Analysis of technical-tactical performance in table tennis based on machine learning and evaluation theory

Using machine learning, the technical effectiveness of table tennis players and their winning probability were modeled. This study adopted a novel algorithm, SHapley Additive exPlanation (SHAP), to analyze the important features based on the gradient boosting + categorical features-tree-structured parzen estimator (Catboost-TPE) with the four-phase evaluation theory, an analysis framework in a table tennis match. A total of 110 singles’ matches (9536 rallies) were analyzed, and 59 male players’ winning rates from 2018 to 2022 were categorized into three levels (high, medium, low) by k-means cluster analysis. The results showed that Catboost-TPE has the best performance (MSE = 7.5e-05, MAE = 0.006, RMSE = 0.008, R 2 = 0.99, and adjusted R 2 = 0.989) among six hybrid machine learning algorithms. Using Catboost-TPE to calculate the SHAP value of each feature, the global interpretation and multiple local interpretations found that the performance of receive-attack and serve-attack phases have essential impacts on the winning probabilities in current matches. Besides, four and three phase evaluation theory are all import framework in table tennis match analysis. To further deepen the theoretical and applied value of the four-phase evaluation theory, this study derived the mathematical equations for converting indicators from the four-phase evaluation theory into the new three-phase evaluation theory. These results provided quantitative references to table tennis matches’ characteristics and winning phases. These methods used in the study can be widely applied to other sports performance analyses, and the equations derived in this study are also instructive for relative sports.