Decoding Critical Strokes in Para Table Tennis Matches: A Machine Learning Approach to Identify 3S-Based Tactical Strengths and Weaknesses

Purpose: This study utilized machine learning techniques to analyze the key factors influencing match outcomes in table tennis (TT) by integrating players' technical characteristics and the 3S attributes (spin, speed, and spot). It aimed to provide a foundation for formulating effective competition strategies for elite athletes with impairments. Methods: A case study was conducted on Taiwanese Para TT player Po-Yen Chen, the silver medalist in men’s singles at the 2024 Paris Paralympic Games and a top 5 world-ranked player as of March 2025. Data were collected using The Intellectual System in Competitive TT from 20 international matches against top 10 players with intellectual impairments (2023–2024). Results: 1. Chen’s scoring rate during service rounds was significantly higher than during receive rounds (χ²=15.85, p< .001). 2. Critical scoring and losing points predominantly occurred between the 2nd and 4th strokes, with the 4th stroke exhibiting the highest occurrence rate (21.3%). 3. Key features influencing stroke outcomes were identified as follows: 1st stroke (serve): Spin (side-back spin); 2nd stroke (receive): Skill (side twist); 3rd stroke: Skill (counter drive); and 4th stroke: Spin (back spin). Conclusion: Machine learning effectively identifies elite athletes’ tactical strengths and weaknesses across strokes, providing valuable insights for coaches to develop targeted training programs. These findings can significantly enhance competitive performance in para TT.