Ernst-Joachim Hossner

Extensive research suggests that humans integrate sensory information and prior expectations in a Bayesian manner to reduce uncertainty in perception and action. However, while Bayesian integration provides a powerful explanatory framework, the question remains as to what extent it explains human behaviour in naturalistic situations, including more complex movements and distributions. Here, we examine whether humans can learn bimodal priors in a complex sensorimotor task: returning tennis serves. Participants returned serves in an immersive virtual reality setup with realistic movements and spatiotemporal task demands matching those in real tennis. The location of the opponent’s serves followed a bimodal distribution. We manipulated visual uncertainty through three levels of ball speeds: slow, moderate, and fast. After extensive exposure to the opponent’s serves, participants’ movements were biased by the bimodal prior distribution. As predicted by Bayesian theory, the magnitude of the bias depends on visual uncertainty. Additionally, our data indicate that participants’ movements in this complex task were not only biased by prior expectations but also by biomechanical constraints and associated motor costs. Intriguingly, an explicit knowledge test after the experiment revealed that, despite incorporating prior knowledge of the opponent’s serve distribution into their behaviour, participants were not explicitly aware of the pattern. Our results show that humans can implicitly learn and utilise bimodal priors in complex sensorimotor behaviour.

Handling motor noise is fundamental to successful sensorimotor behavior, especially in high-risk situations. Research using finger-pointing tasks shows that humans account for motor noise and costs of potential outcomes in movement planning. However, does this mechanism generalize to more complex movement tasks? Here, we investigate sensorimotor behavior under risk in throwing across three experiments with 20 participants each. Their task was to throw balls at a target circle, partially overlapped by a penalty circle. This task challenged participants to find strategies that trade off potential penalties and rewards. In the experiments, penalty magnitude and the distance between the circles were manipulated. We measured the location of their final gaze fixation before movement—as an indicator of their planned aiming point—and the ball’s impact location. Without penalty, the final gaze fixation and the ball’s impact location were both centered on the target. In the penalty condition, the location of the participants’ final gaze fixations and the ball’s impact shifted away from the penalty circle, with larger shifts for higher penalties and smaller distances. Interestingly, the shifts in the ball’s impact locations were not only larger (“more conservative”) but also closer to the statistically optimal (expected gain-maximizing) location compared to the fixated aim points. Movement trajectory analyses show that, in penalty conditions, the shifts away from the penalty zone increased until the final phases of the movement. These results suggest that risk evaluation is not completed in a pre-movement planning phase but is further optimized during movement execution. NEW & NOTEWORTHY We extend the study of sensorimotor behavior under risk from simple finger-pointing movements (Trommershäuser et al., 2008) to a complex throwing task in virtual reality. Our results suggest that, in complex sensorimotor behavior, risk evaluation of potential movements is not confined to a cognitive planning phase before movement but is optimized in action, with the motor system continuously biasing competing action options toward regions of higher expected rewards.