Sports Injury Prediction in Professional Tennis

Zhaoyu Liu, Kan Jiang, Jin Song Dong

Abstract: Each week, there are 2-3 top-level professional tennis tournaments around the world offering attractive prize money and ranking points for the players. Even though match playing is an important part of the professional tennis player’s career development, sports injuries will happen if the player plays too many tournaments. By analyzing historical data and studying cases where players retired during matches, we have been able to identify specific scheduling patterns that correlate with a higher risk of injury. As a result, we propose a dynamic tennis injury prediction model as well as an explainable pattern-mining method to obtain player-specific injury patterns based on historical data. Both approaches take into account features such as the player’s physical condition, match performance, injury history, and recent tournament schedule. With these comprehensive approaches, we aim to balance competitive success and the preservation of players’ well-being.