Research on Momentum Evaluation and Turning Point Prediction in Tennis Matches Based on RSR and GSRF

Abstract

In the men's singles final of the 2023 Wimbledon Open, rising star Carlos Alcaraz defeated veteran Novak Djokovic, an astonishing turnaround and result that sparked an in-depth exploration of the sports phenomenon of "momentum". When one side of a match gains momentum, its technical level and adaptability will be greatly utilized, which will directly affect the victory or defeat of the match. Most of the past researches focus on verifying the authenticity of momentum in sports games and the factors affecting the change of momentum. In this paper, taking the above match as the research object, we use the rank and ratio comprehensive evaluation method (RSR) to mine the match data, and propose a new prediction model GSRF by combining the random forest model and the lattice search algorithm, and based on the independently established momentum model and the momentum difference model, we predicted the turning point of the momentum in the match and the result of the match. The difference effect size of Cohen's d-value of 0.035 and accuracy R² of 0.826 was obtained, which verified the higher accuracy and generalizability of the model. The study helps coaches and instructors optimize training programs, develop game strategies, and provide new ideas and approaches for coaches and instructors.