Polling rate
How often a mouse reports, what the test can measure in a browser, and how to get a higher rate.
What it is
A mouse reports its movement many times a second: 125 times for many office mice, 500 or 1000 for gaming mice, and often 60 to 125 for trackpads. More reports mean smoother movement and less delay.
The number is called the polling rate, and it is measured in hertz: reports per second. At 125 Hz the computer hears from the mouse every 8 ms; at 1000 Hz, every millisecond. On a plain desktop the difference is hard to feel. In a game, or when you draw, a low rate shows as a pointer that moves in small steps and arrives a touch late.
The rate is set by the mouse and by the link it uses. A cable or the mouse's own receiver carries the rate the mouse was built for. Bluetooth often carries less, because the link itself reports at a lower, steadier pace to save power. A hub between the receiver and the computer can lower it too.
How the test measures it
While you move in circles for 5 seconds, the browser hands the page every report it received, each with its own time. The test waits a quarter of a second after the first movement so your hand is already in motion, then takes the typical gap between reports: the median, so a few odd gaps don't sway it. The rate is one second divided by that gap.
The window has to be long enough to trust. The test wants at least 3 seconds of movement and at least 29 gaps between reports; with less than that the result is not tested, and the card says the window was too short. Keep the pointer moving the whole time, in any pattern: circles are suggested because they are easy to keep going.
115 reports a second or more pass, 55 to 115 is a warning, and fewer fail. The bounds sit a little under 120 and 60, so a 125 Hz office mouse passes with room to spare and a 1000 Hz mouse that has fallen to 60 shows as a warning, not a fail.
The card also shows jitter: how far the gaps stray from the typical one, at the 95th percentile. A low number means the mouse reports like a clock; a high one means the reports bunch up and spread out, which you feel as a pointer that moves unevenly even when the rate looks fine.
What a browser cannot see
Browsers normally hand a page one pointer position per screen refresh, which is the display's rate, not the mouse's. To see the mouse's own rate the page asks for the samples the browser folded into that one position, called coalesced events. Where a browser gives none, the page can only see the screen's rate, so it does not judge: the card says this browser shows only the screen's rate and suggests Chrome, Edge or Firefox.
The clock matters too. Gaps of a millisecond need a clock finer than that. When the browser's clock is coarser than 2 ms, the page cannot measure the rate, and says so.
A trackpad's rate is shown but not judged: trackpads report less often by design, and a low number on one is not a fault. The page tells a mouse and a trackpad apart from the way they scroll, and the card says which it saw.
The page also cannot see what the mouse was set to, only what arrived. A mouse set to 1000 Hz that shows 125 on the card is being held back somewhere on the way, and the list below is where to look.
What is normal
A wired mouse, or one on its own receiver, usually shows the rate on its box or in its software. A result far below it, with the mouse set correctly, points at the link or the computer rather than the mouse. A rate that wanders between runs, with high jitter, points at a weak or busy wireless link.
Bluetooth mice are often in the warning band, and that is the link working as designed. Run the test on the cable or the receiver if the mouse has one, and compare.
What to try
- Use the cable or the mouse's own wireless receiver rather than Bluetooth.
- Plug the receiver straight into the computer, not into a hub, and keep it close: a receiver on the far side of a metal desk or behind the computer loses reports.
- Set a higher rate in the mouse's software, if it has the setting.
- Turn off power saving for the mouse, and keep its battery charged.
- Run the test again on the same link before you change anything else: a rate that comes back the same twice is the link's, and one that wanders is the signal.