Start a timed run, Alt-Tab away for a second, and the test ends with “this run has no valid score.” That can feel like a glitch — but it is the opposite. A run that lost your input partway through cannot be judged fairly, so the tool voids it on purpose rather than reporting a number it cannot stand behind. We triggered the four most common interruptions in three browsers to show exactly what happens, and whether it is consistent.
It is: Chrome, Safari and Firefox behaved identically on every one.
Four interruptions, three browsers, one result each
Setup: macOS 15.7.7; Chrome 152, Safari 18.6, Firefox 155. Each interruption was triggered during an active run; the status text below is verbatim and was the same in all three browsers.
| What you do | What the tool reports (all three browsers) |
|---|---|
| Alt/Cmd-Tab to another app (window loses focus) | Window lost focus. This run has no valid score. |
| Switch to another browser tab | Window lost focus. This run has no valid score. |
| Press Escape (keyboard check) | Run canceled with Escape. This run cannot be judged. |
| Press Tab (keyboard check) | Focus navigation ended this run. |
Why voiding the run is the honest choice
- Lost focus / hidden tab: once the window is not focused, input events are unreliable — some never arrive. Counting a partial run would produce a score that looks real but isn't, so the run is voided instead.
- Escape: an explicit cancel — you asked to stop.
- Tab: moves keyboard focus out of the test area, which ends a keyboard run by design rather than silently losing keys.
Crucially, an interrupted run is not dropped in silence: it is saved with its reason, so your exported report shows why a run did not count. That is the difference between a tool that hides its limits and one that documents them.
How to get a clean run
Keep the window focused, close overlays that might steal focus, and finish the run before switching away. If runs keep getting interrupted on their own — without you leaving the window — that is itself worth noting with your report. Start a fresh click speed test or keyboard check once the window is settled.
Single-session captures on one macOS machine; interruptions were triggered through an automation tool. This documents how the browsers handled focus and cancellation, which was identical across the three engines tested. See how we test for the full method, and the open dataset for the verbatim status text across engines.