FlickLab.

Reaction time

Five trials. Wait for the disc, click it, and the gap between the two is measured to the millisecond.

Five trials · normal signal

No account, no download. Your scores stay on this device.

What this test measures

Reaction time is the delay between a signal appearing and your response arriving at the browser. It is not the same as aim speed: a fast flick with a slow start loses to a slow flick with a fast start, and this mode isolates the second part.

The delay before each signal is seeded and random, between roughly 0.7 and 3.4 seconds depending on difficulty. A fixed delay would let you anticipate the signal and produce a number that is not reaction time at all.

How scoring works

Average reaction
Mean of the valid trials, measured from the simulation step that revealed the disc to the pointer event timestamp of the click.
Score
100000 divided by the average reaction time in milliseconds, multiplied by the difficulty value.
False start
Clicking before the signal ends that trial with no reaction time recorded.
Trials
Five per run. The run ends when the fifth trial is finished; the duration selector does not apply.

Reading the result honestly

A single run is noisy. Five trials give a usable average, but the difference between 210 ms and 230 ms in one session is usually posture, sleep and attention rather than a change in ability. Compare weekly averages, not run-to-run numbers.

False starts are worth watching. One is normal; three in a row means you are predicting the delay instead of waiting for the signal, and the average you get from the remaining trials is not trustworthy.

How to practise

Questions

What is a good reaction time?

Typical adult visual reaction time is around 200 to 250 milliseconds on a computer. Under 200 ms is fast; under 170 ms usually means anticipation or a very favourable setup.

Why five trials?

Enough to average out one bad trial without turning the test into a stamina exercise. Five takes about fifteen seconds including the delays.

Why is the delay random?

A fixed delay can be predicted and clicked before the signal appears. Randomising it forces a genuine response.