Predicting Error Rates in Pointing Regardless of Target Motion

In a pointing task with time constraints, it was only possible to predict the user's error rate when pointing to a stationary target. This study presents a novel model for predicting pointing error rates regardless of the target motion. The model assumes that in the last submovement of the pointing trajectory just before the click, the timing to activate the button is anticipated by the user's internal clock decoding the temporal cues present in the relative movement between the cursor and the target. Then, based on the recent theory of temporal pointing, the model can predict the user's pointing error rate with a high R2 for both stationary (0.993) and moving targets (0.986) by analyzing the kinematic characteristics of the last submovement. In addition, empirical parameters obtained from the model fit succeeded in revealing differences in the cognitive characteristics of experts and novices in first-person shooter games.