Fractal architectures in motor expertise: bridging deterministic and stochastic control

Motor expertise is conventionally attributed to enhanced deterministic control and reduced variability. This study extends that view by demonstrating that expertise reflects fractal organization of movement. Fractal analysis of 200 trajectory data points from a haptic ball-striking task revealed that experts exhibit higher fractal dimensions, broader frequency utilization, and stronger scale-invariant properties than novices, despite superior accuracy. Phase-space analysis identified distinct attractor basins separating the two groups, consistent with qualitatively different control regimes. We interpret these findings through a river network model in which expert motor control integrates deterministic goal pursuit with stochastic exploration, producing adaptive accuracy through structured variability. This fractal architecture suggests that skill acquisition involves the development of multiscale control structures rather than error minimization, with potential implications for training and rehabilitation design.