Reaction Time Test

What this test measures

Reaction time here is the delay between the area turning green and your button press, measured with the browser's high-resolution clock. Each round waits a random 1.5 to 4 seconds before switching so you cannot anticipate the change; pressing during the red phase is counted as a false start and the round restarts without a time.

Five rounds are averaged because a single attempt is noisy — one lucky guess or one blink changes it a lot. Your best average is kept in this browser so you can see whether a new mouse, a different display, or simply being more awake shifts the number. Mouse, touch, and the Space bar all work as the trigger.

How the reaction time test works

Our reaction time test runs through a short cycle for each of five rounds. When you click or tap the area, or press the Space bar, the area turns red and a random delay between 1.5 and 4 seconds is drawn. The delay is chosen fresh for every round, so there is no rhythm to learn. At the end of that delay the area turns green and the page records a timestamp from the browser's high-resolution clock at the moment the state changes.

Your next press ends the round. The page takes the timestamp the browser attached to that pointer or keyboard event and subtracts the green timestamp; the difference, rounded to a whole millisecond, is the round's reaction time. Both timestamps come from the same monotonic clock, so system clock changes or timezone differences cannot distort the result. The value is added to the list, and the average and best of the rounds so far are recalculated.

A press while the area is still red is a false start. The round is cancelled, the pending colour change is discarded, the false start counter goes up by one, and the test waits for you to start the round again. No time is recorded for a false start, so it cannot drag the average down. After the fifth valid round the test shows the final average and stops responding to presses until you reset.

Three inputs trigger the test: a left mouse button press, a touch or pen tap inside the area, and the Space bar anywhere on the page. The Space bar listener ignores key auto-repeat, so holding the key down does not fire repeatedly, and it is disabled while you are typing in a text field so it cannot interfere with the rest of the page. If you finish five rounds with a better average than any previous session, that average is saved in this browser.

How to read your result

Three numbers matter. The average across the five rounds is the most representative figure and is the one compared against your saved best. The best round shows your floor — the fastest you managed when everything lined up. The false start count tells you whether you were reacting or guessing: a session with several false starts and a very low average is less meaningful than a clean session with a slightly higher average.

The individual round times are listed so you can spot patterns. If the first round is noticeably slower than the rest, that is usually warming up — your hand was not quite in position or your attention was still settling. If the last round is slower, attention is drifting. If the times swing wildly with no pattern, try again in a quieter moment; the test is sensitive to distraction.

Remember that every value includes fixed hardware delay on top of your own reaction. The monitor needs time to process and draw the green frame, the mouse reports on a fixed polling interval, and the browser delivers events in its own cycle. Those add up to a constant offset for a given setup, which is why absolute numbers from different computers are not directly comparable while numbers from the same machine on different days are.

A gradual improvement over a few sessions is normal and comes mostly from familiarity with the test rather than a genuine change in reflexes. Once your average stops moving, that plateau is a useful baseline. A sudden jump in the other direction — much slower than usual — is more often tiredness or a background process slowing the page than anything to do with your hardware.

When a reaction time test is useful

The classic use is checking yourself before a competitive gaming session. If your average is well above your baseline, you are probably tired or distracted, and knowing that beforehand is better than finding out mid-match. Some players also run a few rounds as a warm-up; the first round tends to be slow and getting it out of the way on a test is cheaper than on a real opponent.

It is also a practical way to compare equipment. Run ten rounds with one mouse, then ten with another, in the same session and with the same display, and compare averages. The biological part of the measurement stays roughly constant over that short period, so a consistent difference between the two sets points to the hardware — polling rate, switch latency, or wireless link — rather than to you. The same approach works for comparing two monitors or for checking whether a game mode or a display setting changes anything noticeable.

Developers and support staff use tests like this to reproduce a user's input lag complaint. A screen that feels slow is hard to describe; a reaction time that is far higher on one computer than on an otherwise identical one is something that can be investigated, for example by checking display scaling settings, wireless receivers, or power-saving modes that throttle USB polling.

Finally, the test is a simple self-check for alertness. It will not replace a medical assessment of any kind, but if you keep a note of your usual average, a morning where it is far worse is a prompt to think about sleep, caffeine, or whether it is a good day to do precision work.

Limits of a browser-based test

The most important limit is that a browser cannot see the physical world. It records the moment it decided to turn the area green, not the moment the pixels actually changed on your screen. Between those two moments sit the browser's rendering pipeline, the graphics driver, the display's own processing, and the pixel response time. On a fast gaming monitor that chain is short; on a television or a cheap office panel it can be considerably longer, and the test has no way to know.

Input has a similar blind spot. A wired mouse at 1000 Hz reports within a millisecond of the press; a wireless mouse in power-saving mode or a Bluetooth keyboard may take noticeably longer, and the browser only sees the event once the operating system delivers it. Touch screens add their own debounce. The timestamp the browser stamps on the event is accurate for when the event arrived, not for when your finger moved.

The page itself competes for the main thread. A heavy tab in the background, a browser extension, or an operating system update running at the same time can delay both the colour change and the delivery of your press by several milliseconds. Keep other tabs closed and avoid running the test while the machine is busy if you want consistent numbers.

Because of all this, the test is best used for relative comparisons — you against yourself, mouse A against mouse B on the same computer — rather than as an absolute measurement of your reflexes. The averages are comparable across sessions on the same setup, and that is where they are most informative.