The Discriminant Analysis And Result Prediction Of The Key Winning Factors Based For ATP Men’s Single Game

Data mining technology has been developed deeply into many aspects ofpeople’s life. And the application in the tennis match is changing the way peoplewatch the game and provide audience an effective and statistical data analysis. Bymining the data that hidden behind the match, coaches and players can obtain a betterunderstanding on the performance of the players. It is helpful for the players toposition the problem and working on further practice to overcome their weaknesses.For research purposes, we set Rafael Nadal, a world No.1Spanish tennis player,as our research object. All the competition data comes from the ATPτAssociation ofTennis ProfessionalsυOfficial Website, including89single matches play against otherthree members of Big4group: Roger Federer Novak Djokovic and Andy Murry.Every single match data set contains9indicators refer to serve and return. Thestepwise discriminant analysis, which can be done by SPSS software, can determineand sequence the key winning factors of the match:2nd serve return point won,2ndserve point won,1st serve return point won,1st serve point won, and break pointsaved. With Fisher’s method and Bayes’s method, we can establish a function todiscriminant and predict the match result. The second step is examining the effect ofdiscriminant function. The result of interactive test shows that the success rate ofdiscriminant is95.5%.Tennis court has three types: hard, red clay and grassland.According to data analysis of30matches on two different court types between RogerFederer and Rafael Nadal, through Nonparametric test of two independent totality weobtained the two important discoveries. First, on the hard court,1st serve point wonand1st serve return point won are the two key success factors. Second, on the red claycourt,1st serve point won,2nd serve point won,2nd serve return point won are thekey success factors. In conclusion, the1st serve point won and the2nd serve returnpoint won represent the statistical significant.