Mouse polling rate vs browser event rate
A mouse may advertise 1000 Hz while a web page reports a lower or less stable number. That does not automatically mean either number is wrong.
Two layers are being compared
USB polling rate describes how often the host and mouse communicate at a lower level. A normal web page does not receive those raw USB reports. It receives pointer events after the operating system, browser and compositor have processed input. BenchRush therefore labels its number a browser event-rate estimate.
Why the browser number moves
Slow mouse movement may not generate enough distinct positions. Browsers can coalesce movement, power-saving modes can alter scheduling, a busy CPU can delay delivery and background tabs can be throttled. Very high-rate devices can also expose more variation because the browser is not a USB analyzer.
How to make runs comparable
- Keep the test tab visible and focused.
- Move the pointer continuously in fast circles inside the test area.
- Use the same browser, display and pointer settings for comparisons.
- Run the test several times and look for a stable range.
- Use vendor software or lower-level tools when you need to verify the configured USB polling rate itself.
What is the useful question?
For a browser utility, the useful question is often “does pointer-event delivery look broadly consistent on this setup?” rather than “can JavaScript certify my USB setting?” A stable browser estimate can reveal a meaningful change after switching ports, power modes or software settings even when it is not a raw hardware measurement.
Pair the result with reaction testing
Polling is only one part of the input path. If you are comparing setups, use the same configuration for a reaction-time baseline and avoid interpreting a single Hz number as total latency.