Look for repeatable patterns, not a one-click diagnosis
Jitter is an experience, while the browser exposes only event timing and delta values. Stroke segmentation and robust statistics can organize clues, but they cannot reveal the physical cause behind them.
How to use the Mouse Scroll Jitter Test
Choose an input profile
Select Notched for deliberate wheel steps or Smooth for free-spin and trackpad-like motion.
Make several one-way strokes
Use one direction, pause briefly between strokes, and collect at least 30 valid events across three strokes.
Repeat candidate patterns
Note reason codes and sample numbers, reset, and see whether a similar pattern returns under the same conditions.
What the scroll jitter test calls a candidate
An isolated reverse candidate is an opposite-sign event with baseline-sign events immediately before and after it. A reverse cluster is a short group of opposite-sign events. A long-gap candidate marks timing that is unusually large relative to the session, while an amplitude candidate uses the median and median absolute deviation rather than a fragile average. Each result includes a reason code and sample numbers so the observation can be explained and repeated.
The scroll jitter test filters near-zero deltas from analysis because tiny numeric noise can dominate sign counts, but the raw event log remains available in the developer tool. It starts a new stroke when adjacent valid events are more than 180 milliseconds apart. These centralized thresholds make results repeatable in code; they do not turn a browser pattern into proof of an encoder fault.
Jitter, skipping, reverse input, and inertia
Jitter usually describes unstable movement, such as a brief reversal or an uneven jump. Skipping describes a missing or unexpectedly large movement from the user’s perspective. Reverse input is specifically an opposite-direction signal relative to a calibrated one-way action. Inertia is continued motion after a touchpad gesture or free-spin action. These experiences can overlap, but treating every reversal as mouse scroll jitter creates false alarms.
Smooth and inertial sources receive conservative thresholds because their event streams naturally contain more small values, changing intervals, and decaying motion. Stop momentum before each stroke and do not deliberately reverse. If you need to explore normal direction changes, use the main test’s Free mode instead of the jumpy scrolling test.
Why robust statistics help
An average can be pulled strongly by one pause or one large event. The median better represents the center of a skewed series, while MAD and IQR describe spread without assuming a normal distribution. The scroll jitter analysis uses those summaries to find values far from the session’s typical pattern. It also requires context around a reverse event and minimum sample quality before returning an interpretation.
Robust does not mean infallible. A deliberate flick, application slowdown, tab focus change, wireless interruption, browser scheduling, or acceleration rule can still create an outlier. The value of the mouse scroll jitter result comes from controlled repetition and from comparing event patterns with what you felt on screen, not from the presence of a colored warning alone.
Reduce false positives in a mouse scroll jitter run
Use the same browser and settings, keep the tab visible, place the device on a stable surface, and perform one-way strokes with clear pauses. Select the correct movement profile. Avoid touching a touchpad while using a mouse and disable no software merely for the test unless you document the change. Collect more than the minimum when possible; a longer stable sample makes typical timing and amplitude easier to characterize.
Repeat the scroll jitter test at least twice. If candidates move randomly or disappear when your technique becomes steadier, user motion or inertia may explain them. If a comparable pattern appears across browsers and another computer, the evidence for an input-path issue becomes stronger, but it still does not identify dirt, wear, firmware, radio conditions, or a particular component.
Safe next steps for jumpy scrolling
Check system scroll settings, application-specific profiles, browser extensions, official driver or firmware updates, and wireless charge or receiver placement. Compare a wired or alternate connection mode if the manufacturer supports it. Test in another application. For external cleaning, disconnect the device and follow its manual; do not spray liquid into the wheel opening or use general-purpose lubricants.
Stop when troubleshooting would require disassembly, solvents, soldering, or a step that could affect warranty coverage. A qualified technician or manufacturer support can inspect components that a webpage cannot. Bring repeated scroll jitter test exports and notes about the environment as observations, not as a final diagnosis.
Mouse Scroll Jitter Test questions
Why does the scroll jitter test require 30 events?
Robust timing and amplitude summaries need enough observations and multiple strokes. Below the minimum, the tool reports insufficient data instead of overinterpreting noise.
What does REVERSE_BETWEEN_BASELINE_EVENTS mean?
One opposite-sign sample appeared between two samples matching the calibrated sign. It is a repeatable description, not a diagnosis.
Should a trackpad use Notched mode?
Usually Smooth is safer because touchpad motion and inertia create dense, gradually changing events that need more conservative thresholds.
Does no candidate mean the wheel is perfect?
No. It means this session did not contain an obvious candidate under the implemented rules. Missing input and intermittent issues may not appear.