Author(s): MURAKAMI, H., NORIMASA, Y., Institution: KINJO UNIVERSITY, Country: JAPAN, Abstract-ID: 1228

Introduction
Fitts’ law describes the speed–accuracy trade-off in goal-directed movements as a linear relationship between movement time (MT) and the index of difficulty [ID=log₂(2A/W)]. Although this framework has been widely applied in both laboratory and applied contexts, it is typically tested under stationary conditions with fixed target distances and widths. In contrast, many sport situations require rapid actions under dynamically changing and partially unpredictable constraints. This study extends Fitts’ paradigm to such uncertainty by employing a dynamic tapping task in which target distance changes during continuous alternation, comparing deterministic and random change patterns.
Methods
A Fitts-type alternating tapping task was implemented in PsychoPy and presented on an iPad mini. One participant tapped between two square targets (width W= 30 pixels) with a stylus for 15s trials, while the distance A between the targets was updated after each tap. Four conditions were tested: (1)Fixed, six conventional ID levels (1–6) presented in separate trials; (2)Increasing, in which A increased stepwise, producing a monotonic increase in ID; (3)Decreasing, in which A decreased stepwise; and (4)Random, in which A (thus ID) changed pseudo-randomly within the same range. For each dynamic condition, the participant completed 100 trials, yielding 2000 MT observations across the three dynamic conditions. For every tap, MT (inter-tap interval) and ID were computed, and a quadratic model MT=aID²+bID+c was fitted to each condition.
Results
In the Fixed control condition, MT increased monotonically with ID and was well captured by the quadratic fit (R²≈0.88), confirming that the present apparatus successfully reproduced the fundamental speed–accuracy trade-off characteristic of Fitts’ law. Similar monotonic trends were observed in the Increasing (R²≈0.52) and Decreasing (R²≈0.96) conditions. Despite trial-to-trial changes in A, the overall MT–ID function remained close to the conventional pattern, suggesting that participants could construct predictive, feedforward strategies when the direction of change was known. In striking contrast, the Random condition showed a markedly poorer fit (R²≈0.17) and a non-monotonic, broadly U-shaped MT–ID profile, with prolonged MT even in low ID ranges.
Conclusion
These results indicate that, under random distance changes, physical difficulty (ID) alone does not fully determine performance. Environmental uncertainty appears to increase effective difficulty and induce flexible, non-deterministic control, in which multiple modes may coexist before target onset. The dynamic Fitts task proposed here provides a tool for probing such “choice-like” or superposition-like aspects of motor control. Ongoing research extends this pilot to multiple participants and additional kinematic and temporal measures to develop a probabilistic framework that integrates classical Fitts’ law with uncertainty-driven control in sport environments.