How to use
- Click each button inside the pad: left, right (the context menu is suppressed there), middle, and the side buttons B4/B5. Each lights up when it registers.
- Roll the scroll wheel up and down; every step is counted, and the ⬆/⬇ badges show which direction registered.
- Do 20–30 slow, deliberate single clicks with the left button. Watch Unwanted double-clicks — it should stay at 0.
- Move the mouse around the pad and watch Cursor position and the Hz figure; jumps or freezes point at the sensor or the cable.
- Reset clears the counters. Repeat the single-click check a couple of times — a switch that is starting to fail is inconsistent.
What the mouse test checks
The pad listens to pointerdown, pointerup, wheel and pointermove events from the browser. Every button that the OS reports — left (0), middle (1), right (2), back (3) and forward (4) — lights up its badge; the wheel counts steps in each direction; the cursor position and the events-per-second figure come from the movement stream. Right-click's context menu is disabled inside the pad so you can test the button without a menu popping up.
The important number is Unwanted double-clicks. The page timestamps every press with performance.now(); when two presses of the same button arrive less than 80 ms apart, it counts one. Humans do not single-click twice within 80 ms — a deliberate double-click is typically 100–200 ms apart — so anything registered in that window came from the switch, not from you.
The double-click bug: why worn switches do it
Almost every mouse uses small mechanical microswitches under the buttons, and the most common ones are Omron's D2FC series, rated by the manufacturer for 10, 20 or 50 million clicks depending on the variant (the rating is printed on the switch, e.g. D2FC-F-7N(20M)). Inside is a thin metal spring that snaps between two contacts. As the switch ages, the contacts oxidise and the spring loses tension, and a single press produces a burst of contact bounces long enough to slip past the mouse firmware's debounce filter. The result is the notorious double-click bug: one click opens two windows, one shot fires two, a drag drops half-way.
Two things make it worse. Higher polling rates and shorter debounce times (some mice let you set 0–2 ms for lower click latency) leave less time to filter the bounce. And butterfly or drag clicking hammer the switch far faster than office use, so gaming mice tend to develop the fault earlier than their rating suggests. There is no reliable public statistic on how many clicks a mouse survives in practice; the ratings above are the manufacturer's, measured under test conditions, not a promise.
Reading the results
| Symptom in the test | Likely cause | What to do |
|---|---|---|
| Double-click counter rises on slow single clicks | Worn switch (contact bounce) | Increase debounce in the mouse software if available; otherwise replace the switch or the mouse |
| A button never lights up | Dead switch, broken trace, or button disabled/remapped in software | Check the mouse software first; then a switch replacement |
| Button lights up only when pressed hard | Switch losing its snap, or a worn button shell | Failing soon; plan a replacement |
| Wheel skips steps or scrolls backwards | Dirty or worn encoder | Blow out dust; contact cleaner on the encoder often helps; otherwise replace the encoder |
| Cursor jumps or freezes while moving | Sensor dirt, bad surface, low battery, cable fault, or USB power saving | Clean the sensor, change the pad, charge; try another port |
| Side buttons don't register | Browser cannot see them, or they are bound to macros | Test in the mouse software or a game; some browsers reserve back/forward |
When to replace your mouse — and what to fix instead
- Replace when the double-click count grows on every run and a debounce setting doesn't fix it, or when a main button needs force to register. A faulty main button costs rounds; it is not worth fighting.
- Repair if you are comfortable with a soldering iron: microswitches cost very little and swapping them is a common fix. Optical switches (Razer, some Logitech and others) do not bounce and are the choice if you butterfly-click a lot.
- Fix in software first: many mouse utilities have a debounce or "click latency" slider — raising it by a few milliseconds often hides early bounce for months.
- Not a fault: the browser may not see B4/B5 if the OS binds them to navigation, and a wheel with free-spin mode will count many steps per flick. Both are expected.
What makes our mouse test different
- Double-click detection with a stated threshold (80 ms, timed with
performance.now()), so you can tell a worn switch from a deliberate double-click. - All buttons, both wheel directions and cursor tracking on one pad, right-click menu suppressed.
- No ads on the pad, no download, no driver — works with any mouse the OS sees, wired or wireless.
- Everything in the browser. No account, no submission.
Privacy: the events are handled by the script on this page and nowhere else. No click, movement or device data is sent to a server or stored beyond the current tab.
FAQ
How do I know if my mouse has the double-click bug?
Do 20–30 slow single clicks in the pad. If the Unwanted double-clicks counter rises — the page counts two presses of the same button within 80 ms — the switch is bouncing. Repeat a few times; early-stage faults are intermittent.
Why does my mouse double-click on its own?
Worn microswitch contacts. The spring inside loses tension and the contacts oxidise, so a single press bounces long enough to pass the firmware's debounce filter. It is wear, not a setting.
Can I fix the double-click bug?
Often temporarily: raise the debounce time in the mouse software if it has the option. Permanently: replace the switch (soldering) or the mouse. Optical switches do not have the problem.
My side buttons don't light up. Are they broken?
Not necessarily. Some browsers and operating systems bind buttons 4/5 to back/forward and don't pass them to the page, and some mice have them mapped to macros. Test them in the mouse software or in a game.
What is the Hz number?
How many movement events per second the browser received while you moved the mouse. Browsers coalesce events, so it is usually lower than the mouse's true polling rate; use the dedicated polling-rate test for a real measurement.
Does the test work with a wireless mouse or a trackpad?
Yes. Any pointing device the OS reports works; a trackpad shows the left/right buttons and movement, but has no wheel or side buttons.