The practical question: controller circularity test. This field note answers it through a defined workflow rather than a copied setting list.
Circularity is a shape description, not a quality grade
A stick sweep plots the X/Y values delivered while the mechanism follows its physical gate. Some gates are circular, some are not, and firmware can rescale diagonals. A trace that departs from a mathematical circle is therefore not automatically defective.
The useful question is whether the same direction clips, stretches or changes across repeated slow sweeps. Keymap divides the path into 24 angular sectors and keeps the farthest radius in each sector. Coverage and opposite-sector differences make repeatable shortages easier to see without hiding the raw path.
Protocol checklist
- Keep contact with the physical gate
- Use two slow full rotations
- Repeat in the opposite direction
Read coverage, outer P95 and symmetry together
Coverage counts sectors that reached at least 75 percent browser-reported radius. Empty sectors can be caused by moving too quickly, so a low result must be repeated before it means anything.
Outer P95 describes the high range without letting one extreme frame dominate. Symmetry error compares opposite sectors. A weak right sector that repeats in both clockwise and counter-clockwise runs is more actionable than one cosmetic bump.
Protocol checklist
- Compare the same stick and connection
- Check whether the weak sector repeats
- Confirm the directional effect in game
Pair the sweep with a recenter sample
Outer range and center drift are different problems. A controller can return cleanly to center but fail to reach one edge, or reach every edge and rest off-center.
Run the untouched center stage before the sweep and again after it. If the resting point changes depending on the direction from which the stick returns, record that recenter behavior separately from the circular path.
Measurement boundary: Browser mappings, gate geometry and firmware processing differ. A sweep cannot certify manufacturing tolerances or predict every game response curve.
Use the result, then repeat it
Keep the device, platform, connection, browser or game build and the previous setting beside the result. If the observation changes, restore the baseline and change one variable. That procedure is more valuable than a precise-looking number without context.
