A Quantitative Assessment Model for Momentum in Tennis Based on Exponential Decay and EWM-TOPSIS Method

Abstract

In sports, the concept of "momentum" describes how athletes or teams, inspired by positive factors during a match, perform better, leading to a "success breeds success" phenomenon. This study analyzes players' momentum in tennis, specifically using data from the 2023 Wimbledon Men's Singles Final. It focuses on quantifying momentum and determining its influence on player performance. The study developed the Player Performance Evaluation Model based on EWM-TOPSIS method, incorporating factors like winning status, movement distance, winning shots, and double faults. The model differentially weighs the winning incentives for servers and receivers and uses an exponential decay accumulation of evaluation indicators, akin to the Momentum algorithm in deep learning. The players' performance score at a certain score point, denoted as momentum score, were then derived using the EWM-TOPSIS evaluation algorithm. For the momentum score calculated by our model, we randomly selected a match to visualize the momentum score of both sides to show the consistency of momentum score and match conditions. Finally, this study examines the relationship between momentum score and the increase or decrease of winning probability to prove that momentum score has a significant correlation with the fluctuation of winning probability.