Emergent Complexity in the Decision-Making Process of Chess Players

TL;DR

This study analyzes chess players' decision-making complexity using engine evaluations, revealing links between decisiveness, performance, and accuracy, and proposing a model to replicate observed behaviors.

Contribution

The paper introduces a decisiveness metric for move evaluation and a simple model that captures the emergent complexity in players' decision-making processes.

Findings

Players exhibit a wide range of decisiveness, indicating complex decision-making.

Decreased decisiveness correlates with lower performance levels.

Players are more accurate in high decisiveness positions regardless of skill level.

Abstract

In this article, we study the decision-making process of chess players by using a chess engine to evaluate the moves across different pools of games. We quantified the decisiveness of each move during the games using a metric derived from the engine's evaluation of the positions. We then performed a comparative analysis across players of varying competitive levels. Firstly, we observed that players face a wide spectrum of the decisiveness metric, evidencing the complexity of the process. By examining groups of winning and losing players, we found evidence where a decrease in complexity may be associated with a drop in players' performance levels. Secondly, we observed that players' accuracy increases in positions with high values of the decisiveness metric regardless of competitive level. Complementing this information with a null model where players make completely random legal moves allowed us to characterize the decision-making process under the simple strategy of making moves that minimize the decisiveness metric. Finally, based on this idea, we proposed a simple model that approximately replicates the global emergent properties of the system.