Stick drift test

Stick drift is when the camera or your character moves while your thumbs are off the sticks. This stick drift test measures it instead of asking you to eyeball it: rest the controller on a table, start the capture, and for three seconds the page records where each stick sits when nobody is touching it. The result is the resting offset of each stick, in the same units a game reads, placed beside the dead zones that vendors actually publish, so you can see whether the drift you have will show up in play. A second pass, rolling each stick around its rim, checks that both sticks still reach full deflection. Nothing is uploaded.

How to run the test

Connect the controller by cable or Bluetooth and press any button on it, so the browser lets the page see it. Put the controller down on a flat surface and take your hands off both sticks. Start the rest capture. For three seconds the two plots show each stick's position magnified four times around centre, with the documented dead zones drawn as dashed rings, and the page records every reading. If a stick moves more than half its travel during those three seconds the capture is refused with the reason, because a resting offset measured while a thumb was on the stick is not a resting offset.

When the rest capture is accepted the range pass unlocks. Pick the controller up and roll each stick slowly all the way around its rim, pressing against the edge, for six seconds. The plots switch to full range and the result draws the furthest reach at every angle. Skip this pass if you only care about drift; the rest result stands on its own.

Reading the result

The first line for each stick says where it rests, as a direction and a distance from centre, where 1.0 is full deflection. A stick that reads 0.06 to the right and 0.03 up is resting about six percent of its travel off centre. Under that line the page compares the offset with the dead zone values Microsoft publishes for its XInput controllers: 0.24 for the left stick, 0.27 for the right. A dead zone is the small circle around centre that a game ignores, so an offset inside it does not move anything in a game using those defaults, and an offset outside it does. Sony publishes no equivalent figures, so none are listed; a table of invented numbers would be worse than none.

The page does not call a stick faulty. Whether a given offset matters depends on the game's dead zone, which you can often change, and on how much worse it gets over the coming months, which this page cannot know. What it gives you is the number, the comparison, and the numbers behind it: the mean position on each axis, the furthest the stick wandered during the capture, the spread around the mean, and the reporting step of the stick, which is about 1/127 of full deflection on an 8-bit stick. This stick drift tester reports what it measured and refuses when it could not trust the capture.

The range result is simpler. It shows the furthest and shortest reach around the rim and whether the stick reaches full deflection. Some controllers report readings past 1.0 on the diagonals; that is a square reporting range, a shape rather than a fault, and the result says so when it sees it.

What causes controller drift

Most thumbsticks read position with two small potentiometers, one per axis, and a wiper that slides across a carbon track. The track wears where the wiper rests, dust and skin oil collect on it, and the return spring tires, so the resting position creeps and the readings around centre get noisy. Controller drift usually starts small and grows. Hall effect sticks measure position magnetically with no contact to wear, which is why replacement modules built on them have become a common repair.

What to do next

Start with the game's own dead zone setting if it has one; raising it hides drift inside the range the game ignores. Recalibrating in the console or the controller's desktop software can recentre a small offset for a while. The stick base can be cleaned, though results vary and this page makes no promises about it. Beyond that the options are a stick module replacement, a warranty claim if the controller is still covered, or a new controller. Whichever you choose, run this test again afterwards: a number you measured before and after is the only way to know a repair worked.

What the browser cannot see

A browser reads a copy of the controller's state on its own schedule; it never receives the controller's reports. That is fine for a resting offset, which does not change quickly, and it is why this page reports no latency and no polling rate. The gamepad tester reads the same controller for buttons, triggers and live stick movement, and the methodology page states exactly what has and has not been validated on physical controllers.

Frequently asked questions

How much stick drift is normal?

New controllers commonly rest a few hundredths off centre; a reading around 0.02 to 0.05 is ordinary. What matters is where the offset sits against the game's dead zone. Microsoft's published XInput dead zones are 0.24 for the left stick and 0.27 for the right, so an offset well inside those will not move a camera in a game using them. No published figure exists for PlayStation controllers, so the page lists none.

Can stick drift be fixed?

Sometimes, for a while. Raising the in-game dead zone hides it, recalibration can recentre a small offset, and cleaning the stick base helps some controllers. A worn potentiometer track does not heal, so the lasting fixes are a stick module replacement, ideally a Hall effect module, or a new controller. Measure before and after so you know.

Why did the rest capture refuse?

Either a stick moved more than half its travel during the three seconds, or the capture was too short to read. Rest the controller on a table, keep your hands off both sticks, and start again. A refusal is the page declining to print a resting offset it cannot stand behind, not a verdict on the controller.

Does this work with PS5 and Xbox controllers?

Yes, and with any controller the browser recognises, over a cable or Bluetooth, on a computer, a phone or a tablet. Press any button on it first; browsers hide controllers from pages until one is pressed. The dead zone comparison uses Microsoft's published values because those are the ones that exist; the measurement itself is the same for every controller.

Does this upload anything?

No. The stick positions are read in your browser and nothing about your session leaves your machine. The "Copy result" button puts a plain-text summary on your clipboard so you can paste it into a forum post or a support ticket, and that is the only place the numbers go.