Mouse acceleration test
Pointer acceleration means: the faster you move, the more the system scales your movement up. So the way to test it is to push the same physical distance at two speeds, slow and fast, and see whether the cursor covers the same distance on screen. If it does, there is no acceleration; if the fast run clearly goes further, there is.
This page does not need acceleration switched off first — what it measures is precisely whether acceleration is on right now. It does need a ruler, and both runs have to cover the same distance.
How to test
- Pick the distance below (10 cm by default), rest the mouse at the ruler’s starting mark, then click Start measuring on the first run. Push the whole distance slowly, then stop.
- About 0.7 seconds after you stop, it settles on its own. Put the mouse back at the start, click Start measuring on the second run, push the same distance this time quickly, and stop the instant you reach the end.
- Once both runs have a number, the verdict appears below.
How the verdict is decided
The reading is the fast run’s estimated DPI ÷ the slow run’s estimated DPI. Both runs cover the same physical distance, so with no acceleration on the system the ratio should be around 1 — not exactly 1, because a hand-pushed distance is always a few millimetres out.
| Ratio | Verdict | Meaning |
|---|---|---|
| < 1.15 | No acceleration found | the difference is within what a change of hand speed and measurement error can explain |
| 1.15 – 1.50 | Slight acceleration | Windows’ “Enhance pointer precision” is around this order of magnitude at moderate speeds |
| ≥ 1.50 | Clear acceleration | with acceleration on and a fast flick it reaches 2× or more |
Why not just tell me “acceleration is on”
Because a web page cannot read that switch. Whether “Enhance pointer precision” is ticked is not something the browser has an API for. Nor does it tell you whether your mouse driver adds a layer of acceleration of its own — some gaming mouse drivers bypass the system setting and handle it separately.
All that can be observed is behaviour: whether the cursor answers the same hand movement consistently at two different speeds. So this page can only give a behavioural verdict; it cannot state the state of that checkbox. If you do measure acceleration, then wherever the switch sits, you have felt it in your own hand — and that is enough to decide with.
The cases where the reading drifts
- The two speeds were not far enough apart. This is the most common one. If the first and second runs differed only slightly, acceleration barely shows at all and the verdict leans towards “none”. The page lists both actual speeds, and when the gap is too small, treat the verdict as void — it is not that the mouse is fine, it is that this run did not measure it.
- The cursor hit the edge of the screen on the fast run. With acceleration on the cursor travels a long way, and once it hits the edge its coordinates stop changing: the measured travel comes out short and the acceleration gets hidden. Choosing a shorter distance avoids it.
- The hand-pushed distance was not the same. This one cannot be detected automatically — the page only knows how many pixels were covered on screen, not how far your hand actually pushed. So both runs have to be lined up on the same stretch of the ruler.
Common questions
Quite the opposite — it only measures anything with acceleration on. What this page asks is “does my system have acceleration on right now?”. So measure with the mouse set up the way you normally use it. If the verdict afterwards is “acceleration on”, then decide whether to turn it off.
Because what is being compared is “the same physical distance — does it come out as the same travel on screen?”. With different distances the two readings are not comparable. If you pushed 10 cm the first time and 12 cm the second, the difference has your own hand error mixed into it and no conclusion can be drawn. That is why the page locks both runs to the same distance.
With acceleration on, the same 10 cm pushed quickly sends the cursor much further — further than the screen may have room for, so the cursor hits the edge part-way through. Once it hits the edge, the coordinates the browser reports stop moving, the measured travel comes out smaller than the real one, and the result is low rather than high. So pick a shorter distance, and stop the instant you reach the end.
No. All this page can say is “within the speed range you just used, no scaling was observed”. If your two runs were in fact at much the same speed, the acceleration stayed hidden — it only shows itself once the speed difference is big enough. So the page shows both actual speeds, and you can see for yourself how big a speed difference that ratio came from. If the speed difference was too small, the verdict does not hold.
