In order to explore the mystery of momentum in the competition, so that the athletes can better maintain good momentum and win the competition, we established two models: the momentum model based on entropy weight TOPSIS method, and the performance prediction model based on BP neural network. On the basis of data preprocessing, we identified ten indicators affecting momentum, evaluated them with the TOPSIS method, and determined the target weights with the entropy weighting method, and established the momentum model to determine the performance of a player at a specific time. A BP neural network-based performance prediction model is built to analyze whether the success of the game showed a certain pattern or trend. By introducing all the data in the training set into the BP neural network, allowing it to train and learn, and then predicting the data in the test set, we validated the high accuracy of the model's predictions.