What we keep

What is stored on our side

One row per completed test, containing four things: which test it was, the score, whether you were on a mouse, a trackpad, or a touchscreen, and a random token generated in your browser. Nothing else.

That row exists for one reason. The percentile this site is built around requires a population to compare you to, and the population is made of these rows. Without them the site has nothing to tell you.

What is deliberately not stored

• No account. There is nothing to sign up for.
• No IP address is recorded against a result.
• No device fingerprint. Nothing is probed to tell one browser apart from another.
• No advertising tracker, no third-party script of any kind.
• No cookies. The site sets none, and neither does the page-view counter.

Page views

The site counts page views using Counterscale, an open-source counter running on a server we operate at cs.benchself.com. Nothing is sent to a third party, and there is no advertising network involved.

Each page view records the page address, the referring site if there was one, and the broad browser and device class your request already announced in its headers, plus the country your network resolves to. That is the whole record. It is never joined to your test results, which live in a separate table with no request data attached to them.

There is one derived value: whether a request is the first one from your network in the current day. It is computed at the moment of the request and stored as a single yes or no. Your address is not kept alongside it, and the calculation resets daily.

The counter exists so we can tell which pages people actually use and which ones are not worth maintaining. If your browser or extension blocks it, nothing on this site stops working.

The random token

A random identifier is generated the first time you complete a test and kept in your browser's local storage. It is not derived from anything about you and it cannot be linked to any other site or service.

It does one job: it lets us count only your first attempt at each test toward the population curve. People who practise get faster, and counting every attempt would drag the distribution toward the practised and away from the truth. Clearing your site data removes the token, at which point a future attempt would be counted as a first one again.

What stays only in your browser

Your run history, your best and average scores, and your device preference are stored in local storage on your machine. They are never sent anywhere. The CSV export is generated from that local data directly, so exporting does not involve a network request.

Clearing your browser's site data for this domain deletes all of it permanently.

Reference distributions

Where a test does not yet have enough real results to rank against, the curve includes a fitted reference distribution based on published summary statistics. Any page that does this says so directly next to the percentile. We never present fitted data as measured data, and the baseline is dropped automatically once enough real results exist.

Contact

Questions about any of this can go to support@benchself.com, or see who runs this site.