Communicating Chess Strategies in Natural Language

2026-07-13

AI Summary

This study addresses the lack of interpretable strategic reasoning in chess engine move recommendations, which hinders human comprehension. It introduces and formally defines, for the first time, the task of “chess strategy verbalization,” proposing an end-to-end pipeline that integrates engine analysis, natural language generation, and structured evaluation, alongside an objective assessment framework. The research demonstrates that conveying strategic intent accurately cannot rely solely on principal variations or abstract concepts, and that large language models currently cannot fully substitute human judgment. Experimental results validate natural language as an effective medium for communicating chess strategy, offering a more intuitive and explainable pathway for human players to understand AI-generated strategic insights.

Abstract

Chess engines have long achieved superhuman playing strength. However, the underlying strategy behind their move suggestions is difficult for human players, even skilled ones, to comprehend. Motivated by this, we propose the task of chess strategy verbalization, which is to describe chess strategies in natural language. We design (i) a pipeline for verbalizing strategies and (ii) an evaluation framework for objective evaluation of generated strategy descriptions. Our experiments show that natural language is a promising and interpretable medium for communicating strategic information to both human and LLM players. We glean additional interesting insights, including (a) the importance of evaluating strategies beyond the main line, (b) the limitations of pure concept-based descriptions, and (c) the limitations of relying on LLMs rather than humans for evaluation.

Problem

• chess strategy
• natural language
• verbalization
• interpretability
• human comprehension

Innovation

• chess strategy verbalization
• natural language generation
• interpretability
• evaluation framework
• human-AI communication