Mouse Polling Rate Test

How to use

1. Choose a 3, 5, or 10 second capture. Pointer Lock is optional and only helps reduce cursor clipping during fast movement.
2. Press Start and, after the capture begins, move a physical mouse continuously inside the focused test area. The page prefers raw pointer updates when the browser exposes them and otherwise uses pointermove.
3. Stop freezes the browser-exposed samples. A blur, hidden page, or unexpected required Pointer Lock exit interrupts the run and does not save history.
4. Review the browser-observed update rate, measurement path, interval distribution, sample status, exclusions, and dispatch/coalescing context. The result is not a USB or sensor report.
5. Reset clears the current run but preserves aggregate local history. Repeat with the same browser, device, motion pattern, display, and system load for comparison.

What the result means

• Browser-observed update rate is 1000 divided by the median interval between accepted, nonzero, mouse-like browser updates from one measurement path. Mean, P95, and minimum intervals show other parts of the accepted timing distribution.
• Raw pointer updates are preferred when supported; pointermove is the fallback. If a dispatch exposes coalesced PointerEvents, their individual timestamps supply the detailed stream and the outer dispatch is not counted again.
• Zero movement, invalid intervals, and gaps longer than 250 ms are excluded and counted. Touch and pen samples are excluded from primary stats; empty or unknown pointer types are accepted as mouse-like and counted separately.
• Browser dispatches and browser-exposed updates remain separate context. Sample status describes the amount of usable browser evidence only; it is not an accuracy, hardware-health, or outcome grade.

Limitations

• Browser pointer updates are scheduled and may be coalesced; pointerrawupdate support still does not expose raw USB polling. Touch and pen samples are excluded, while some trackpads may appear mouse-like.
• The primary stream is browser-observed PointerEvent timing from pointerrawupdate when supported or pointermove as fallback—not USB, HID, firmware, sensor, or device-report telemetry.
• Event.timeStamp and browser-exposed coalesced events provide more detailed browser evidence, but timer precision, browser implementation, scheduling, and coalescing can still change what is observed.
• Operating-system load, browser work, CPU scheduling, display timing, movement pattern, focus, background activity, and device behavior can all affect the sample stream.
• Pointer Lock only reduces cursor clipping; it does not change the measurement into a hardware polling report. Permission denial continues as an unlocked run.
• Touch and pen samples are excluded, but pointerType is a browser label and some trackpads may appear mouse-like. This page does not certify a physical mouse.
• Zero movement, non-positive intervals, and gaps over 250 ms are excluded from interval statistics. A short or paused run can have no usable rate.
• A high peak or a lower browser-observed rate is not proof of a good, bad, failed, or misconfigured device. Do not map results to a nearest hardware tier.
• Fast device reports such as 2,000, 4,000, or 8,000 Hz can be under-observed by a browser event stream; theoretical intervals are shown only as context.

Troubleshooting

Observed Hz is blank or low

Move continuously with nonzero motion, keep the surface focused, and allow enough time for accepted intervals. Reset and repeat with the same motion path and conditions.

A specified 1,000, 4,000, or 8,000 Hz mouse looks lower

Browser dispatch, scheduling, coalescing, display timing, and system load can reduce observed events. This page cannot certify the USB or sensor rate.

Coalesced browser updates were returned

Their individual browser timestamps replace the outer dispatch in the detailed timing stream. Dispatch counts remain separate, and neither value reveals raw USB reports.

Pointer Lock is denied

Continue unlocked or allow Pointer Lock and retry. It is only a clipping aid, so the browser-observed measurement remains available without it.

Results vary between runs

Use the same duration, mouse, browser, system load, focus state, and movement pattern, then compare aggregate rows in Local history. Variation is expected in browser timing.

Related tools

• Mouse Test

  Use this online mouse tester to check buttons, side buttons, held states, wheel events, and pointer movement. See browser counts, hold times, movement path, and session status.

• Mouse DPI Test

  Estimate mouse DPI/CPI from Pointer Lock movement over a measured ruler distance. Compare three trials, median, drift, input mode, and an optional configured DPI.

• Scroll Wheel Test

  Check browser-received wheel direction, reverse events, horizontal input, delta units, relative magnitude outliers, and optional wheel-event cadence.

• Mouse Double Click Test

  Check a mouse for unintended repeat clicks with slow single-click screening. Inspect browser-observed intervals by button, then optionally measure intentional double-click timing.

FAQ