Alliance in Three Person Games
V.I. Zhukovskiy, Moussa Larbani
Abstract
Open-access reader
V.I. Zhukovskiy, Moussa Larbani
Abstract
Open-access reader
In the present paper, we suggest a new concept of an optimal solution (that we call "coalitional equilibrium") based on the concepts of Nash and Berge equilibria. We apply the concept of an optimal solution where the outcome of a deviant coalition cannot increase. Then we determine sufficient conditions of existence of a coalitional equilibrium using the Germeier convolution. The convolution transforms the problem of determining a coalitional equilibrium into finding a saddle point of a special antagonistic game that can be effectively constructed based on the mathematical model of the initiall game. As an example of application, we suggest the proof of existence of a coalitional equilibrium in mixed strategies under "regular" mathematical programming limitations: continuity of players' outcome functions and compactness of sets of strategies. This work is intentionally limited to three persons to avoid cumbersome notations and calculations, even though application of the suggested method to games with more than three players is promising for solving problems of creating stable coalitions.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In the present paper, we suggest a new concept of an optimal solution (that we call "coalitional equilibrium") based on the concepts of Nash and Berge equilibria. We apply the concept of an optimal solution where the outcome of a deviant coalition cannot increase. Then we determine sufficient conditions of existence of a coalitional equilibrium using the Germeier convolution. The convolution transforms the problem of determining a coalitional equilibrium into finding a saddle point of a special antagonistic game that can be effectively constructed based on the mathematical model of the initiall game. As an example of application, we suggest the proof of existence of a coalitional equilibrium in mixed strategies under "regular" mathematical programming limitations: continuity of players' outcome functions and compactness of sets of strategies. This work is intentionally limited to three persons to avoid cumbersome notations and calculations, even though application of the suggested method to games with more than three players is promising for solving problems of creating stable coalitions.
Key concepts: Nash equilibrium, Outcome (game theory), Solution concept, Mathematical economics, Best response, Computer science, Compact space, Saddle point