How This Browser Reaction Time Test Measures Results

The five-attempt flow

Each session contains five successful attempts. When you start an attempt, the test waits for a randomized delay so the ready moment is harder to predict. If you respond before the green signal is ready, the test reports “too soon,” cancels the pending transition, and does not add that activation to the attempt count or statistics.

After each successful response, the page shows the measured milliseconds and current progress. The fifth result ends the session. The summary lists all five values in attempt order and reports the best, rounded arithmetic mean, and median. For five values, the median is the third value after numeric sorting.

How the start time is aligned

The random waiting timer does not become the measurement start. When that timer finishes, the test enters an internal arming state and requests an animation frame. In that frame it commits the green ready state and records a high resolution timestamp from the browser performance clock. The response time uses the same clock domain.

This approach avoids deliberately counting the earlier timer-to-render queue as part of your reaction. The W3C High Resolution Time specification defines the monotonic timing model used by performance.now(), while the HTML Standard defines animation-frame processing within browser rendering.

What the result includes

The number is a browser-observed visual response time. It can still include:

• display refresh and pixel-response delay;
• mouse, keyboard, touchscreen, USB, and driver delay;
• operating-system and browser event scheduling;
• main-thread work near the ready signal or response;
• your perception, decision, and physical input time.

Research comparing simple reaction-time systems shows that hardware and software choices can materially affect measured latency. This test does not calibrate or subtract those device-specific delays, so it does not claim laboratory precision.

Privacy and reset behavior

Attempt results remain in memory in the current page session. They are not uploaded to an account, leaderboard, or server by this feature. Reset clears the five attempts and calculated statistics. Reloading the page also starts a fresh session.

Sharing uses the browser's native share interface when available, otherwise a clipboard copy. Only the generated summary text is shared or copied after you choose the action.

Use the result for informal practice and like-for-like personal comparison, not for medical, occupational, driving-safety, or diagnostic decisions.