Anticipatory Judgment of Tennis Serve Displayed by the Human Models of Computer Graphics (CG)

Abstract

This study examined the perceptual effect of color and /or figuration information in human models created using computer graphics (CG) on anticipatory judgments of tennis serves. Fourteen skilled and fourteen novice tennis players attempted to anticipate serve direction by observing one live-action image and four CG animations (polygon, shadow, stick, and point-light) edited in accordance with a temporal occlusion paradigm. The dependent variable was the correct responses of serve directions (anticipation accuracy). The shadow, stick, and point-light models were simplified variations of the color and/or figuration information within a polygon model. At an occlusion of 30 ms before the moment of racket-ball impact, the skilled player's anticipation accuracy was superior to the chance level (50%) in the live-action and polygon model conditions, but not in the other three model conditions. In contrast with the results of skilled players, the novice player's anticipation accuracy did not exceed the chance level in all the conditions (model and occlusion). Moreover, the skilled players made more accurate judgments than their novice counterparts in the live-action and polygon model conditions. These results suggested that the skilled players used the color and figuration information in human models as visual information when they picked up the essential kinematic information from opponent's movement patterns. In addition, there were no significant differences in the skilled player's anticipation accuracy between the live-action and polygon model conditions. This led to the conclusion that our polygon model had sufficient quality to evaluate the superior anticipatory performance of skilled tennis players.