The purposes of this study were: 1. to explore the differences of the offensive and defensive records among different positional players, 2. to explore the predictions of playing minutes per game and scores for the offensive and defensive records among different positional players. The subjects were from the all season's games of male players from teams of NCAA from 2005 to 2010 of basketball final games. The offensive and defensive record data were analyzed by independent one-way ANOVA and multiple stepwise regression analysis. The results were as follows: 1. For differences: there were significant differences among different player in two-point shot percentage, three-point shot percentage, total rebounds, offensive rebounds, defensive rebounds, assists, blocked shots, and fouls; 2. for predictions: different offensive and defensive records could predict playing minutes per game and scores for different positional players. Conclusion: every basketball player in different position has exactly special performances. Finally, suggestions are made according to the study purposes and results which contribute to disciplines and selections for domestic basketball.