A continuous motor task was reexamined to increase the fidelity of a previous model of the task. Previous modeling work was successful in matching qualitative performance, but not all aspects of the movement profile. When a new set of subjects were brought into the lab and eye-tracked, their motion data informed further modifications to the model’s performance strategy as well as changes to the ACT-R architecture. This in turn produced a model with a higher fidelity movement profile. The change in strategy modeled identified capabilities that could be implemented in future work as well as brought new inspiration to strategy refinement on this task.