Which phrases the board uses

A phrase is lowercased. Every page title, heading and paragraph is lowercased and joined into one string. If the phrase appears anywhere in that string, the page counts. The result is a number of pages.

That is the entire rule and it is deliberately crude. There is no weighting, no position bonus, and no attempt to distinguish a phrase used carefully from one dropped in. Crude is appropriate here, because a more sophisticated version would suggest the result meant more than it does.

Five phrases, each with the number of pages that use it. They are nexus market mirrors, nexus market links, a nexus market url, a nexus market onion address, and a nexus market login. The phrases are the ones people actually search for when looking for this kind of site: mirrors, links, a url, an onion address, and a login.

The gauge shows the proportion of all pages that name the market at all, which is high, because a board about one subject mentions it often. That is what you would expect and it is not a finding.

Because it looks like authority and is not. A high count says a site talks about something a lot. It says nothing about whether the site is correct, useful, current, or written by anybody who knows anything.

It is the metric that most rewards the worst behaviour in this space. A page can be constructed entirely to maximise a count like this, and such pages exist in large numbers, and they would score well here while being worthless.

So the board publishes its own figure and immediately says what it is worth. Showing a weak metric with the weakness attached is more honest than not showing it, and considerably more honest than showing it as a strength.

The board also tracks how many page titles carry a given phrase, which is a different and smaller count. A title is one line and there are only so many of them, so title coverage is genuinely scarce in a way that body coverage is not.

It is still a measure of usage rather than of quality. The only reason to track it separately is that it is a different question, and conflating the two is a mistake that sites in this space make constantly: they check that a phrase appears somewhere on the site and assume that covers everything.

• What anybody searched for. This site has no access to that and never will.
• Whether a phrase is used naturally or awkwardly. A count cannot see the difference.
• How often a phrase appears on a page. Only whether it appears at all.
• Whether the pages using a phrase are the pages you would want.

Every site in this space is optimised for phrases like these, and almost none of them say so. This one says so, publishes its own numbers, and explains why they are not evidence of anything.

That is the most useful thing a phrase coverage panel can do. It cannot make the metric better, and it can stop a reader mistaking a count for a qualification. The page on what this board covers makes the same argument from the other direction.

Does a high count mean this board is thorough?

No. It means a phrase appears on many pages. Depth is in the panels and no count can see it.

Why publish a metric you say is worthless?

Because it is computable and because a weak metric shown with its weakness attached is more honest than one hidden or one dressed up.