Mouse Polling Rate Estimate
Estimate how frequently pointer-move events reach this browser. This cannot read or verify the mouse's USB polling-rate setting.
Move the pointer continuously here
Use steady circles or side-to-side movement until at least 200 intervals are collected.
0
Samples
—
Median interval
—
P95 interval
—
Estimated event rate
Start, then move the pointer continuously inside the area.
Interval distribution
P10–P90 stability range: —
Coalesced Events API: not exposed by this browser
How to use
- Choose Start new sample, then move the pointer continuously inside the test area.
- Collect at least 200 intervals while keeping the page focused and the movement steady.
- Stop the sample or allow it to stop at 1,200 intervals, then review validity before saving a local report.
Read the results
- Median and p95 summarize browser-delivered event intervals; estimated Hz is 1,000 divided by the median interval.
- The distribution and P10–P90 range show how tightly recent delivery intervals cluster.
- A valid label only means the browser sample met the page's sample-count, focus, and delivery-density rules.
Limits and retesting
- Browser event coalescing, display refresh, system load, pointer speed, operating-system scheduling, and background throttling can change the estimate.
- Sparse delivery above a 20 ms median is rejected because it cannot support a useful event-rate estimate.
- Use manufacturer software or device documentation to inspect configured hardware polling rate.
Common situations this test can investigate
Use the page to collect a repeatable browser observation before changing cables, settings, software, or devices.
The estimate is lower than the mouse setting
Compare browser delivery with the configured value without assuming they should match. Rendering cadence, coalescing, scheduling, and pointer movement can lower the observed estimate.
Repeated runs vary widely
Use the p95 and P10–P90 range to see whether delivery was uneven, then repeat with the same motion after closing heavy background work.
The result is marked invalid
Keep the page focused, move continuously, collect at least 200 intervals, and avoid a run whose median delivery interval exceeds the disclosed 20 ms validity boundary.
A repeatable troubleshooting sequence
- 1Confirm the manufacturer's configured rate separately, then close overlays or applications that may compete for rendering time.
- 2Start a new sample and move the pointer in continuous circles or long side-to-side strokes until at least 200 intervals are collected.
- 3Review median, p95, the distribution, and P10–P90 together; do not compare only the rounded estimated Hz.
- 4Repeat three runs using the same browser, display, power mode, pointer speed, and movement pattern before changing one variable.
- 5Compare another browser or the manufacturer's device tool if the estimate remains sparse or inconsistent.
Browser event reference
- Sample
- A positive elapsed interval between consecutive pointer events exposed to this page.
- Median interval
- The middle sampled interval; used as the stable input for the browser event-rate estimate.
- P95 interval
- A slower-end interval: 95% of retained samples are at or below approximately this value.
- Estimated Hz
- 1,000 divided by the median interval. It describes this browser run, not USB configuration.
- P10–P90
- A central stability range that reduces the influence of the fastest and slowest tails.
Frequently asked questions
Why can a high-rate mouse show a much lower browser estimate?
Browsers and operating systems may combine or schedule pointer events around rendering. Display refresh, system load, pointer speed, power settings, and background throttling can also limit delivery before this page observes it.
Is estimated event rate the same as USB polling rate?
No. The value is calculated from browser-delivered pointer-event intervals. It cannot read device firmware or prove the rate configured in manufacturer software.
Why does the test require at least 200 samples?
A very short run is sensitive to startup movement, a few delayed events, and ordinary scheduling variation. Two hundred intervals provide a more useful browser-level median and p95, while still remaining an estimate.
What does the Coalesced Events message mean?
Some browsers expose multiple underlying pointer samples through getCoalescedEvents(). Support does not guarantee grouped samples will appear in every run, and lack of support does not imply a mouse problem.