blitz50 tools

Reaction time

Five lights come on. When they go out, tap. Five goes, and the average is your score.

How it works

Five lightsOne at a time, then a hold

They go outThat is your signal

243ms

Five goesThe average is the score

The rules in full

Five red lights come on one at a time, a Formula 1 start. When the last one is lit they hold for a moment you cannot predict — somewhere between 0.9 and 3.4 seconds — and then all five go out at once. Tap, click or press space the instant they do.

You get five goes, and your score is the average in milliseconds. Tapping before the lights go out is a jump start: that go simply runs again, so a twitch costs you nothing but the wait.

The holds are generated from a seed the server issues, not by your browser, which is why a link you share deals the receiver the same five waits in the same order.

What counts as a good time?

  • <200msvery fast
  • 200–250msquick
  • 250–300msordinary
  • 300ms+tired, or on a phone

Anything under 90ms is rejected: that is faster than sight, so it was a guess that happened to land rather than a reaction.

A simple visual reaction is the sum of fixed costs: light reaching the retina, the signal crossing to the visual cortex, a decision, and the motor command travelling back down your arm. Measured in a lab that adds up to roughly 180 to 250 milliseconds, and practice moves it only a little. Sound is faster than sight by about 40ms, which is why sprint starts use a gun.

Screens add their own delay — a phone's touch sampling and a display's refresh together cost tens of milliseconds — so a browser test reads slower than a laboratory one. Everyone here is measured through the same overhead, which is what keeps two scores worth comparing.

The same waits for both of you

The hold before the lights go out is random every go, not fixed for the day — a wait you have already sat through is not a reaction test. A link you share replays your exact five holds, so a head-to-head is still fair.

Also here