Reaction Time Test

What the number on the screen is actually made of

A browser cannot measure your reaction time. It can measure the gap between the frame in which a stimulus was drawn and the timestamp the browser attached to your click, and that gap contains three things: you, your display, and your input device. Every reaction test on the web reports that sum and calls it the first term.

The two parts that are not you are small but not negligible. A 60Hz screen paints once every 16.7 milliseconds, so the green you are waiting for becomes visible somewhere inside that window rather than at the instant the page decided to show it. A mouse polled at the usual 125Hz waits up to another 8 milliseconds before it mentions that you clicked.

Added together that is a systematic offset of roughly twelve milliseconds on a typical desktop, and rather more on a phone. This site works it out from your own measured refresh rate and prints it beside your score instead of quietly folding it in, which is the only reason a number here can be compared honestly against a number from somewhere else.

Reading your own result

The headline figure is the mean of your five rounds, because averaging is what makes a single lucky click stop mattering. The median is shown next to it, and when the two disagree by much it is telling you that one round went badly — a blink, a glance away, a moment of doubt — rather than that your reaction time is unstable.

The margin printed after the score is not padding. It is half the width of the interval your equipment leaves open, shrunk by the square root of five because averaging independent rounds narrows random error. If your score is 268 with a margin of 5, the honest reading is that a repeat of this test on this machine would very likely land between 263 and 273.

The percentile underneath is the part people actually came for, and it is computed against attempts made in the same input mode as yours rather than against everybody. Until a mode has enough attempts to draw a curve from, the chart says so in plain words instead of showing you a shape.

Why you score differently on two of your own devices

Move from a 144Hz gaming monitor to a 60Hz laptop panel and your measured time gets worse by around six milliseconds without anything about you changing. The stimulus simply spends longer waiting to be shown. This is the single largest reason two scores from the same person disagree, and it is invisible on every site that does not report the refresh rate it measured.

Input device is the second reason. A key press is usually a little faster than a mouse click for the same person, partly because the finger is already resting on the key, and a touchscreen is slower than both because the panel samples on its own schedule and then has to decide the contact was a tap rather than the start of a scroll.

That is why the results here are kept in separate distributions per input mode and never pooled. Comparing a phone tap against a desktop click is comparing two different measurements, and averaging them together produces a curve that describes nobody in particular.

This is a measurement toy, not a medical or psychological test. No score here says anything about your health, attention or cognition, and it is not a screening tool of any kind.

Questions

What is a good reaction time?

For a visual click test in a browser, most adults land somewhere in the 200 to 300 millisecond range, and anything under about 200 is genuinely quick. The chart on this page is the better answer than any single number, because it shows the whole spread rather than one threshold.

Why is my reaction time worse than it used to be on another site?

Usually the equipment rather than you. A slower display refresh, a different input device or a phone instead of a computer each move the measured figure by several milliseconds, and most sites neither measure nor mention any of it.

Does the test work on a phone?

Yes, and taps are recorded as their own input mode with their own curve, because touch hardware adds more delay than a mouse does. Pinch-zoom is left enabled and nothing is hidden behind a hover.