FlickLab.

Tracking

The target does not wait for you. Hold the cursor inside it and the score is the time you managed to stay there.

Normal speed · 30 seconds

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

What Tracking trains

Flicking is a discrete movement with a clear end. Tracking is a continuous correction with no end at all, and it fails differently: the cursor overshoots, then over-corrects, then oscillates around the target while the target keeps moving. That oscillation is the habit this mode exposes.

The measurement is unforgiving in a useful way. There is no partial credit for being close, and no click to reset the situation. Either the cursor is inside the target at that instant or it is not.

How scoring works

On target
The share of simulation samples where the cursor was inside the target. The simulation samples 240 times per second, independent of your frame rate.
Score
On-target percentage multiplied by the difficulty value (Easy 1, Normal 1.5, Hard 2), then by ten.
Time on target
The same share expressed in seconds of the run.
Clicks
Clicks are ignored in this mode; the cursor position is the input.

Speed, not size

Difficulty changes how fast the target travels, not how large it is. A slower target with a small radius rewards precision; a fast target with a normal radius rewards prediction. Both are useful, but they teach different things, and mixing them up makes progress hard to see.

The path is generated from the run seed, so two runs at the same difficulty and seed follow the same route. That makes it possible to compare attempts fairly instead of guessing whether the target was simply easier this time.

How to practise

Questions

Why do clicks not count?

Tracking is a position task. Adding clicks would let a player pause the cursor and convert it into a flick drill, which is a different skill.

What is a good on-target percentage?

Above 60 percent at Normal is a solid starting point, and above 75 percent is strong. The absolute number matters less than whether it rises across a week at the same speed.

Does the target follow the same path every run?

With the same seed, yes. The seed is shown on the results screen, and the layout is generated deterministically from it.