Mouse steadiness target challenge
Click start, move the pointer into the target, then keep as still as you can — the page waits for your hand to stop, then times 5 seconds and measures how far the pointer travelled in that stretch. If the pointer leaves the circle mid-round the round is void — start over.
The figure to watch is total travel: every little tremor back and forth adds up notch by notch, which says more about a steady hand than “how far did it drift at most”. It measures how steady the pointer is — the pointer has already been through the system’s smoothing and acceleration, so the system’s share is mixed in, and it is only good for comparing one machine with itself.
Only the result number and confidence level are uploaded; raw samples stay in your browser.
How to read these numbers
| Reading | What to look at | Notes |
|---|---|---|
| Total travel | How steady the hand is | The total distance the pointer covered over this stretch. With your hand still, the pointer still walks back and forth over a few pixels from hand tremor, desk vibration and sensor noise, and every bit of it counts — so it is far more sensitive than “how far did it drift at most”, and it is the main figure on this page. |
| Tremor span | What the tremor looks like | The distance between the two furthest of all the sample points. Read it together with total travel: large travel with a small span means the pointer is trembling fast in place; when both are large, the hand is really moving. |
| Samples | How many times the pointer moved | 0 is the best result — it means the sensor produced not a single count over those 5 seconds. But that 0 only means anything once the “move into the target” step has been completed; the reason is in the FAQ below. |
Why holding a target gives no score
This page deliberately gives no score, no grade and no pass mark. Turning it into “steadiness 82/100” needs a set of weights, and those weights would be made up — this site does not invent numbers for that. All it can say is: this run was steadier than the last, or it was not.
There is a more practical reason too: the same tremor covers four times as many pixels at 3200 DPI as at 800 DPI. So that number is only worth comparing against your own history, and ideally at the same DPI setting.
If you want to be steadier
- Rule out the mouse itself first. Run the “Stationary drift” test on the trail-test page with your hand completely off the mouse — that one measures whether the mouse creeps on its own. If the mouse is moving by itself, any practice afterwards is wasted.
- Try again on a dark cloth pad. On glass, plain white paper or a very slick desk, most mice read noticeably worse.
- Don’t float your wrist. This is not something this page teaches, but it is something this page can measure: take one run with your arm supported and another with it floating, and the two numbers come out far apart.
Common questions
Steadier is smaller, but there is no pass mark. Two reasons: first, the number is tied to your mouse’s DPI — the higher the DPI, the more pixels the same tremor covers, so a lovely score at 800 DPI looks far worse at 3200 DPI; second, it also depends on how you grip the mouse and whether your arm is supported. So this page does same-machine before/after comparisons only: against last time, against before you changed the mousepad. Comparing a tremor distance across devices compares device settings, not hands.
Because there is an operating system in between. Every tiny movement of your hand is first put through smoothing and (if it is switched on) acceleration by the system, and only then becomes the pointer position on screen. What this page gets is the result after that processing, so it measures how steady the whole chain is, with the system’s, the driver’s and the screen’s share mixed in. To look at the mouse hardware alone for self-oscillation, use the “Stationary drift” section on the trail-test page — that one asks you to take your hand completely off the mouse.
Because it is not a difficulty setting, it is a boundary. The moment the pointer leaves the circle the round is void and you start over — and if the circle were small you would be fighting the edge the whole time instead of concentrating on holding still. Drawing it 110 pixels across means normal tremor (a few pixels) cannot reach the edge at all, so the circle is there for one reason only: to make sure you really are aiming at the same point for those 5 seconds.
Because “received 0 samples” is ambiguous on its own. With the pointer outside the browser window, or the mouse not connected, the page also receives not a single event — which is indistinguishable in the data from “your hand was rock steady”, and the conclusion is the opposite. So the flow has an extra step: first it waits for you to move the pointer into the target and confirms the page really is receiving a signal, then it waits for you to stop, and only then does it start timing those 5 seconds. Once that step is done, 0 really does mean 0.
