Quantitative Analysis and Prediction of Momentum Based on TOPSIS and SVR Models

The purpose of this study is to investigate the factors that influence the outcome of a tennis match, with a special focus on the role of "momentum" in the match. By combining entropy weighting, AHP and game-theoretic combinatorial weighting, a comprehensive model is designed to assess the key factors of player performance in a match. Using the TOPSIS model for the synthesis process, it is found that Player 1 performs better in the game and can turn the situation around even after suffering a setback. Correlation tests showed that momentum was positively correlated with win percentage and turnover percentage. An indicator of player momentum difference was introduced, and the SVR prediction model was used and validated by cross-validation to verify the accuracy of the model. The most critical metrics found in the study were first serve return rate and unforced errors, suggesting effective tactical strategies for players.