A statistical model to predict the results of Novak Djokovic's matches in the Australian open tennis event using the binary logistic regression

Abstract

The purpose of the research was to construct a model that could forecast the probability of winning in the case of Novak Djokovic in the men’s singles grand slam event of the Australian Open and to determine the relative relevance of the match data that contribute to victory. A total number of 147 matches were recorded for all nine years i.e., from 2013 to 2021, from the first round to the exit round over the years. One of the few assumptions in logistic regression is that the dependent variable must be binary in nature. Therefore, the dependent variable selected for this study was Match Outcome (Win/Loss). Ace, (DF) Double Fault, (FS) First Serve, (FSPW) first serve point win, (SSPW) second serve point win, (BPC) Breakpoint converted, and (TPW) Total point win were selected as the predictor variables. All the data were collected from ATP world tour.com. In order to accomplish the goals of the research, the only matches that Novak Djokovic competed in during the Grand Slam AO (Australian Open), were analyzed. The prediction of the likelihood of Mr. Novak Djokovic winning or losing in the men’s singles Australian open grand slam by fitting the logistic regression model. According to the statistical significance of the predictor variables, they were numerically weighted and can be used to predict the match outcome. Out of seven predictor variables, only the variable Breakpoint Converted was included in the prediction model with a coefficient of determination (R2) of.424 (Cox & Snell) and .588 (Nagelkerke). The case adds seven independent variables and one dependent binary logistic variable for all the Australian Open Grand slam matches played from 2013 to 2021. The given result of it verifies conclusive evidence that the prediction fits quite well as it classifies an 88.9% winning probability.