Using Principal Component Analysis to Study Tennis Shot Anticipation

Resumen analysis to study tennis shot anticipation

What visual information underpins complex sporting acts such as the anticipation of tennis forehand stroke direction? Previous attempts to identify the dynamic information have relied on spatial and temporal occlusion methodologies that treat body segments or windows of time independently. To overcome this limitation Principal Component Analysis (PCA) has recently been used because it can reveal patterns of movements that are otherwise masked when parts of the body are analysed in isolation. When PCA is used to identify patterns of movement in tennis players forehand ground stroke kinematics, collected from shots that have been directed to two different locations, three macroscopic spatio-temporal structures are found that capture around 90% of tennis shot variance. Inherent within these dynamic structures are reliable movements that are indicative of particular shot directions. In the other dynamic structures that capture less of the tennis stroke variance trial-to-trial variability is predominately found. In resulting psychophysical experiments where information for anticipation of shot direction in these dynamic structures have been perturbed; expert tennis players can be differentiated from novices when perceiving linear combinations of these three dynamic structures, the ability to anticipate emerges when intermediate and less skilled players view the information for anticipation in these dynamic structures but not when only information for anticipation in movement amplitude is present, and finally global information pick up may offer a reliable way in which these dynamic structures can be accurately perceived.