Input Lab · five-second center sample

Stick Drift Test

Measure the observed center offset, maximum movement and stability of both analog sticks with a repeatable browser test.

Center stability analyzer Gamepad API · local session
Connect a controller, then press any button.
Waiting for a controller. Some browsers expose a gamepad only after a button press.

Left stick

X 0.000Y 0.000

Right stick

X 0.000Y 0.000

Buttons & triggers

B0B1B2B3B4B5B6B7B8B9B10B11B12B13B14B15B16B17B18B19
Left average
Left maximum
Right average
Right maximum

How to use: Stick Drift Test

Place the controller on a stable surface and release both sticks.

Press “Run center sample” and do not touch the controller for five seconds.

Review both sticks, repeat once, then compare in another browser or connection before deciding on repair.

What the result actually measures

The test records both stick axes while the controller rests untouched. It summarizes average center distance, maximum observed distance and variation instead of treating one noisy frame as a diagnosis.

Measurement boundary: The result reflects the current browser, connection, driver and controller state. Built-in game deadzones can hide drift in play, while calibration utilities can change the values the browser receives.

How to interpret the output

Under 3%

The observed center is tight in this sample. Re-run if you still experience movement in a game with a very small deadzone.

3–8%

A minor offset was observed. A modest software deadzone may absorb it, but watch whether the direction is stable.

Above 8%

The offset is more likely to affect low-deadzone games. Confirm it with a second sample and a wired/alternate connection.

Troubleshooting unexpected results

Result changes every run

Wireless interference, a moving cable, worn mechanics or resting a finger on the stick can create an unstable sample.

Game moves but test is centered

The game may use another input, a different calibration layer or settings such as gyro. Compare the in-game deadzone and controller profile.

Frequently asked questions

What drift percentage is acceptable?

There is no universal threshold. Competitive games with tiny deadzones reveal smaller offsets than games with larger built-in deadzones.

Should I set my deadzone to the maximum reading?

Not automatically. Start slightly above a repeatable center offset, then confirm that slow intentional movement still registers.

Does the test move or vibrate the controller?

No. The center sample only reads exposed axes while the controller rests.