Skip to the content

Stick drift test

Free stick drift test: leave the controller alone for 5 seconds, then circle each stick. See how far it drifts and save a dated PDF report.

  • Nothing is stored or uploaded
  • Free, no sign-up, nothing to install
  • Chrome, Edge, Firefox and Safari

Waiting for a button press

Press any button on your controller

Plug it in or pair it by Bluetooth first. A browser only shows a controller after one of its buttons has been pressed.

Pressed a button and nothing happened?
  1. Check the cable or the Bluetooth pairing. Some USB cables only charge and carry no data.
  2. Close other programs that may be using the controller, such as a game launcher, then press a button again.
  3. Reload this page with the controller already connected, then press a button.
  4. Try another browser. Chrome, Edge and Firefox all read controllers.

More help: controller not showing up

This test runs in your browser. Nothing is stored or uploaded.

This stick drift test reads both sticks while nobody touches them, then while you turn them in full circles, and tells you plainly whether there’s drift: Okay, Check this or Not working. It works as a stick drift tester for Xbox, PlayStation and Switch Pro controllers, and anything else your browser can read. It takes about a minute, and you can save the result as a dated PDF.

How to test for stick drift

Connect your controller by USB cable or Bluetooth and press any button, so the browser can see it. Then find a flat, steady surface. A desk is better than your lap or a sofa cushion, because the test is about to notice the smallest movement. If the controller doesn’t appear when you press a button, see controller not showing up.

  1. Put the controller down and press Start the test. You get a three-second countdown to let go, then the test reads both sticks for five seconds. Don’t touch the controller or knock the table while it counts.
  2. Pick the controller up and turn each stick slowly round its outer edge, two or three full circles each. A percentage shows how much of each circle you’ve covered.
  3. When both circles are complete, let go of the sticks and the result appears. You can press Show the result at any point, or Skip this step if you only want the hands-off reading.
Step 2 of the stick drift test: a round plot for each stick, part of each outer circle traced in a solid line and the rest dashed. The left reads 77.1% of the circle covered, the right 39.6%.
Step 2: turn each stick round its outer edge until the solid line closes the circle. The percentage under each plot shows how much you’ve covered.

If the controller disconnects during the hands-off reading, start again: those five seconds have to be one unbroken stretch. And if you’re not sure whether you nudged it, run the test twice. Real drift shows up both times.

What your result means

For each stick you get four readings, each a percentage of the stick’s full travel. Only the first one decides whether there’s drift; the others flag things worth a second look.

Resting offset: the number that decides

The resting offset is how far from centre the stick sat, on average, during the five hands-off seconds. A perfect stick would read 0%. A stick resting 12% of the way towards its edge is telling the game you’re pushing it gently, all the time, and that’s what drift is: a camera that creeps, a character who walks off on their own, a menu cursor that won’t stay put.

A scale of resting offset from 0% to 20%: under 5% is Okay, 5% to 10% is Check this, and over 10% is Not working.
The resting offset decides the answer. Under 5% is Okay, 5% to 10% is borderline, and over 10% is enough to move a camera or character in many games.

Largest jump while untouched

Some sticks sit still most of the time and then twitch. The largest jump is the furthest the stick strayed at any single moment in those five seconds. The test flags a jump to 15% or more, and well beyond where the stick was resting, with Check this. If nothing bumped the controller, a jump like that points to drift that comes and goes. A jump above 50% almost always means a hand or a knock, and the test says so and suggests running it again.

A line chart of a stick’s distance from centre over five seconds. It sits steadily around its resting offset, then spikes once above the 15% line before settling again.
Over the five hands-off seconds, the average gives the resting offset and the highest point gives the largest jump. The test flags a spike past 15% to check.

Roundness and reach: the controller circularity test

The circles step compares the outer path of each stick with a perfect circle. Roundness error is how far that path strays from the circle overall. Shortest reach is how far the stick got in its weakest direction. If a stick reaches less than 85% of its travel somewhere, the test flags it with “A stick stops short of its full travel”, which usually means dirt, wear or a calibration that has wandered.

A stick’s traced outline drawn over a dashed perfect circle. The outline follows the circle closely except for a dent at the lower left, marked as the shortest reach.
The circles step: the stick’s outline against a perfect circle. A dent like this one is the shortest reach. Roundness is there for reference and doesn’t change the answer.

Both numbers only appear once you’ve covered at least 90% of the circle, so go all the way round. They’re comparisons made in the browser, not lab measurements, and plenty of healthy sticks trace a slightly squared-off path. That’s why roundness is there for reference and doesn’t change the answer.

Okay, Check this or Not working

  • Okay: “No stick drift found”. Both sticks rest within 5% of centre, which a game’s dead zone hides.
  • Check this: “Borderline: worth keeping an eye on”. A stick rests between 5% and 10% off centre, or jumped past 15% while untouched. Most games hide a resting offset this small, but one with a small dead zone may not. You also get Check this, as “A stick stops short of its full travel”, when a stick reaches less than 85% of its travel in some direction.
  • Not working: “Stick drift found”. A stick rests more than 10% off centre, enough to move a camera or character by itself in many games.
The result: the left stick’s dot rests right of centre, the answer reads Not working: stick drift found, and a bar marks the left stick at 14% against the 5% and 10% limits, the right at 0%.
What a drifting stick looks like in the result: the dot rests off centre, and the bar shows where 14% falls against the 5% and 10% limits.

Each answer names the stick and the figure behind it, for example “Left stick rests 12% off centre”. These limits are starting values that we’re still tuning, and How this test decides, under the result, always shows the ones in use.

How much stick drift is normal?

Some, and that surprises people. A thumbstick is a small mechanical part held in the middle by springs, and its centring has a tolerance, so few sticks come to rest at exactly zero. A reading of one or two percent is ordinary on a perfectly good controller, and it’s the reason every game ignores a small area around the centre.

So “is stick drift normal?” has two answers. A small resting offset, under 5% here, is normal and harmless. Drift you can see in a game, where the view moves while your thumbs are off the sticks, isn’t normal on a working controller, and it tends to get worse with use rather than better. The useful question is whether the offset is bigger than what your games ignore, which is what the dead zone is about.

Dead zones: why stick drift shows in one game and not another

Games ignore a small area around each stick’s centre, called the dead zone. It’s typically somewhere between 5% and 25% of the stick’s travel, and many games let you change it in their controller settings. Drift smaller than a game’s dead zone is invisible in that game. Drift larger than it shows up as movement you didn’t ask for.

Two enlarged stick centres, each resting 7% off centre. On the left the shaded 5% dead zone is smaller than the drift, so the camera drifts. On the right the 10% dead zone covers it and hides it.
The same stick, resting 7% off centre, in two games. A 5% dead zone lets the drift through; a 10% one hides it.

This is why a controller can drift in one game and behave in the next, and why players who set small dead zones for quicker aim notice drift first. Read the resting offset next to the dead zone in your game’s settings. If the offset is bigger, you’ll see drift; raising the dead zone just above it hides it, at the cost of a little precision around the centre.

The stick drift test works as a dead zone test, too. Under the result, What a game’s dead zone would hide sets each stick’s offset against the limits, so you know roughly how big a dead zone you’d need.

Testing a PS5, Xbox or Switch controller for stick drift

The test reads any controller your browser can read, so the steps are the same for every brand. The differences are in how you connect and what you can do afterwards.

PS5 DualSense and DualSense Edge

For a PS5 stick drift test, connect the DualSense to a computer with a USB-C cable, or pair it by holding Create and the PS button until the light bar flashes. Then run the test as above. On a DualSense Edge, a stick that fails doesn’t mean a new controller: Sony sells replacement stick modules, and its page on replacing the Edge’s stick module shows the swap. Run the stick drift test again afterwards to confirm the new module rests at centre.

Xbox controllers

For an Xbox controller drift test, plug in by cable or pair over Bluetooth with the pair button on the top edge. Before you decide a stick is worn, update the controller’s firmware with Microsoft’s Xbox Accessories app on Windows or on the console, then test again. On an Elite controller the same app sets stick sensitivity curves, which change how the stick feels in play but don’t move where it rests.

Switch Pro Controller and Joy-Con

A Switch Pro Controller pairs with a computer over Bluetooth, using the sync button on its top edge. If it drifts, the console can recalibrate it: System Settings, then Controllers and Sensors, then Calibrate Control Sticks, as set out in Nintendo’s steps for calibrating the controllers. Then run the stick drift test again to see whether the offset came down. We haven’t checked Joy-Con with this test yet, so we can’t promise how one will read.

Checking a used controller before you buy it

Drift is the fault sellers least often mention, because it can be mild enough to miss in a quick go. If you’re buying in person, bring a laptop, connect the seller’s controller and run the stick drift test on the spot. Then open the gamepad tester to check every button and trigger too. Both take a couple of minutes.

Save the report while you’re there. It records the date, the controller’s reported name and the readings, which is handy if a problem turns up a week later. If you’re buying online, a report the seller sends you is better than nothing, but remember that it can’t prove which controller the seller tested. Watching the test run yourself, or over a video call, is the stronger check.

What to do if a stick drifts

Start with the free fixes: recalibrate where the maker offers it, raise the dead zone in your game, clean around the base of the stick and update the firmware. Run the stick drift test again after each one, so you know which helped. If the offset keeps growing, the sensor inside the stick is usually worn and needs replacing. The guide to how to fix stick drift goes through every option, from cleaning to repair and warranty.

What a browser drift test can and cannot tell you

The test reads the stick positions your operating system hands to the browser, much the same raw movement a game starts from before it applies its own dead zone. That makes the resting offset a fair guide to drift. It has limits, and it tells you when it hits them.

  • Filtering can hide slight drift. If your controller or system already ignores small movements, the test can’t see them either. When a stick reads exactly zero the whole time, the result says so, because a real stick almost never does.
  • Drift that comes and goes may not happen in five seconds. Run the test a few times, including straight after a gaming session.
  • Roundness and reach are comparisons made in the browser, not lab measurements.
  • The test can’t say why a stick drifts, whether that’s dirt, wear or damage, and it can’t predict how quickly it will get worse.
  • If the browser doesn’t recognise the controller’s layout, the test reads the first four axes as the two sticks, which is the usual arrangement, and says so. If it can’t tell which readings are sticks at all, it says “The sticks couldn’t be measured in this browser” and suggests Chrome or Edge rather than guessing.

The stick drift report

After the result, Download the report (PDF) saves a dated one-page report, on A4 or US Letter paper. It shows the date and time, the controller’s name as it reported it, its layout, your browser and system, each stick’s readings, the answer, how the test ran and how it decided the answer. It’s meant for a return, a repair or a second-hand sale.

Your browser makes the report and saves it straight to your device. Nothing is stored or uploaded, as the privacy page explains. And because a browser cannot read a serial number, the report shows what the test measured in one session, not proof about a specific controller. Whether a shop or maker accepts it is up to them.

Questions

How do I know if my controller has stick drift?

Put it down, run the test and read the resting offset. Under 5% is fine. Over 10% is drift that many games will show, and between the two it depends on the game’s dead zone. In a game, the giveaway is a view or character that moves while your thumbs are off the sticks.

Can I use the report for a return or warranty claim?

You can attach it as a dated record of what the sticks did, for a return, a repair or a second-hand sale. It can’t prove which controller the test read, because a browser can’t read a serial number. Whether a shop or maker accepts it is up to them.

My controller drifts in games but passes here. Why?

The drift may come and go, so run the stick drift test a few more times, ideally straight after playing. Your game may also use a dead zone smaller than 5%, so look at the resting offset, not just the answer. If a stick read exactly zero, something is filtering small movements before they reach the browser.

Why does my stick drift when I press R2 or L2?

This test reads the sticks with nothing pressed, so it won’t catch drift that appears only while a trigger is held. To look for it, open the gamepad tester, open its raw values, hold the trigger and watch the stick’s numbers without touching the stick. If they move while the stick is untouched, the drift is real and the stick needs the same fixes as any other drift.

Does it work with Hall effect or TMR sticks?

Yes. The test reads any stick, whatever sensor is inside. Hall effect and TMR sticks don’t wear in the same way as the usual kind, but they can still rest off centre: SCUF’s own support page on stick drift and jitter says they don’t fix mechanical centring by themselves. So they’re worth testing too.

Can this tool fix stick drift?

No, and no website can: a web page can read a controller but can’t change it. What the test does is tell you how bad the drift is and whether a fix worked. Cleaning, recalibrating or a larger dead zone can help with mild drift, and the stick drift guide covers each option.

Are the values different between browsers?

They can differ a little, because each browser gets the controller from the system in its own way. Compare results from the same browser, and if one browser shows a stick at exactly zero while another shows movement, trust the one that shows movement. If a browser doesn’t recognise the layout, try Chrome or Edge.

Why did it flag a jump when I didn’t touch anything?

A jump above 15% while the controller sat still points to drift that comes and goes. A knock to the table can cause one too, so run the test again and keep clear of the controller.

Can I test Joy-Con, a Switch 2 controller or a Steam Deck?

Not with any promise yet: we haven’t checked any of them with this test. If your browser shows the controller after a button press, the test will read it, and it says plainly when it doesn’t recognise the layout. For a Steam Deck’s own sticks, the browser has to see them as a controller first.

Does it test one stick if my controller only has one?

Yes. With a single stick, the answer is about that stick alone (“The stick rests within 5% of centre”). On a controller whose layout isn’t recognised, the test takes a reading that rests at the end of its range for a trigger and leaves it out, so it doesn’t mistake it for a drifting stick.