The predictive outfielder: a critical test across gravities

Abstract

Intercepting moving targets, like fly balls, is a common challenge faced by several species. Historically, models attempting to explain this behavior in humans have relied on optical variables alone. Such models, while insightful, fall short in several respects, particularly in their lack of predictive capabilities. This absence of prediction limits the ability to plan movements or compensate for inherent sensorimotor delays. Moreover, these traditional models often imply that an outfielder must maintain a constant gaze on the target throughout to achieve successful interception. In this study, we present a new model that continuously updates its predictions, not just on the immediate trajectory of the ball, but also on its eventual landing position in the 3D scene and remaining flight time based on the outfielders real time movements. A distinct feature is the models adaptability to different gravitational scenarios, making its predictions inherently tailored to specific environmental conditions. By actively integrating gravity, our model produces trajectory predictions that can be validated against actual paths, providing a significant departure from previous models. To compare our model to the traditional ones, we conducted experiments within a virtual reality setting, strategically varying simulated gravity among other parameters. This gravity variation yielded qualitatively distinct predictions between error-nulling optical-based heuristics and our model. The trajectories, kinematic patterns and timing responses produced by participants were in good agreement with the predictions of our proposed model, suggesting a paradigm shift in our understanding of interceptive actions. Significance statementCatching a moving target, a challenge consistently faced across various species, exemplifies the complex interplay between perception, prediction, and motor action in dynamic environments. Prevailing models have been largely rooted in optical cues, often overlooking the predictive capacities essential for understanding real-world human behaviors and sidestepping crucial physical variables such as gravity. Our research introduces a novel model that emphasizes both the predictive component and the broader gravitational dynamics allowing for a more holistic understanding of interception tasks. This innovative approach not only holds implications for refining existing models of interception but also carries broader significance for training platforms, ensuring relevance across diverse settings, from Earth to altered gravity environments.