Select one stick
Choose left or right. The test resets whenever you change sticks so data from separate mechanisms is never mixed.
72-direction outer-edge analysis
Measure joystick circularity, not just draw a circle.
Rotate one stick around its full edge to capture the outer envelope. Gamepad Bench reports angular coverage, roundness error, maximum and minimum radius, directional asymmetry and the actual trajectory.
Live trajectory5° direction binsLocal processing
Outer-edge traceLeft stick
PausedThe thin circle marks radius 1.00. The bright trace is the maximum radius captured in each 5° direction.
Roundness error—mean absolute radial deviation
Maximum radius—largest captured outer value
Minimum radius—smallest covered direction
Directional asymmetry—requires 80% coverage
01
Repeatable method
Trace the mechanical edge slowly, not a freehand circle.
The test needs the stick pushed against its physical gate. Moving inside the edge measures your hand path instead of the controller's available outer range.
Rotate against the edge
Push fully outward and make two or three slow rotations. Reverse direction once to reduce gaps caused by hand speed or corner transitions.
Stop above 80% coverage
Review the trace and metrics only after most directions are captured. Clear and repeat under the same connection when comparing changes.
Metric definitions
Each number answers a different question.
No single circularity score proves a stick is good or bad. Response curves, square ranges and firmware normalization differ by controller.
02
Interpret the shape
Joystick circularity test FAQ
What does joystick circularity error mean?
This page divides the full rotation into 72 angular bins, keeps the largest radius captured in each direction and reports the mean absolute deviation from the average radius. It describes radial consistency, not aiming accuracy or input latency.
Why must I reach at least 80% angular coverage?
Maximum, minimum and directional comparisons are misleading when large parts of the circle were never sampled. The page withholds final metrics until at least 80% of the 5-degree direction bins contain an outer-edge sample.
Is a radius above 100% automatically bad?
No. Browser axes can form a square range where diagonal magnitude exceeds 1.00, and firmware may clamp or reshape the output. Compare shape consistency and behavior in the games you use rather than treating one universal percentage as a pass/fail limit.