Quantifying the Effect of Anticipatory Eye Movement on Successful Ball Hitting Using Fine-Scale Tracking and SHAP-Analysis

Abstract

Previous studies have suggested that anticipatory eye movement is important for success in dynamic interception tasks, such as ball sports. However, there are few quantitative studies that explore the connection between specific eye movement and successful ball interception. In this study, we used a state-of-the-art motion-capture and eye-tracking system to monitor the eye and body movements of 91 individuals playing ping-pong against a wall. We analyzed 24 features related to eye movement and action, using nonlinear machine learning models to understand the relationship between these features and successful hits. Our key findings showed that the accuracy and timing of anticipatory saccades to the ball’s bounce points were crucial for success. Successful hits were characterized by more precise, just-in-time anticipatory looks and longer duration of ball pursuit compared to unsuccessful hits. This quantitative framework is the first step towards understanding how humans plan their actions based on visual perception in realistic, dynamic tasks.