Strategic Decision-Making and Expertise in Chess: An Engine-Based and Behavioral Analysis Across Skill Levels

Abstract

Chess is widely regarded as an effective model for studying human cognition, expertise, and strategic decision-making under conditions of complexity and time pressure. This study examines how decision quality differs across players of varying skill levels by integrating engine-based performance analysis with survey-based behavioral insights. The research compares elite players, titled master-level players, and intermediate or lower-rated players using metrics such as move accuracy, average centipawn loss (ACPL), blunders, mistakes, and inaccuracies. In addition, the study explores behavioral dimensions of decision-making, including reliance on intuition versus calculation, time management strategies, perceived causes of errors, and differences between online and offline play. The study adopts a mixed-methods research design combining quantitative engine evaluations with primary survey data collected from chess players across multiple rating categories. The findings indicate that stronger players consistently demonstrate lower ACPL, fewer severe errors, and more stable decision-making patterns across games. Survey-based insights further reveal that higher-rated players rely more effectively on intuition, pattern recognition, and selective calculation, while lower-rated players experience greater indecision, over-calculation, and time pressure. The results also suggest that expertise influences not only the frequency of errors but also their severity and consistency under complex playing conditions. Overall, the study contributes to the literature on chess expertise and bounded rationality by linking objective performance metrics with underlying cognitive and behavioral processes. The findings highlight that expert performance depends not only on technical knowledge but also on efficient cognitive processing, adaptability, and consistent decision-making under pressure.