How the measurement works
October 5, 2026 at 14:18
A reaction time test is only as trustworthy as its clock. This page explains exactly what Tap the Cap measures, what it cannot measure, and how much of your score is you rather than your hardware.
The timing loop
Each round follows the same sequence:
- You start a round. The game enters a waiting state and picks a random delay between 1000 and 4000 milliseconds.
- When the delay expires, the browser repaints the mushroom cap in a new colour and the game records a timestamp. This timestamp is the start of your measurement.
- Your click or key press fires an event. The game takes a second timestamp and subtracts the first. The difference, in milliseconds, is your score.
The random delay is not decoration. If the wait were fixed, you could count it out internally and press at the right moment without ever reacting to anything. Randomising it forces the response to be genuinely stimulus-driven. The range also matters — too short and you would still be settling after the previous round, too long and attention drifts.
Why the colour changes rather than appears
A new object appearing on screen is detected partly by peripheral vision, so you would react before you consciously identified it. A colour change on an object you are already staring at forces the visual system to process the same shape twice and notice the difference. It is a cleaner stimulus, and the resulting numbers are more comparable between attempts.
Consecutive rounds also never use similar colours. The new hue is always at least 60 degrees away on the colour wheel from the previous one, so the change is unambiguous even for players with reduced colour discrimination.
What the number actually contains
Your displayed score is not purely biological. It is the sum of several delays, only some of which belong to you:
| Stage | Typical cost | Yours? |
|---|---|---|
| Frame rendering (60 Hz display) | 0–16.7 ms | No — hardware |
| Display panel response and processing | 2–20 ms | No — hardware |
| Retina and visual pathway | ~40–70 ms | Yes |
| Decision and motor command | ~60–100 ms | Yes |
| Nerve conduction and muscle contraction | ~30–50 ms | Yes |
| Input device and USB polling | 1–15 ms | No — hardware |
| Browser event dispatch | 1–10 ms | No — software |
In practice, somewhere between 10 and 60 milliseconds of your score is equipment rather than physiology. A 144 Hz gaming monitor with a wired mouse genuinely produces better numbers than a 60 Hz office display with a cheap wireless one — on identical reflexes. This is why comparing your score against a friend on different hardware is only roughly meaningful, while comparing your own scores over time on the same setup is quite reliable.
Why sub-100 ms scores are not real
Nerve conduction is not instantaneous. Signals travel along motor neurons at a finite speed, the muscle needs time to develop force, and the switch in the mouse or key needs to physically close. Adding up only the unavoidable stages gives a floor of roughly 100–120 ms for a simple visual reaction.
This is not a niche opinion — competitive athletics encodes it in the rules. World Athletics treats a sprint start faster than 0.100 s after the gun as a false start by definition: the assumption is that the athlete moved in anticipation rather than in response, because responding that quickly is not possible.
On this site: the ranking rejects submissions below 100 ms for exactly this reason. If you see a time under 100 ms anywhere on the internet's reaction tests, it is a lucky guess during the waiting phase, a bug, or a manipulated submission — not a reflex.
Where your scores are stored
Your personal best times and your last three attempts live in your browser's local storage. They never leave your device, they are not tied to an account, and clearing your browser data deletes them permanently. There is no way for us to recover them, and no way for us to see them.
Only a time fast enough to enter the global ranking is sent to the server, and the only thing sent is the number itself — no name, no identifier, no profile. See the privacy policy for the full detail, and the leaderboard for the ranking itself.
Keyboard or mouse?
Both work, and the Space bar is usually marginally faster. A key press has a shorter mechanical travel than a mouse click for most people, and keyboard events tend to reach the browser with slightly less overhead. The difference is small — on the order of a few milliseconds — but if you are chasing a personal best, pick one input device and stay with it so your numbers stay comparable.
