Machine Learning Algorithm Based Prediction of Players' Scores in Tennis Matches

During the men's singles final of the 2023 Wimbledon Open, the match situation fluctuated dramatically, and "momentum" was cited as the cause of these fluctuations. In order to specifically describe and analyze the impact of "momentum" on the match situation, first, we created a set of metrics and applied logistic regression for significance analysis to clearly demonstrate that the system has a significant impact on players' real-time scoring situations. In addition, we used several machine learning algorithms and cross-validation methods based on this system to test the algorithm performance. As a result, we employ LGBM models to predict player "momentum" while providing visualizations of the predicted results. Preliminary Pearson's correlation tests confirm that there is a significant correlation between "momentum" and a player's actual score. In addition, a simple linear regression model was constructed with "momentum" as the independent variable and the player's actual score as the dependent variable. Therefore, it can be concluded that the view expressed by tennis coaches that "a player's swing and success in a match is random" is biased.