Before you buy
Checking a used controller before you buy
A second-hand controller is sixty dollars of hardware with one fault that makes it worthless and half a dozen that are fine to live with or cheap to fix. The one that matters is stick drift, it is measurable in two minutes, and a seller cannot hide it from a test the way they can from a photograph. The rest of this list is the order that finds the deal-breakers first and leaves the haggling points for last.
The run, in order
Measure both sticks at rest and after they return
This is the check. Drift is the reason most used controllers are for sale, it is the fault that cannot be lived with, and it is the one a seller cannot show you is absent. Let go of both sticks for the first measurement, then swing each to the rim and release for the second: a stick that rests correctly and lands off centre after being moved is the earlier symptom and the one a rest-only test misses. Anything at or above 8% is a stick module replacement, which is a soldering job, and a reason to walk away at anything near full price.
Press every button once and roll both sticks to the rim
The counter says how many of the controller's buttons have registered, which is the only reliable way to find one that never does. Roll each stick round its edge and watch the dot: a stick that does not reach the rim in one direction has lost travel, which is the same module fault as drift arriving from the other side. Squeeze the triggers slowly and watch the value climb — a trigger that jumps from 0 to 1 has lost its analogue travel, which matters in any game with a throttle.
Look at the stick caps and the thumb rubber
Worn rubber on the stick caps is cosmetic and cheap — replacement caps are a few dollars. But a cap that is worn smooth on one side tells you how hard the stick was used and in which direction, and a controller with one bald cap is a controller whose module has done a lot of miles. Read it as a mileage gauge, not a fault.
Pair it, then check the battery
Pair with the console or phone you brought, or over Bluetooth to a laptop; a controller that will not pair is either a dead battery or a dead radio, and you cannot tell which in the room. On a DualSense or an Xbox with a rechargeable pack, ask how long it holds a charge — the battery is replaceable on some, sealed on others, and a controller that lasts an hour is a controller with a dead cell. On Xbox with AA batteries this step is free: bring your own.
Check the rumble and, on PlayStation, the adaptive triggers
Neither is testable from a browser, and both are faults people accept until they notice. Start any game with rumble; on a DualSense, one with adaptive trigger resistance. A dead motor is a cheap part and a fiddly repair; dead adaptive triggers are the reason some DualSense controllers are sold cheaply, and a fair reason to pay less rather than to walk.
Open the battery door, or look at the screws
Evidence of a previous repair. Scratched or chewed screw heads, a seam that does not sit flush, a rattle when shaken gently. A controller that has been opened once has usually been opened for drift, and a repair done well is fine — but a repair done badly is a controller that will drift again soon, and the seller knows which it was.
Check the headphone jack and the port
Plug headphones into the jack and a cable into the port, and wiggle each gently. A loose port on a PlayStation or Xbox controller is common and irritating rather than fatal — it charges intermittently and drops wired mode — and it is a reason to pay less. A jack that crackles is dirt or a cracked solder joint, and cheap either way.
Ask what it was used for
A controller from a fighting-game player has a battered D-pad and fresh sticks. One from a shooter player has the opposite. Neither answer is disqualifying, but it tells you which part to look hardest at, and the answer is usually honest because sellers do not think it matters.
If you brought a laptop, measure the input path
Not a controller fault as such, but a check that costs a minute: a wireless controller that measures far slower than a wired one on the same machine has a poor radio or is on the wrong Bluetooth profile. It is also a way to confirm the controller really is talking to the computer rather than just lighting up.
Why the order
The drift test is first because it is the only finding that ends the deal, and because it takes two minutes. There is no point examining the stick caps of a controller that fails it. Everything after that is about setting the price, in rough order of how much each finding is worth.
The two browser tests do the measuring; the rest is hands and eyes. Bring a phone or a laptop — the drift and button tests run on either with the controller paired over Bluetooth — and bring your own batteries if it takes them.
What each finding is worth
Drift at or above 8% — walk away, or buy it as a project at parts price.
Drift between 2% and 8% — real, climbing, currently hidden by game dead zones. Worth a discount; worth knowing you are buying a controller with a year or so in it.
A dead button or a trigger with no travel — a repair, and not a hard one on most controllers. Worth a discount matching the repair.
Weak or dead rumble, dead adaptive triggers — live with it or pay less. Not worth a walk on their own.
A loose port or a crackling jack — irritating, cheap to fix, worth something off.
Worn caps, scuffs, a missing back paddle — cosmetic. Replacement caps cost almost nothing. Mention it, do not price it.
Opened before — ask why. A drift repair done well is a plus. One done badly, or one the seller will not talk about, is a controller that will drift again.
The one thing a photograph cannot show
Everything on a listing can be photographed except the fault that matters. Drift is invisible in a picture, invisible in the seller’s hands, and often invisible in the first minute of a game, because the game is hiding it. It is visible in a two-minute test with the controller lying on the table, and that is the whole reason to run one before paying.
Common questions
What should end the deal immediately?
Drift at or above 8% on either stick, measured after the swing-and-release step. That is a stick module replacement, a soldering job on nearly every controller, and it will get worse. Everything else on this list is a reason to pay less; this is the reason to leave. The one exception is a controller with Hall-effect sticks, which do not drift by wear — if the seller claims that, the drift test will confirm it in two minutes.
How much should drift knock off the price?
All of it, unless you intend to replace the modules yourself. A pair of modules is cheap; the labour is not, and a shop repair costs a good fraction of a new controller. Under 2% is a healthy stick and not a bargaining point. Between 2% and 8% is drift that games will hide with their dead zone — it is real, and it is climbing, and it is a fair reason to take something off.
Are Hall-effect sticks worth paying more for?
Yes, if the controller has them. They sense magnetically rather than through a wearing contact, and they do not drift with use. Some third-party controllers ship with them; the first-party PlayStation and Xbox controllers mostly do not. A used controller that has been fitted with Hall-effect modules by a previous owner is a repair worth having, and the drift test will show a stick that sits dead centre.
The seller says it was only used for a month.
The stick caps say how long it was used, and the screws say whether it was opened. Neither can be argued with. Drift also does not follow age: a controller used hard for a month can drift, and one used gently for two years may not. Measure it.
Can I run these tests without a computer?
The drift and button tests run on a phone with a controller paired over Bluetooth — Android and iPhone both support the standard pads. Pair it to the phone you bring, open the tests in the browser, and press a button so the phone shows it. The rumble and trigger checks need a console or a game.
Is a used controller worth it at all?
At half of new and passing the drift test, yes. At three quarters of new, a new one with a warranty is the better buy, because the one fault that matters can arrive at any time and only a warranty covers it. The tests here are what make the first case safe.