Extremality and equilibrium with respect to given functions and game- theoretic models. II. Strong equilibria (Q799501)

The weak active equilibria derived in the first part of this article [Autom. Remote Control 44, 358-363 (1983; Zbl 0529.90102)] constitute a generalization of such standard equilibria as Nash equilibrium, saddle points, and ''objection-counterobjection'' equilibrium and are applicable, as we have seen by examining various examples, to a number of decision- making games. The motivation for other equilibria stems from the need to reduce the set of equilibrium decisions (for instance the symmetrical weak active equilibrium is generally not unique); moreover, the reduction of the equilibrium set should be performed in such a way as to produce a stronger and more stable equilibrium. If the two players face a certain situation \(x\in U\) to which the second player has agreed, then the first player also will be forced to agree to this situation, at least in the framework of the ''preference game'' [see part I of this article]. The more beneficial the point x for the second player, the more reason for him to agree to this point and therefore the more likely it is that both players will accept this point. Therefore, in order to isolate from the set \(U_ 1\) those points which are preferred by the second player, we have to identify the ''more stable'' points in the set D, which constitute ''better equilibria''.