Prediction of Basketball Skill Using Biomechanical Variables

Abstract

This study was designed to examine the use of selected biomechanical variables in the prediction of basketball skill. The subjects were college women in three mutually exclusive groups of basketball skill: an elite group of six competitors on the United States team in the World University Games, a good group of seven players on a varsity team, and a poor group of nine members of an instructional class. An accuracy test and digitized film records provided the data for 12 variables related to the process or product of free throw shooting. Discriminant analysis was employed to predict the categorical variable of skill. The most discrimination came from variables of accuracy, stability, and height of release rather than from variables of projection. Poor shooters were distinguished by instability; elite shooters were characterized by a high point of release and accuracy under pressure. Depending on the method of prediction, rates for correct classification of subjects ranged from 76–100%. Thus, it appears that discriminant analysis using biomechanical variables can be a successful tool in the prediction of basketball skill.