Editorial policy

Two kinds of figure, held to different standards: measurements this site takes, and figures it reports from elsewhere.

A measurement is anything the instrument produces from your own input — a rate, an interval, a gap between presses. Those are computed in your browser from performance.now() and are reproducible on demand: run the same interval again and you will get a comparable number. The method is published on the About page and the clock source and resolution are printed under the instrument itself.

A reported figure is anything quoted from outside — an average, a world record, a switch specification, a game's combat behaviour. Those carry the standard described below, which is lower, and are labelled as such wherever the distinction matters.

• Every rate is stated with its interval. A clicks-per-minute figure without the span it was measured over is meaningless — 480 CPM from a one-second burst and 480 CPM held for a minute describe very different hands — and we treat quoting one without the other as an error, not a stylistic choice.
• The clock is disclosed. Timing comes from performance.now() read at the input event, to 0.1 ms. The clock starts on your first press, so reaction time is never charged against a rate.
• Ranges, not single numbers, for technique. Where a figure varies by hand and hardware — what jitter or butterfly clicking produces — we publish the range a practised hand reaches rather than a headline figure.
• Hardware limits are named as limits. Above roughly 1,000 CPM on one button the reading describes a switch rather than a person, and the calibration scale says instrument-limited instead of naming a human tier.

It certifies nothing, and it says so on the pages where people most want it to.

Click-speed records live on open submission sites. No governing body certifies a single-button record, submissions are not independently timed, and several widely quoted figures fail a single-button standard on inspection. Where we report those numbers we say where they came from and what is wrong with them — we do not present them as established.

A reading taken in your browser is not evidence of a record either. It is a measurement of your browser, on your hardware, at that moment, and we will not issue a certificate, badge or verification for one. The daily board filters obvious autoclicker output, and that filtering is a defence against nonsense rather than a claim of authenticity.

Hardware specifications — debounce windows, switch reset times, USB polling intervals — are published manufacturer and standards figures, quoted as ranges. This site has not measured them in a lab, and does not claim to have.

If a number here is wrong, we want to know, and the fix is a change to the page rather than a note at the bottom of it.

Report an error through the contact page. A measurement bug is most useful with the browser, the device, the interval and the reading you expected — the instrument stores nothing server-side, so your message is the only evidence available.

Substantive corrections to an article are made in the article and its published date is left as it is, so the change is visible in the page's history rather than hidden behind a fresh timestamp. Anything that changes what a figure means gets a line in the text saying what changed.

• No advertising. There are no ads on this site and no plan to add any.
• No affiliate links and no sponsored content. Where hardware is named — a switch manufacturer, a mouse behaviour — nobody has paid for the mention and no link earns a commission.
• No gated results. Every tool, every figure and every article is available without an account, a sign-up or a paywall.
• Nothing is collected to sell. Measurements stay in your browser. The single exception is the anonymous daily top score, described in full on the privacy page.

Articles carry the byline The Clicks Per Minute bench, which is the maintainer of this site rather than a newsroom. There is no editorial board, no panel of reviewers and no claimed professional credential behind the measurement claims here — and saying otherwise to satisfy a checklist would be the first editorial failure worth reporting.

What stands behind the numbers instead is that they are reproducible. Every measurement figure on this site can be re-derived by anyone with a browser, using the same tools and the same stated method, which is a stronger guarantee than a name would be. Where that is not enough — the reported figures above — the limitation is stated on the page.