Tracking Accuracy

Tracking accuracy describes performance on continuous manual control tasks, where the goal is not a single discrete response but sustained correspondence between an operator's input and a moving target over an extended period. Poulton (1974) synthesized decades of tracking-skill research, distinguishing compensatory tracking, where only the error signal is visible, from pursuit tracking, where both the target and the response cursor are visible, and showing that pursuit tracking reliably produces tighter accuracy because the operator can see target motion directly rather than inferring it from error alone. Tracking accuracy degrades predictably as target velocity and unpredictability increase, a signature Poulton used to characterize an operator's control bandwidth.

Reference: Poulton, 1974. The NeuroRank implementation holds the canonical form and scales interference via task-irrelevant stimulus density.

1. Keeping a crosshair on a strafing target for the full duration of a sustained spray rather than only landing the first shot.
2. Following a jungler's pathing on the minimap continuously through a gank rather than checking it in isolated glances.
3. Maintaining a tight follow on an evasive target in a hero shooter's tracking-based weapons, where damage accrues only while the reticle stays on target.

NeuroRank's Tracking module is a pursuit-tracking task by Poulton's definition: both the target and the cursor are visible throughout. timeOnTarget and avgNormDistance jointly produce the Tracking Accuracy dimension score, capturing both binary on-target time and continuous positional error.