Reaction time test

Exactly what this test measures

Each round, the panel goes dark and holds for a random delay between 1.5 and 4 seconds. When it flips to cyan, a timestamp is recorded, and the moment your pointer goes down a second timestamp is taken. The difference, rounded to the whole millisecond, is your time for that round. You play five rounds and your score is the average of the five. Lower is better, and your best average is saved in your browser so you can chase it later.

One detail matters more than it sounds: the clock stops on pointer down, not on a completed click. You do not need to press and release, you only need to make contact. That is deliberate. A full click adds the release of the button to your time, which has nothing to do with your reflexes.

How to actually get faster

Clicking early does not cost you anything except a restart of that round, so the temptation is to gamble. Resist it. The delay is uniformly random across the 2.5 second window, which means there is no rhythm to learn and no tell to read. Guessers occasionally land a 90 ms fluke, then post a 380 on the next round after the panel scolds them, and the average absorbs both.

What does work: hover your finger a millimeter over the button with light pressure already loaded, soften your gaze on the center of the panel rather than staring hard, and react to the change in brightness rather than consciously registering "that is cyan." Color naming is slow; luminance detection is fast. Also, the between-round screen waits for you, so breathe. Rushing into round three carries tension from round two, and tense forearms are slow forearms.

What your average means

The median for a simple visual test like this one sits around 250 ms. Anything under 220 is quick, under 200 is genuinely rare, and a sub-150 average almost always means anticipation crept in. Before comparing yourself to anyone, compare setups. A 60 Hz monitor redraws every 16.7 ms, so the cyan frame itself can appear late. Wireless mice buffer, trackpads debounce, and phone touchscreens add their own scan latency. The same person can swing 30 to 50 ms between devices. Your number is only meaningful against your own number on the same hardware, ideally at the same time of day, since sleep and caffeine move reaction time more than practice ever will.

Design notes and honest limits

The test uses the browser's high resolution timer, which is as precise as a web page can get, but it cannot see past the browser. Display latency, input hardware, and the operating system's event pipeline all sit inside your measured time, which is why we average five rounds instead of crowning a single attempt: one blink or one lucky twitch should not define the score. There is no leaderboard here and no server-side record. Your best lives in your own browser's storage, so clearing site data resets it, and a score from your gaming rig will not follow you to your phone. If you want a fair rivalry with a friend, use the copy button under the score and make them play on the same machine.