Fair sharing in cooperative games (Q960683): Difference between revisions

In this paper, the author considers cooperative games. He works with the notion of fair sharing and suggests a method for determining it. He does this taking into account a known conflict equilibria. His starting point is that sharing must be individually and coalitionally stable in the sense of the core provided by \textit{D. B. Gillies} [Some theorems on \(n\)-person games. Ph.D. thesis, Princeton (1953)]. He begins by recalling how \textit{J. von Neumann} and \textit{O. Morgenstern} [Theory of games and economic behavior, 2nd. ed., Princeton University Press, Princeton, N. J. (1947; Zbl 0063.05930)] define the characteristic function associated to a strategic game in which players cooperate. Moreover he also considers the definition of solution or sharing of the cooperative gain proposed by von Neumann and Morgenstern. He is critical with this definition mainly having into account that it does not always exist and when it exists, it is essentially always nonunique. Even, he points out some problems with the Shapley value [\textit{L. S. Shapley}, Ann. Math. Stud. 2, 307--317 (1953; Zbl 0050.14404)] considered as fair sharing because it does not satisfy certain natural stability conditions. The author considers that it is important to construct a satisfactory theory of game equilibria and not only have into account the characteristic function, for developing an adequate theory of cooperative games. Following \textit{E. R. Smol'yakov} [The theory of conflict equilibria, Editorial URSS, Moscow (2005)], he considers the called sequence of optimal coalitions associated to a characteristic function that is obtained by using the mean increments in the income of the different subsets of players. A sequence of optimal coalitions imposes natural constraints on the fair sharing of a cooperative gain. Using the notions of equilibrium of the problem introduced in [\textit{Smol'yakov} (2005)] he provides a procedure to determine the unique fair sharing in any game. An example illustrates the complete procedure.