Research on Quantifying and Evaluating Tennis Match Momentum Based on Comprehensive Analytical Methods

Abstract

Momentum is crucial in tennis matches, influencing player performance and match outcomes. However, quantifying and evaluating momentum has long been a challenge in sports science research. This study aims to quantify momentum in tennis matches by establishing a hierarchical analysis-based model for momentum flow. The model analyzes factors affecting momentum through five criteria: Service Level, Catch Level, Competition Result, Finishing Ability, and Winning Streak, with respective weights of approximately (0.1, 0.1, 0.2, 0.4, 0.2). Subsequently, the study assigns scores to these criteria based on different actions within a match. By multiplying the weights and scores, the momentum of each player is quantified, and the ratio of the two players' momentum provides a measure of momentum flow during the match. To assess the impact of momentum on match outcomes, the study employs k-means++ clustering analysis, violin plot analysis, and point-biserial correlation analysis. The point-biserial correlation analysis reveals a positive correlation coefficient of 0.66 between a player's momentum and their victory. Similar results are derived from the violin plot and clustering analysis. Therefore, it can be concluded that there is a significant association between the direction of momentum flow and a player's success in a match.