Press a controller button after connecting it. If inputs remain absent, confirm that the computer recognises the controller and try another supported connection. An empty browser result does not establish that the controller is broken. Use the gamepad test to check whether buttons and axes arrive.
Part of the Controller and gamepad test. Run the full gamepad test first for a complete device check.
No controller appears The result changes after each release The camera moves in only one game
Stick drift test
Use this stick drift test to record where both controller sticks rest without being touched. Compare the resting offset and suggested dead zone, then verify the symptom in the game where it occurs.
Also known as: stick drift test, stick drift tester, controller stick drift test, controller drift test, controller deadzone test, joycon drift test. Free, online, and no download required.
No permission neededNothing uploaded
Connect a controller, press a button to expose its inputs, then leave both sticks untouched.
01
What the stick drift test checks
| Check | How it is measured | Fails when |
|---|---|---|
| Resting offset | Distance from centre of the mean X and Y values received during the rest sample. | A stick reaches this page's 0.10 review threshold, which is not a manufacturer specification. |
| Radial noise | Median absolute deviation of the recorded stick radii. | Movement during capture makes the sample unsuitable for a resting comparison. |
| Suggested dead zone | Resting offset plus three times radial noise, rounded upward to two decimal places. | The game uses a different scale or dead-zone shape, so the suggestion needs in-game verification. |
How to run the stick drift test
- Connect the controller and press any button so the browser can detect it.
- Select a 5 s or 10 s rest duration and place the controller on a stable surface.
- Start the rest check and keep both sticks untouched until it finishes.
- Compare left and right offset, noise and suggested dead zone.
- Repeat after moving and releasing each stick, then check the affected game.
If something failed
Move the stick, release it fully and repeat the complete rest sample without touching it. Record whether the offset returns in the same direction. Do not mix movement samples with untouched samples when choosing a dead zone; the two answer different questions.
Compare that game's input settings and active controller profile with another game. A browser rest check can look quiet while a game reacts to smaller offsets. Change one dead-zone setting at a time and retest both unwanted movement and deliberate small movements.
02
How to read the result
How to run a stick drift test
Begin with a controller that the browser already detects. Put it down before starting the timer and avoid resting your thumb on either stick. The selected five- or ten-second capture is meant to measure the untouched position, not the path you draw while moving the stick. Movement during this period changes the mean and makes the rest result misleading.
After the first capture, move each stick gently away from centre and let it return. Start a fresh rest capture and compare the direction and size of the offset. A similar result after separate releases is more useful than one sample taken while handling the controller. Keep the connection and software profile unchanged during the comparison.
Drift thresholds: 0.05 and 0.10
This stick drift tester marks offsets below 0.05 as below its warning threshold, values from 0.05 to below 0.10 for review, and values of 0.10 or more as reaching its failure threshold. These are this site's diagnostic rules. They are not universal acceptance limits, a manufacturer's tolerance or a promise that a particular game will ignore the input.
The offset is calculated from the mean X and Y coordinates. Noise describes variation in radial distance, so it does not describe every possible change in direction. Read both numbers and repeat the capture when the visual movement looks inconsistent with the summary. A controller can pass this site's threshold and still produce unwanted input in a game with different settings.
Choose a dead zone and verify the trade-off
The displayed suggestion adds three times the measured radial noise to the resting offset and rounds upward to two decimal places. It is a starting point for comparison. A game may use a different scale, apply separate horizontal and vertical settings, or combine another software profile with its own controls. Do not copy the number blindly into an unrelated setting.
Increasing a dead zone can suppress unwanted movement but can also remove small intentional movements near centre. Test both effects in the affected game. If comfortable aiming or steering requires a large change, record the behaviour and consult the controller's support guidance. The controller stick drift guide explains how to organise the next checks.
Outer range and circularity
This rest check answers whether an untouched stick stays near centre. It does not ask you to rotate the stick during capture. Use the gamepad test for a separate visual check of movement and outer range. A full sweep and an untouched rest sample need separate observations because combining them makes the mean offset difficult to interpret.
An unusual outer shape does not by itself prove a centring fault. The page receives coordinates after driver and controller processing, so the visible range is not a photograph of the physical sensor. Describe resting movement separately from restricted travel when reporting the issue; those symptoms may need different investigations.
Measurement limits
The browser receives mapped Gamepad API values, not raw sensor voltages or access to calibration memory. Firmware and software can apply their own transformations before the page sees input. This controller drift test cannot recalibrate the device, write a repair or certify its mechanical condition.
A missing controller or missing axis is an incomplete test, not a drift measurement. Support depends on the controller, connection and browser together. Keep results tied to that setup, and verify changes in the application where you noticed the problem instead of treating a quiet browser sample as a complete repair.
03
Stick drift test questions
How long should I leave the controller untouched during a stick drift test?
Leave it untouched for the full selected duration, either five or ten seconds. Put it on a stable surface before starting and keep your thumbs clear. If you bump a stick during capture, discard that sample and restart. Use the same duration when comparing two connections or controller profiles.
What resting value counts as stick drift?
This page uses 0.05 as its warning threshold and 0.10 as its failure threshold for mean resting offset. Those are local review rules, not universal hardware tolerances. The practical question is whether the game's input settings allow unwanted movement, so compare its behaviour with the recorded direction and offset.
Why does the stick drift tester suggest a dead zone?
A dead zone can make a game ignore small inputs near centre. The suggestion combines the measured offset with radial noise to provide a comparison point. It does not repair the stick. Check the game's scale and test small intentional movements too, because a larger dead zone can reduce fine control.
Can this test check Joy-Con drift?
Only when your computer and browser expose the connected controller and its stick axes. Verify that the visible inputs respond before starting a rest check. If an axis is absent or mapped unexpectedly, the result is incomplete. This page cannot infer the condition of a Joy-Con it cannot read.
My stick drifts in only one game. What does that mean?
The game may handle centre offsets differently from the browser or another game. Compare the selected controller profile and dead-zone settings, then repeat the same scene with one setting changed. The browser result helps describe the input it received, but cannot inspect or override the game's own input processing.