2023/07/17 by Dan Zwillinger, Zwillinger, Dan, Paul San Clemente +1
Computer Science · #Artificial Intelligence in Games #Computer Science and Game Theory (cs.GT) #FOS: Computer and information sciences
paper · pdf · doi:10.48550/arxiv.2307.10257
openalex publication_date 2023/07/17 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
In classical game theory, optimal strategies are determined for games with complete information; this requires knowledge of the opponent's goals. We analyze games when a player is mistaken about their opponents goals. For definitiveness, we study the (common) bimatrix formulation where both player's payoffs are matrices. While the payoff matrix weights are arbitrary, we focus on strict ordinal payoff matrices, which can be enumerated. In this case, a reasonable error would be for one player to switch two ordinal values in their opponents payoff matrix. The mathematical formulation of this problem is stated, and all 78 strict ordinal 2-by-2 bimatrix games are investigated. This type of incomplete information game has not -- to our knowledge -- been studied before.