Two-person positive shortest path games have Nash equlibria in pure stationary strategies

Endre Boros,Khaled Elbassioni,Vladimir Gurvich,Mikhail Vyalyi
2024-10-12
Abstract:We prove that every finite two-person positive shortest path game has a Nash equilibrium (NE) in pure stationary strategies, which can be computed in polynomial time. The existence result holds also for graphs with finite out-degrees. Moreover, we prove that a terminal NE exists provided at least one of two players can guarantee reaching a terminal. If no one can do it, in other words, if each of two players can cut all terminals from the initial position $s$, then, obviously, a cyclic NE exists, although its cost is infinite for both players, since we restrict ourselves to positive games. We conjecture that a terminal NE exists too, provided there exists a directed path from $s$ to a terminal. However, this is open.
Discrete Mathematics,Multiagent Systems,Optimization and Control
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to prove the existence of Nash Equilibrium (NE) in the finite two - person positive shortest - path game, and this Nash Equilibrium can be calculated in pure strategies and fixed strategies. Specifically: 1. **Existence of Nash Equilibrium**: - The author proves that every finite two - person positive shortest - path game has a Nash Equilibrium in pure fixed strategies. - This Nash Equilibrium can be calculated by a polynomial - time algorithm. 2. **Existence of Terminal Nash Equilibrium**: - If at least one player can ensure reaching the end point, then there exists a terminal Nash Equilibrium (i.e., a Nash Equilibrium where the game ends at a certain terminal node). - If neither player can ensure reaching the end point, then obviously there will be a cyclic Nash Equilibrium, but at this time the costs for both are infinite. 3. **Assumptions and Conditions**: - All local costs are positive (i.e., \( r_i(e)>0 \) for all edges \( e\in E \) and players \( i\in [n] \)). - Every vertex and edge in the graph can be reached from the initial position \( s \). 4. **Conjecture**: - The author proposes a conjecture: if there is a directed path from the initial position \( s \) to a certain terminal in the graph, then there exists a terminal Nash Equilibrium. This conjecture has not been proven yet. ### Formula Summary - **Effective Cost**: \[ R_i(p)=\sum_{e\in p}r_i(e) \] where \( R_i(p) \) is the effective cost of player \( i \) on path \( p \), and \( r_i(e) \) is the local cost of edge \( e \). - **Nash Equilibrium Condition**: Nash Equilibrium \( \sigma = (\sigma_1,\sigma_2) \) satisfies: \[ R_i(p(\sigma))\leq R_i(p(\sigma'))\quad \text{for all players }i\in [n] \] where \( \sigma' \) is a strategy combination that differs only in the strategy of the \( i \)-th player. ### Conclusion This paper, through strict mathematical proofs, shows the existence and computability of Nash Equilibrium in the two - person positive shortest - path game under specific conditions. This is of great significance for understanding strategy selection and equilibrium states in multi - person games.