ABSTRACTPrediction of play outcomes is fundamental for sports science, engineering and practice in ballgames. Predicting who obtains a ball after a shot failure called rebound in basketball, is one of the important research subjects. To obtain the rebound, players often compete and move towards the ball drop position. Researchers have analysed important factors of a rebound using basic game statistics and video analysis. However, the most critical factors in the players’ movement to obtain a rebound are unknown. The purpose of this study is to determine the important factors to obtain a rebound and to develop a method to predict who obtains it with player’s positional data. The factors were quantified and divided into three categories; individual position, individual movement and interpersonal relationship, and at least one factor from each category was significantly related with obtaining rebounds by logistic regression. Furthermore, our method predicted who obtained rebounds using logistic regression (70.3%) and support vector machine (71.1%). We also spatially visualised the value of factors related to the shooter’s position through heat maps. Our method has a potential to provide a ground for the evaluation of the effectiveness of practical techniques (e.g. box out) in rebound practice and game execution.