Button Latency Test — Press to Register Delay
Press the button when the cue fires and the tool times how fast it lands. This button latency test measures the end-to-end delay between a visible cue and a registered input, in milliseconds.
The reading captures the whole chain — your reaction, the switch, the controller's firmware, the USB or Bluetooth transport, the OS, and the browser. Absolute numbers include your reaction time, so the useful comparisons are between buttons on the same pad, or between wired and wireless on the same controller. The tool below runs 10 trials, then reports the median, the spread, and the standard deviation. Everything runs in your browser — no download, no signup, nothing leaves your device.
Press a button to begin
Connect via USB or pair over Bluetooth, then press any button so the browser detects your pad.
Choose a button to test, press Start, then press it the instant the cue turns orange.
Absolute numbers include your reaction time. Compare buttons or wired vs wireless runs — the differences are diagnostic, not the raw figure.
Your reading is a sum of several delays
A press → register number isn't a single thing. It's the total of every link in a short chain between your thumb and the browser. Knowing the chain is what makes the reading useful instead of just a curiosity.
The latency stack
Roughly how a single trial breaks down — typical figures, not exact. The biggest piece is almost always you.
From the orange cue lighting up to your brain telling your thumb to move. By far the largest slice — and the reason absolute totals always look "slow."
Your monitor's input lag and frame timing. A 60 Hz screen adds more here than a 120 or 240 Hz one.
The physical switch closing and the controller's firmware packaging the press into a report. This is where a worn switch shows up.
How the controller sends that report to your machine. USB tends to sit near the low end; Bluetooth varies more.
The operating system handing the input to the browser, and the browser polling the Gamepad API.
Because human reaction dominates, a single absolute number doesn't tell you much. The controller-specific parts only become visible when you compare two runs against each other — same hands, same screen, only one variable changed.
Wired versus wireless, side by side
The classic comparison. Run the test once over USB, once over Bluetooth, same controller, same button — the difference between the two median values is what the transport is actually costing you.
- Lowest, most consistent transport delay.
- Tighter standard deviation across the 10 trials.
- Recommended for any latency-sensitive game.
- The baseline you measure other connections against.
- Adds a few milliseconds versus USB on average.
- Wider spread — the standard deviation usually grows.
- Worse with interference, distance, or low battery.
- Fine for casual play; not ideal for competitive timing.
A wired-versus-wireless gap of 10–25 ms in median is normal. A much larger gap, or a runaway standard deviation, points at a poor radio environment rather than the controller itself.
What to compare, what to ignore
- Button vs button on the same pad. A single button noticeably slower than the others is a worn switch.
- Wired vs wireless on the same pad. Measures what your transport is costing.
- Standard deviation across runs. Lower is better — a steady stick is a sign of a reliable input chain.
- Before vs after a change — firmware update, new cable, lower polling rate.
- Your absolute number. It includes your reaction; it isn't your controller's latency in isolation.
- Comparing to someone else's score. Different reflexes, screens, and rooms — the figures aren't comparable.
- A single trial. One slow press doesn't mean a slow pad — only the median over 10 does.
- The best trial alone. That's your reflex ceiling, not your controller's typical behaviour.
A protocol you can trust
The trick with latency testing is doing it consistently. Follow this five-step protocol and your numbers will actually mean something.
Settle in first
Sit normally, hands on the pad, eyes on the cue. A warm-up run gets your reflexes out of the way.
Pick one button
Test one button at a time. Comparing across buttons later is how you find a worn switch.
Run 10 trials
Don't stop short. The median only stabilises across the full ten — fewer trials are noisy.
Change one thing
Wired vs wireless, button A vs button B — change one variable, then run another 10.
Compare the medians
The gap between two medians is your real reading. The absolute number on its own isn't.
If your numbers look bad
Two readings tell different stories. Find which one is the problem on your run, then follow its branch.
An unstable input chain
Times all over the place mean something between you and the browser is varying trial to trial. The fixes target consistency.
- Re-test wired. Bluetooth interference is the most common cause of a wide spread.
- Charge the controller fully. Low battery widens wireless variance noticeably.
- Move closer to the receiver. A few feet matters on Bluetooth, especially through walls.
- Close background tabs and apps. A busy CPU adds frame-to-frame jitter to the browser.
- Disable browser extensions. Some monitor every keypress and can slow the input chain.
A slow but steady chain
Times that are consistent but slow mean something is reliably adding delay. The fixes target the source of that constant lag.
- Check display refresh rate. 60 Hz adds frame time versus 120/144/240 Hz on the same machine.
- Test a different button on the same pad. A specific slow button points at a worn switch.
- Update the controller firmware. Vendors occasionally cut polling delay in updates.
- Switch USB ports or cables. A bad cable can add a steady few milliseconds.
- Compare on another machine. If the median drops elsewhere, your machine is the slow link.