Prediction of Linear Motion

Abstract

The paper briefly discusses the perception of velocity and prediction of motion as processes involved in the performance of fast-ball game skills. Two experiments utilizing a motion-prediction paradigm, in which the subject was allowed to view a moving tennis ball cross the display and disappear from view, are reported. The subject was required to estimate when the ball would reach a target set on the extended motion track. The experiments examined the effects of viewing distance, prediction distance, and stimulus velocity upon the accuracy and variability of performance. The first experiment found that both prediction distance and stimulus velocity were significant variables, their additive effects showing a linear relationship between prediction time and both constant and variable errors. The second experiment demonstrated that these linear relationships were replicable phenomena and that motion prediction became more accurate when subjects were given knowledge of results and allowed to practice the task.