Abstract
Eye movements are often investigated with simple and controlled stimuli, under the assumption that results can be generalized to the real world. We previously tested this by comparing tracking of the same trajectories, either as an isolated target or embedded in a naturalistic video, the movement of a puck in an ice hockey game. We observed that the oculomotor system was able to leverage the contextual cues available in the naturalistic scene to produce highly predictive eye movements. Here, we wanted to asses which factors are critical for achieving this predictive advantage by manipulating (1) expertise, (2) the amount of available peripheral information, and (3) contextual and kinematic cues. When peripheral information was manipulated by decreasing the visible area of the video, the predictions of all observers decreased. However, expert ice hockey fans were consistently more accurate and better in predicting than novices and also benefitted more by additional peripheral information. Artificial contextual cues about the position of the players did not lead to a predictive advantage, whereas impairing the causal structure of kinematic cues by playing the video in reverse led to a severe impairment. Watching the video upside down, potentially making the kinematic cues more difficult, did not significantly affect predictive behavior. Together, these results demonstrate that when contextual information is available in naturalistic scenes, the oculomotor system is successfully integrating them and is not only relying on low-level information about the target trajectory. Critical factors seem to be the amount of available information, expertise with the stimuli and especially for our ice hockey videos the availability of intact kinematic cues for player movements to predict what is going to happen next.