Logistic回归模型对女子网球制胜因素模型的构建及分析——以澳大利亚网球公开赛为例

The Construction and Analysis of the Logistic Regression Model to the Winning Factors Model of Women’s Tennis——A Case Study of Australia Open

本文旨在运用二元logistic回归模型对2018、2019、2020、2021年澳大利亚网球公开赛女子双打的共256组选手,252场比赛进行分析,探究影响女子双打比赛的制胜因素指标,为我国女双运动员备赛提供建议和参考。以澳网官网公布12项技术指标进行相关性分析,剔除三个指标,将9个技术指标作为自变量,比赛结果为因变量,其中1代表获胜,0代表失败,构建女子双打制胜双打模型。回归分析得出影响比赛结果最为显著的四个指标分别是一发成功率,一发得分率,二发得分率以及破发点成功率。通过对建立的方程拟合优度检验,发现方程拟合度较好,以及回归模型预测正确度分析,胜负结果预判较准确。建议增加提高双打一发成功率的专门性练习,加强接发球局接发的能力,提高运动员破发的能力。

This article aims to use the binary logistic regression model to analyze the 256 groups of players and 252 matches of the 2018,2019,2020,and 2021 Australian Open Women’s Doubles,and to explore the factors affecting the winning factors of the women’s doubles competition and provide suggestions for Chinese women’s doubles players.The 12 technical indicators announced on the official website of Australia Open were used for correlation analysis,three indicators were eliminated,9 technical indicators were used as independent variables,and the result of the game was the dependent variable,of which 1 represents victory and 0 represents failure.The model of women’s doubles winning doubles was constructed.The regression analysis shows that the four most significant indicators that affect the game results are the first serve success rate,the first serve scoring rate,the second serve scoring rate and the break point success rate.By testing the goodness of fit of the established equation,it is found that the fit of the equation is better,and the prediction accuracy of the regression model is analyzed,and the prediction of the outcome of the outcome is more accurate.It is recommended to increase the special practice to improve the success rate of first serve in doubles,strengthen the ability of receiving and serving rounds,and improve the ability of athletes to break serve.