Static multilevel reverse Stackelberg games: existence and computations of best strategies

TL;DR

This paper investigates the existence and construction of strategies in multilevel reverse Stackelberg games, extending the theory to nonconvex cases and providing methods for multiple strategy development.

Contribution

It develops conditions for affine strategy existence and introduces a method to construct multiple strategies in multilevel reverse Stackelberg games, including nonconvex scenarios.

Findings

Established conditions for affine strategy existence

Extended analysis to nonconvex sublevel sets

Provided a method for multiple strategy construction

Abstract

The multilevel reverse Stackelberg game is considered. In this game, the leader controls the outcome by announcing a strategy as a function of decision variables of the followers to his/her own decision space. Corresponding to the leader's strategy, the player in the next level presents his/her strategy as a function of decision variables of the remaining players. This procedure is repeated until it is the turn of the bottom level player in the hierarchy, who reacts by determining his/her optimal decision variables. The structure of this game can be adopted in decentralized multilevel decision making like resource allocation, energy market pricing, problems with hierarchical controls. In this paper conditions for existence and construction of affine leader reverse Stackelberg strategies are developed for such problems. As an extension to the existing literature, we considered nonconvex sublevel sets of objective functions of followers. Moreover, a method to construct multiple reverse Stackelberg strategies for the leader is also presented.