proposed for the purpose of unification. The competitors' dynamic behavior and posterior performance for some general classes of games is then derived, and the relationship to the results implied from game-theoretic considerations is discussed. Illustrative examples are given, too.