Predicting Tennis Serve directions with Machine Learning

Researchers developed a machine learning method to predict professional tennis players' first serve directions, achieving 49% accuracy for male players and 44% for female players. The study provides evidence that top players use mixed-strategy serving decisions and suggests contextual information plays a larger role in tennis strategy than previously understood.