Research on the prediction of sports competition results based on momentum modeling

Abstract

This study examines the role of “momentum” in sports competitions. Using a trained model to predict all samples, we analyzed the correlation and effect of “momentum” with athletes' scores through Pearson's correlation coefficient and one-way linear regression. The results showed that “momentum” was moderately correlated with athletes' scores and had a significant effect on the outcome of the game. We also built a judgmental descriptive model, aggregated and normalized the data, and performed binary logistic regression analysis. The results show that the model has good accuracy in predicting whether an athlete will be able to win a race. In addition, we validated it using machine learning algorithms and plotted ROC curves and AUC values. The results show that traditional machine learning models perform well in predicting the outcome of a single match. Finally, suggested strategies are proposed for different types of players, including studying the opponent's game video in depth, emphasizing on psychological preparation, making specific tactical plans, and maintaining flexibility and adaptation to improve performance in the game.