English

Grouped Satisficing Paths in Pure Strategy Games: a Topological Perspective

Computer Science and Game Theory 2025-09-30 v1 Machine Learning Multiagent Systems

Abstract

In game theory and multi-agent reinforcement learning (MARL), each agent selects a strategy, interacts with the environment and other agents, and subsequently updates its strategy based on the received payoff. This process generates a sequence of joint strategies (st)t0(s^t)_{t \geq 0}, where sts^t represents the strategy profile of all agents at time step tt. A widely adopted principle in MARL algorithms is "win-stay, lose-shift", which dictates that an agent retains its current strategy if it achieves the best response. This principle exhibits a fixed-point property when the joint strategy has become an equilibrium. The sequence of joint strategies under this principle is referred to as a satisficing path, a concept first introduced in [40] and explored in the context of NN-player games in [39]. A fundamental question arises regarding this principle: Under what conditions does every initial joint strategy ss admit a finite-length satisficing path (st)0tT(s^t)_{0 \leq t \leq T} where s0=ss^0=s and sTs^T is an equilibrium? This paper establishes a sufficient condition for such a property, and demonstrates that any finite-state Markov game, as well as any NN-player game, guarantees the existence of a finite-length satisficing path from an arbitrary initial strategy to some equilibrium. These results provide a stronger theoretical foundation for the design of MARL algorithms.

Keywords

Cite

@article{arxiv.2509.23157,
  title  = {Grouped Satisficing Paths in Pure Strategy Games: a Topological Perspective},
  author = {Yanqing Fu and Chao Huang and Chenrun Wang and Zhuping Wang},
  journal= {arXiv preprint arXiv:2509.23157},
  year   = {2025}
}
R2 v1 2026-07-01T06:00:27.672Z