Browser input diagnostics · browser-wheel-events-v2
Find out how fast you can scroll — in pixels per second and scrolls per second, with a rating — or run the guided diagnostic to catch a wheel that scrolls in reverse, skips, stutters or drops input. Everything runs in this tab.
Local measurement workbench
Normal pace · down
After arming, the first wheel event starts a 5 s window. Keep the declared down direction and normal pace until it ends.
0 recent browser events shown. Up events appear above the axis, down events below it, and horizontal-only and true-zero events use distinct marks on the axis.
Timer starts on the first wheel event. Events at or after the exact boundary are excluded. While armed or recording, wheel input anywhere on this page is captured. Press Escape or use Cancel run to stop.
One page per question
The workbench above answers every scroll question at once. These pages each answer one, with the tool already set to the right mode and only the numbers that question needs.
Three questions, one instrument
A direction check, a repeatability comparison and a speed challenge answer different questions. Results are kept separate by intent, duration, input type, setup and metric version.
Run controlled down-and-up phases at a normal and a faster pace. The report separates up, down, horizontal-only and true-zero events, flags reversals inside a declared phase, and marks an under-sampled phase inconclusive rather than guessing.
Run the wheel diagnostic →Complete at least three protocols before and after a change. Compatible completions are kept locally for comparison and are never mixed across setup, input type, duration or metric version.
Read the methodology →Hover and go: no setup, no button to press. You get pixels per second, scrolls per second, a Slow-to-Extreme rating and the best single second of the run.
Take the speed test →Beyond the scroll wheel
The same local-first approach, applied to the rest of your setup. Every tool runs entirely in this tab — no installs, no uploads, no sign-in.
Read the signal, not a verdict
Browser observations are useful for retesting a symptom and comparing the same setup. They are one piece of evidence, not a universal hardware grade.
Check browser compatibility and limits →Before you diagnose
Keep the browser, device, operating-system settings and test duration the same when comparing runs.
On a detented mouse wheel, 4–7 scrolls per second is normal browsing pace, 8–12 is fast, 13–20 is very fast, and anything above 21 usually means a free-spin wheel or a high-resolution wheel that emits several events per notch. Touchpads sit higher again because a single flick can generate a long tail of momentum events. Compare your own runs rather than chasing someone else’s number.
Measured as distance rather than events, ordinary browsing runs at roughly 600–1,500 CSS pixels per second — about one to two laptop screens a second. Skimming usually lands between 1,500 and 3,500 px/s, and a hard free-spin flick can exceed 7,500 px/s. The figure moves with your operating system’s lines-per-notch setting, so treat it as a same-machine comparison rather than a universal constant.
The ceiling is set by your wheel and your operating system rather than your hand. A free-spin wheel released at full speed can push tens of thousands of pixels per second for a second or two, while a detented wheel is limited by how quickly the detents pass the sensor. Run the scroll speed test and read the best one-second burst — that peak measures your ceiling far better than a whole-run average.
No. A browser wheel event is a software-observed input message. One detent can produce different event sequences depending on the mouse, operating-system settings, driver, browser and free-spin mode. Touchpads have no physical wheel notches at all.
A consistent flip is usually a setting rather than a fault: macOS calls it “natural” scrolling and Windows exposes a per-device flip in Device Manager. A wheel that reverses only sometimes, in the middle of a steady scroll, points instead at a dirty or failing encoder. The mouse wheel test runs declared up and down phases and reports any events that travelled the other way, so you can tell a consistent flip from an intermittent one.
WheelEvent.deltaMode tells the page how to interpret deltaX and deltaY. ScrollTest keeps those browser-reported units separate instead of relabelling every value as pixels or inventing a conversion to physical movement. The speed tools sidestep the problem by measuring how far a real scroll container actually travelled, which the browser always reports in CSS pixels.
No browser test can see a physical movement that produced no browser event, and software can alter event timing. The diagnostic can highlight observed reversals, gaps or bursts worth retesting. Confirm a persistent symptom in another browser, port or computer before drawing a hardware conclusion.
Raw events stay in memory in this browser tab unless you explicitly download an export. Compatible result summaries and personal bests may be saved in local storage for history and repeatability comparisons until you clear them in the workbench or remove site data.