Accelerating Skill Assessment in Chess: A Drift-Diffusion-Enhanced Elo Rating System

본문 미리보기

arXiv:2606.26267v1 Announce Type: new Abstract: Rating systems such as Elo serve as the gold standard for matchmaking in competitive chess. However, they inherently suffer from response lag due to their exclusive reliance on match outcomes, neglecting the granular quality of gameplay. Nevertheless, incorporating move-by-move information into rating adjustments presents a significant challenge given the substantial noise and the vastness of the game-state space. To address this, we propose the D