How These Mouse Tests Work
Useful evidence starts with a clear method and an honest limit.
Mouse button response score
The required buttons are primary, secondary and middle. Five complete cycles each are needed for a score; ten clean cycles each enable Excellent. Score = 100 × (1 − (short repeat signals + unmatched events) / complete cycles), rounded and clamped to 0–100. Any detected signal produces Suspected issue even if the numeric score is high. Side buttons are displayed but are optional.
Paced double click test
Make twenty single clicks on the selected button with about half a second between clicks. A full cycle followed by another press within the selected 80–250 ms interval is flagged. The default is 150 ms. Timing cannot reveal user intent; fast intentional clicks can look identical. The score formula matches the button test. These thresholds are heuristics, not manufacturer tolerances.
Directional scroll test
Receive at least ten events during the Up phase and ten during Down. Score = 100 × correctly directed events / all received events. At least forty total events without a direction signal enable Excellent. Deltas retain their reported pixel, line or page unit; no physical notch count or hardware polling rate is claimed.
Physical-distance mouse DPI estimate
Request pointer lock with unadjustedMovement:true using the Promise API. Use a physical ruler on the desk. Estimate = absolute signed horizontal movement / physical inches; three valid passes produce their median. Reject a pass with no movement, under 50 horizontal units, signed displacement under 95% of horizontal path, or over 20% vertical path. Browser updates may be coalesced, so event count is not used as hardware sampling evidence. The distance must be 1–12 inches. Browser and driver scaling can still affect units.
Mouse DPI repeatability score
Spread = 100 × (highest − lowest estimate) / median estimate. Consistency = max(0, round(100 − spread)). Up to 5% spread is Excellent; up to 15% is Normal; larger spread asks for a repeat measurement. These are transparent product bands, not verified sensor-health limits. High/low mouse DPI does not change the grade.
What the browser cannot establish
These tests cannot read firmware settings, inspect internal switches, identify your device model, distinguish a mouse from every trackpad, or certify hardware health. Incomplete or cancelled samples have no score and are not stored as completed results. Compare ports, applications and computers when investigating a reproducible symptom.
Related mouse tests
Sources and test limits
Methods are browser observations, not hardware certification. See the scoring formulas and limitations.
Return to Mouse Test to start another check.