Spacebar Counter

Five rules decide what goes into the number.

1. Only separate presses count. Every counted press needs the key to come back up first. Keyboard auto-repeat events are recognised by the browser’s repeat flag and ignored, and so is any further key-down that arrives while the key is still held, for systems that do not set the flag.
2. The first press starts the clock and counts. It is press number one of the window, not a free warm-up, and the page says so before you start.
3. The window is set by press timestamps. Each press is stamped with the time the browser recorded the key going down. A press stamped at or after the end of the window is left out even if it arrives before the closing timer runs, so a 10-second result is presses divided by exactly 10 seconds.
4. One input per run. A run started on the keyboard counts only the space bar; a run started on the tap pad counts only taps. Mouse clicks on the pad do not count at all.
5. Trailing presses cannot wipe the result. For 1 second after a window closes, extra presses are ignored, because nobody stops at the exact instant the clock runs out.

Two runs with the same total can feel completely different. The stats under the result describe the gaps between your presses, which is where technique and fatigue show up.

• Presses per second — presses ÷ window seconds on a timed run. In Unlimited mode it is the gaps ÷ the time from first to last press, because N presses spread over that span contain N − 1 gaps.
• Best 1-second stretch — the most presses that fit inside any one second of the run, wherever it started. Your burst ceiling.
• Mean gap and spread — the average time between presses, and its standard deviation. If your gaps were 120, 140 and 130 ms, the mean is 130 ms and the spread is ±10 ms.
• Variation (CV) — spread ÷ mean × 100. In that example, 10 ÷ 130 is 7.7%. Lower is steadier, and because it is a ratio you can compare it between a fast run and a slow one.
• Second half vs first — on 10, 30 and 60-second runs, the change in rate between the two halves. −15% means the second half was 15% slower.
• Presses in each second — a bar for every second of the run, with the busiest second highlighted, so you can see exactly where you slowed down.

A gap under 20 ms is flagged on its own. Releasing a key and pressing it again by hand takes longer than that, so on a keyboard such a gap can mean the switch registered one press twice. Keyboards filter normal contact bounce in their firmware; a worn or dirty switch can still slip a double press through, and that is worth knowing if you are using this page to test a space bar.

The space bar is the longest key on the keyboard and sits under both thumbs, so it can be pressed by either thumb, by both thumbs alternating, or by fingers from both hands drumming on it. A mouse button takes one or two fingers. The key also travels further than most mouse switches and has to rise past its reset point before it can register again, and long keys use stabilisers that can feel different at the edge than in the middle.

All of that makes a spacebar result its own measurement. Compare it with your own spacebar runs on the same keyboard, and use the click speed test when you want mouse clicks per second. Neither number is a reaction time: for the delay between a signal and your response, use the reaction time test.

• One thumb — the simplest to repeat, and the fairest way to compare runs across days.
• Alternating thumbs — left and right thumbs take turns. Each thumb gets more recovery time, which usually helps on the 30 and 60-second timers.
• Two-hand drumming — fingers from both hands alternate on the bar. Treat it as a separate category from thumb runs.

Whatever you use, keep attempts short, take breaks, and stop if you feel pain, numbness or lasting discomfort.

A touchscreen has no space bar, so the pad itself is the button. Each finger that lands counts as a press, which lets you alternate thumbs the way you would on a keyboard. The pad blocks double-tap zoom and the long-press menu so neither interrupts a run. Tap scores are saved apart from keyboard scores, because tapping glass and pressing a switch are different movements.

After a run, Share score copies a link (or opens your phone’s share sheet) that shows your result to whoever opens it, with a prompt to beat it on the same timer. The link carries only the timer, the number of presses and the input you used.

Anyone can edit a link, so a shared score is labelled as unverified, and the page reads it strictly: each value must appear once, the timer must be one this page offers, the count must be a plain whole number, and a timed score above 30 presses a second is rejected. A link that fails any check is ignored with a note saying so. Shared scores are never saved.