Fight Stick Tester / Test Arcade Stick Inputs

A fight stick has more to go wrong than a standard controller, and less margin for error when it does. A microswitch that's started chattering, a direction that registers late, or SOCD behaviour that doesn't match tournament rules can all cost you a match, and none of them are obvious until you're already losing.

Testing takes a couple of minutes in a browser, needs no software, and covers things a normal controller test doesn't. Here's what to check on an arcade stick or leverless controller, and what the results actually mean.

What to test, and why sticks are different

A fight stick isn't just a controller in a bigger box. Three things make testing it different.

Every input is digital. Arcade buttons and microswitches are on or off, with no analog range. That means faults present differently: instead of drift or a weak signal, you get a button that doesn't register, registers twice, or sticks.

Directions come from microswitches too. A traditional lever pushes against four microswitches, one per cardinal direction, and diagonals are produced by two firing at once. So a "bad diagonal" is usually one microswitch, not a mysterious calibration problem.

Leverless controllers can do something a lever physically can't: send two opposite directions at the same time. That's the SOCD issue, and it's the single most important thing to test if you compete.

Test the buttons first

Start with the simplest check, because a failing microswitch is the most common fight stick fault.

Connect your stick and open a browser button test, then press each button individually and watch what registers. You're checking three things per button.

Does it register at all? A button that needs a hard press, or works intermittently, has a worn or dirty microswitch.

Does it register once? A single press producing two inputs is switch chatter, the same bounce phenomenon that causes mice to double-click, and it's caused by worn contacts bouncing beyond the debounce window. On a fight stick this is brutal, since a doubled input can turn a single attack into a repeat or break a combo.

Does it release cleanly? A button that stays registered briefly after you let go is on its way to sticking, and it will cause dropped inputs in sequences that depend on precise release timing.

Arcade microswitches are rated for millions of actuations but are also cheap and, crucially, usually replaceable without soldering on quality sticks, since they connect with quick-disconnect terminals. That's one of the genuine advantages of a fight stick over a pad: a failing button is a two-minute fix with a $2 part.

Test the directions

With a traditional lever, push each cardinal direction one at a time and confirm each registers cleanly and releases when you return to neutral.

Then test the diagonals deliberately. Since diagonals come from two microswitches firing together, a diagonal that fails while both cardinals work individually points at timing or a marginal switch rather than a dead one. This matters enormously for motion inputs: a quarter-circle forward needs down, down-forward, and forward to register in sequence, and a weak down-forward is exactly why a special move "doesn't come out."

If you use numpad notation, which most fighting game communities do, it's worth mapping what you see to it. The keypad layout treats 5 as neutral, 1 to 3 as the bottom row, 4 and 6 as left and right, and 7 to 9 as the top row. So 236 is the quarter-circle forward you'd use for a fireball. Testing each numbered position individually tells you precisely which part of a motion is failing.

On a leverless controller, test each direction button the same way you'd test any button, since that's what they are.

Test SOCD behaviour

This is the check that matters most competitively, and the one a standard controller test won't cover.

SOCD stands for Simultaneous Opposing Cardinal Directions, meaning left plus right, or up plus down, pressed at once. A lever physically can't do this. A leverless controller can, and without handling it, pressing back and forward together could let you block both directions simultaneously, which the game was never designed to allow.

So leverless controllers include SOCD cleaning: firmware that resolves the conflict before the input reaches the game. The common methods are:

Neutral, where opposing directions cancel to no input. Last Input Priority, where the most recently pressed direction wins. First Input Priority, where the originally held direction wins.

To test yours, press left and right simultaneously and observe what the tester reports, then repeat with up and down. Note the behaviour for each pair separately, because many controllers use different rules for the horizontal and vertical axes.

What tournament rules actually require

Here's where accuracy matters, because getting this wrong can mean a disqualification.

Evo's ruleset states that a controller may not send simultaneous opposing cardinal directions, and that this is typically enforced through cleaning firmware that removes one input before it reaches the game. Stock gamepads such as the DualShock 4 and DualSense are explicitly exempt from the rule.

Broadly, all SOCD methods that prevent simultaneous opposing inputs are acceptable, including Neutral, First Input Priority, and Last Input Priority, unless a specific game's ruleset says otherwise.

That caveat is doing real work. Capcom Pro Tour modified its policy to require that up plus down resolves to neutral rather than to up. The reasoning is worth understanding: on a standard pad, pressing up and down together gives neutral, but on many leverless controllers it gave up, which let charge characters retain charge without returning to neutral and made certain inputs faster than intended. The rule change aligns leverless behaviour with what a pad can physically do.

Also worth knowing: down plus right are not opposing cardinals, so a lever or controller sending that combination is fine, and it can legitimately be bound to a single input.

The practical takeaway is that leverless controllers are tournament legal at major events including Evo and Capcom Pro Tour provided their SOCD cleaning complies, and many ship with a compliance mode. But rulesets differ between organisers and evolve between seasons, so check the specific event's current rules rather than assuming your controller's default is legal everywhere. Many sticks also include a tournament lock switch that disables system and menu buttons during matches, which some events require.

Check for input lag

Fighting games are frame-sensitive in a way most genres aren't, so latency matters more here than almost anywhere else.

Connection type is the main variable. Most competitive fight sticks are wired for exactly this reason, and if yours supports both wired and wireless, wired is the correct choice for anything serious. An input lag test lets you compare your setup's actual performance rather than trusting a spec sheet, and it's worth testing both connection modes to see the real difference on your hardware.

Converters and adapters are the other common source. If you're using an adapter to run a stick on a console it wasn't built for, that adapter sits in your input path and adds latency, and quality varies enormously between them.

Reading your results

Pulling the checks together:

Button doesn't register or needs force: worn or dirty microswitch. Usually a cheap, solder-free replacement on a quality stick.

Button registers twice from one press: switch chatter from worn contacts. Same fix.

Button stays registered after release: a sticking button or a failing switch, and it will cause dropped inputs.

A cardinal works but its diagonals don't: one microswitch is marginal, or the lever's actuator isn't engaging both switches cleanly.

SOCD produces an unexpected result: check your controller's SOCD mode setting, since many are configurable, and verify against your target event's ruleset.

Everything tests clean, but inputs still drop in game: the problem is the game's settings, your connection, or an adapter, not the stick.

If your stick reports button indices rather than familiar labels and you're unsure what maps to what, our guide on what gamepad tester numbers mean explains the standard mapping and why indices sometimes don't match printed labels.

The short version

Fight sticks are all-digital, so faults show up as buttons that don't register, register twice, or stick, rather than as drift. Test every button individually for registration, doubling, and clean release, then test each cardinal direction and each diagonal separately, since diagonals depend on two microswitches firing together. If you use a leverless controller, test SOCD behaviour on both axes, because that's what tournament rules govern: Evo requires that opposing directions never reach the game simultaneously, and Capcom Pro Tour additionally requires up plus down to resolve to neutral. Check event rules directly, since they differ and change. And keep your stick wired, since fighting games punish latency more than most genres.