Predictive and Reactive Control During Interception

<b>Background/Objectives:</b> Successful interception of moving targets requires combining predictive control, which anticipates future target states, and reactive control, which compensates for ongoing sensory discrepancies. How these components evolve over time and are distributed across gaze and manual behavior remains unclear. We aimed to explore the time-resolved dynamics of predictive control during continuous interception and to dissociate eye and hand contributions. <b>Methods:</b> Human participants intercepted a moving target in a two-dimensional arena using a joystick while eye movements were recorded. Target speed was systematically varied, and visual information was selectively reduced by occluding either the target or the user-controlled cursor. Predictive control was assessed using two complementary metrics: a geometric strategy index capturing moment-to-moment spatial lead or lag relative to target motion, applied separately to gaze and manual trajectories, and root mean square error (RMSE) computed relative to current and forward-shifted target positions to quantify predictive alignment. <b>Results:</b> Successful interception was characterized by structured, speed-dependent transitions between predictive and reactive control rather than a fixed strategy. Predictive alignment emerged early and was dynamically reweighted as temporal constraints increased. Gaze and manual behavior showed complementary but partially dissociable predictive signatures. Occluding the target decreased predictive alignment, whereas occluding the user-controlled cursor had comparatively minor effects, indicating strong reliance on internal state estimation rather than continuous visual feedback of the effector. <b>Conclusions:</b> Predictive and reactive control are continuously and dynamically reweighted during interception. Their interaction unfolds within single trials and depends on target dynamics and sensory availability. These findings provide quantitative evidence for time-resolved coordination between anticipatory and feedback-driven control mechanisms in goal-directed behavior.