The test measures the browser's main thread for 10 seconds idle, then for 30 seconds with the CPU stage's workload on every core but one.
- Three probes run throughout: a timer ping every 50 ms, a synthetic input every 200 ms timed to the next frame, and every frame interval.
- A loaded 95th percentile delay under 50 ms reads as responsive; that line is ours. At or above 200 ms, Google's Interaction to Next Paint good line, it reads as unresponsive.
- It measures input handling inside the browser. It cannot see the mouse, the USB path or the display's own delay.
The short answer
When a computer is busy, it can still be quick at the work and slow to respond to you. The responsiveness test measures that second thing: how late the browser gets to your input and to the next frame, first with the machine idle and then with every core but one working. The difference is the delay the load adds.
What the three probes measure
All three run on the page's main thread, which is the thread that handles clicks and draws the page.
- Ping: how late a timer and a message callback run after they were due. A busy main thread shows as delay.
- Input: the time from a synthetic pointer event to the next animation frame, the shortest path from an input to a paint.
- Frames: every frame interval. One longer than 1.5 times the display's normal interval counts as missed.
Reading the result
The idle phase is the machine's own baseline in this browser, so the loaded figures are read against it rather than against another computer. A small gap means the system kept the interface responsive while the processor was busy. A large gap means work competed with input. One core is left free for the main thread on purpose; if workers cannot start, load and probes share one thread, and the result is flagged as not comparable.
What it does not measure
End-to-end input lag runs from the mouse or keyboard, through USB and the operating system, to the browser, and then to the display's own processing and pixel response. Only the browser's part is visible to a web page. Measuring the whole chain needs a high-speed camera or a dedicated latency tester.
Terms used here
Questions
- Is this the same as a mouse or monitor input lag test?
- No. It measures the delay the computer adds inside the browser under load. The mouse's polling, USB and the monitor's processing are outside what a web page can time.
- Does the responsiveness test affect my score?
- No. It runs on its own page, is not a stage of the Standard or Deep run, and contributes nothing to the score.