Decoding Player Decision-Making in Team Sports

Team performance hinges on teammates’ ability to decide when and how to act; context-dependent decisions differentiate experts from novices in high-pressure domains such as military operations and team sports. Unlike deliberative reasoning, action decisions emerge spontaneously as individuals and teammates co-adapt to achieve task goals, relying on situational awareness cultivated through expertise. Yet, no validated method exists to objectively identify and assess decision-making strategies during complex, fast-paced team tasks. Recent work by the Investigators demonstrated that artificial neural network (ANN) models can predict players’ future actions in a multi-player video game, and that interpretable-AI techniques can reveal the informational cues driving those decisions. Building on these findings, the project will employ interpretable machine learning to predict, understand, and evaluate fast-paced decision behaviors and information-attunement processes of individuals in competitive team sports. The results will establish a generalizable framework for optimizing training and performance in high-pressure settings.