Abstract: Using BP neural network and multiple regression, this study analyzed the technical and tactical ability of elite male table tennis players. Results show that: (1) The models based on both BP neural network and multiple regression are highly efficient in fitting and forecasting. The BP neural network-based model performs better than the multiple regression-based model in both aspects; (2) The primary technical and tactical ability of elite male players is their attack-after-service ability, and the second their rallying ability and receiving and blocking ability in serving rounds, the fourth their rallying ability in multiple rounds of serving and returning; (3) There is a compensation effect between table tennis technical and tactical abilities. In a match featuring different technical and tactical levels, a total score of over 13 points means the game is over.

Key words: BP Neural Network, Multiple Regression, Table Tennis, Technique and Tactics