Stick Drift Test
Let go of both sticks, run a 10-second hold-still test, and get a real number instead of a feeling: max and mean deviation per stick, a none/minor/repair verdict, and a dead zone to type into your game's settings.
No permissions. A connected controller; press any button to wake it up.
No controller detected yet
Connect a controller by USB or Bluetooth, then press any button. Browsers hide gamepads until you interact with one.
What causes stick drift, and how to fix it
Most drift comes from the stick's internal potentiometer wearing down after months or years of use -- the resistive material thins out at whatever center point you rest on most, so the stick reports a false position even when it's not touched. Dust and debris under the stick module cause the same symptom without any real wear, and a drop or hard impact can bend the housing enough to throw off the center reading.
For minor drift, an in-game dead zone set to this test's recommendation is a free, instant fix. For dust-related drift, a short burst of electrical contact cleaner sprayed into the gap around the stick, followed by working the stick through its full range a few dozen times, clears many cases. For a stick that reads "needs repair," the real fix is a new stick module or potentiometer -- through the manufacturer if it's still under warranty, or a third-party repair kit otherwise; some replacement modules use Hall-effect sensors, which read magnetic position instead of resistance and don't wear the same way. If you open the controller yourself, unplug it first and check whether doing so voids your warranty before you start.
Want the full picture, not just the sticks? The controller tester checks every button and trigger on the same pad.
How the verdict is calculated
While the test runs, each stick's distance from center (√(x² + y²)) is sampled on every animation frame. After 10 seconds, that gives a mean and a max deviation per stick. If the max stays at or below your threshold (0.05 by default), the verdict is none. Above the threshold but within 3× it, the verdict is minor -- a dead zone will hide it. Beyond 3× the threshold, the verdict is repair. The recommended dead zone is the larger of your max deviation or threshold, padded 50% and rounded to the nearest 0.01, so brief run-to-run variance doesn't leave residual creep.
Worked example: a run with threshold 0.05 comes back left stick mean 0.020 / max 0.040 (0.040 ≤ 0.05, so none) and right stick mean 0.060 / max 0.090 (0.090 is above 0.05 but at or below 0.15, so minor). The recommended dead zone for the right stick is max(0.090, 0.05) × 1.5 = 0.135, rounded to 0.14.
Frequently Asked Questions
What counts as drift versus normal resting noise?
Every analog stick reports a tiny amount of noise even when it's healthy, usually well under 0.02 at rest. Drift is when that number stays elevated instead of settling near zero, or when the stick visibly creeps in a direction on its own. The 10-second hold-still test separates the two by sampling the whole window instead of a single reading.
Why did I get "minor" on one run and "none" on the next?
A stick right at the edge of a threshold can flip verdicts between runs, especially on a worn but not badly worn potentiometer. Run the test two or three times and look at the mean deviation, not just one verdict -- a mean that's consistently rising over several runs matters more than a single borderline result.
What dead zone should I actually set in my game's settings?
Use the number this test recommends after a run showing minor or repair -- it's your measured max deviation padded by 50% so small run-to-run variance doesn't leave residual creep. Setting it higher than needed just wastes stick travel at the center; setting it lower than your real drift brings the creep back.
Does a dead zone actually fix the drift?
It masks the symptom, not the cause -- the stick still drifts, but the game ignores movement below the dead-zone radius so you don't feel it. That's a fine permanent fix for minor drift. For a stick that reads "needs repair," a dead zone large enough to hide it usually also eats real precision near the center, which is worse for aiming than living with a smaller dead zone and getting the part replaced.
My controller shows no drift here but still feels off in-game.
Some drift only appears under load -- heat from extended play, or pressure from how you hold the stick -- and won't show up in a cold 10-second test. Run the test again right after a long session. If it's still clean, the issue is more likely game-specific sensitivity settings than the hardware.