How a number gets onto this site

The test a number has to pass

Three conditions. A number has to be stated plainly, so there is something concrete to disagree with. It has to be sourced or derived in the open, so anybody can check the working without asking permission. And it has to carry a label saying whether it was counted or judged, so a reader knows whether they are holding arithmetic or somebody choosing an assumption.

Nothing else qualifies a figure. Not how interesting it is, not how often it gets repeated, not whether it would make a better page. A number that cannot be derived in front of the reader has no place here even when it is probably right.

The conditions are also why the pages look alike. A number page states the value, shows where it came from, says what it is good for, says what people conclude from it that does not follow, and says what would change it. That shape is what the rules produce rather than a style anybody chose.

What gets rejected

Three kinds of figure never appear on this site, and between them they cover most of what gets published elsewhere in this subject.

• Anything the site would have to measure. Availability, response times, order volumes, anything requiring a connection to something and a watch on it afterwards. That is a decision rather than a gap, and the zero page gives both the reasons and the price.
• Anything a market says about itself that cannot be checked from outside. User counts, vendor counts, dispute rates, resolution times. Any of it might be true. There is no way to tell from here, and repeating a claim would launder it into a fact.
• Anything that moves faster than a static page can follow. A fee adjusted weekly, a limit changed without notice, a balance. Where a quantity of that kind matters, the page names it and says the current value has to come from the market itself.

A fourth category sits at the edge and is handled differently: numbers that are real but weak. The count of published addresses is knowable and has its own page, and that page spends most of its length on what the count cannot tell you. Publishing a weak number with its limits attached is more honest than leaving it out and letting somebody else publish it bare.

Why the rules are written down

Because the alternative is a site that sounds confident. This subject rewards confidence and punishes accuracy, and the fastest way to look authoritative is to print figures nobody can check. Writing the rules down makes that harder to do by accident.

It also gives readers something to hold the site to. A page here that states a number without saying where it came from is breaking a rule published on this page, and you can discount the page accordingly and tell somebody why.

The rules are not a claim to neutrality. Choosing which quantities deserve a page is a judgement. So is the order of the sections, and so is the decision to spend a whole section on ownership rather than on prices. What the rules do is keep those choices visible instead of dressed up as facts, and the keys section is the clearest example of one.

Questions people ask

Why is there no page about prices?

Prices move faster than a static page can follow and vary by vendor, by product and by day. Such a page would be out of date before it was read, and the units page covers the general rule.

Can a number change its chip later?

Yes. If a value that was judged turns out to fall out of a definition, or the reverse, the chip changes and the opening lines of the page say so.

Who decides which numbers are worth publishing?

Whoever writes the page, which is a judgement and is stated as one. The rules above constrain how a number is presented rather than which subject gets picked.