Acceleration disrupts a stable device-to-screen mapping

What it is

Acceleration disrupts mapping consistency. Under fixed gain, one device displacement always yields the same screen displacement; under acceleration, speed also determines the result. People cannot reproduce an endpoint by moving “that far” alone, but must control movement tempo too.

Why it happens

A stable mapping supports predictable visual-motor calibration. Velocity-dependent gain adds a second variable: tiny speed changes alter the result despite equal amplitude. This compresses long transport but increases errors in muscle-memory prediction, especially for tasks requiring repeated trajectories.

Skilled control rests on an internal predictive model: the brain treats "hand movement" and "cursor displacement" as a fixed function it repeatedly calibrates. Under fixed gain that function has one free variable — displacement. An acceleration curve turns speed into a second free variable, so the model must remember both "how far" and "how fast" to predict the outcome, which slows learning and makes the model more fragile to any change in context. Worse, once that internal model is built, it can be invalidated by a change nobody sees — an OS update, a driver change, a DPI setting — and people just feel that "it doesn't feel right anymore" without being able to say what changed, because the change happened in the curve's parameters, not in some visible setting.

Where it stops holding

Consistency is not the only goal: acceleration can help one-off discrete choices across a large display, and experts can learn a fixed curve. The concern is not dynamics alone, but an unpredictable curve, cross-platform inconsistency, or a task requiring reproducible amplitude.

Competitive aiming disables acceleration outright precisely because players need to reproduce the same sensitivity across different events and different machines; pixel-level retouching and precise alignment tasks depend on the same reproducibility — introduce velocity-dependent gain and the same editing gesture lands on a different pixel each time it is repeated.

Applying it

• Do not silently replace or reset an acceleration curve within a continuous task.
• Pair trajectory-reproducible functions with an available fixed mapping.
• Have people repeat equal-distance actions at different speeds, examining whether endpoint spread impedes the task.
• Before and after a version or driver update, run the same displacement-versus-speed matrix as a regression check, confirming the curve parameters were not silently changed.

Related

• Same group: C1.04.1 Acceleration dynamically changes gain with movement speed · C1.04.3 Acceleration is a net loss for tracing and drawing · C1.04.4 Acceleration must be disableable
• Nearby: C1.03 Control-display gain · C1.05 Cursor hotspots and hit testing
• Search terms: mapping consistency · muscle memory · pointer acceleration