Reaction Time Test
Wait for the signal, then react. Five visual trials take about 20 seconds; your median is the headline so one lucky tap does not define the run.
Use the same device and input method when comparing sessions.
—
Tests · Visual 5
Five visual trials · median is the headline.
Results, history & settings
Compare like with like
Finish a Clean session to build a same-mode, same-input local baseline.
What this browser can measure
- Monotonic clockStimulus and response use
performance.now(). - Pointer-down timingMouse, touch and pen are captured on press, not release.
- Device latency stays visibleDisplay, browser scheduling, audio path and input hardware remain part of the observed score.
- No invented percentileYour strongest comparison is repeated Clean runs on the same setup.
Recent sessions
Keep repeat tests comparable
Audio mode includes speaker/headphone latency. Bluetooth can vary. F1 mode measures response to lights-out; it is a different task from the classic visual test.
Last 30 Daily 5 sessions
Classic first
The page opens directly on Visual 5. Extra test types stay one disclosure below the core run instead of pushing the reaction surface out of the first screen.
False starts stay honest
Early responses are refused, the current trial is retried with a fresh deterministic wait, and the run becomes Review rather than silently improving the score.
Exact continuity
Recorded trials, retry count, input, mode and integrity state save locally. An interrupted active session restores paused so a hidden-tab cue never becomes a score.