The Band That Calls Its Shot

Every post that goes through the Authority Index comes out wearing a band: Exceptional, Elite, High, Medium, Low. The band is decided from the words on the page, before a single person has scrolled past the post. Which invites the fair and slightly hostile question you should ask of anything that grades writing. Does that judgement have any relationship to what the post then does in a real feed? We ran the check, and this is what came back.

What a band is, and what it is not

The band is a model judgement, published as a band. It reads the text and predicts quality. It has no sight of reactions, comments, follower counts, the hour you posted or anything else that happened afterwards. Inside the project the honest label for it is a predicted quality band, and we labour that point because the alternative reading, that the band reports a measured result, would be false.

So the check is a comparison of two independent readings of the same post. One is the band, made from the writing. The other is what the audience did.

The ladder holds, rung by rung

Every post was placed against its own creator's history rather than against the corpus. A post at that creator's p90 is one that out-earned nine tenths of everything else the same person published. Measuring within creator this way means a large account and a small one are each judged against themselves, so audience size cannot flatter either of them.

The average top-band post lands at its creator's 91.5th percentile. The average bottom-band post lands at the 14.2nd. The three bands in between fall in order and never cross: 69.2, 61.0, 38.6. Five bands, five rungs, no shuffling.

The top band is sharper than its average

An average can hide a messy distribution, so here is the harder reading. Take each post and ask a yes-or-no question: did it clear its own author's p90, the personal top tenth of everything that person has published? By band, the share that clears it runs 88.9%, then 6.4%, then 0.7%, then 0.2% and 0.2%.

Close to nine in ten top-band posts land in their own author's top decile. That is a much stronger claim than a mean, and the gap between the first two rungs, 88.9% against 6.4%, is the part worth staring at. The top band is not the good end of a gentle slope. It is a cliff edge, and 2,867 posts in the analysis cohort are standing on it.

What a writer should take from it

You can get the read before you publish. The whole value of a text-only judgement is its timing. Anything derived from reactions arrives when the post is already cold and the only available lesson is for next time.

A small account's top band means what a large account's top band means. Because every percentile here is drawn inside one person's own history, the ladder is not a restatement of who has the biggest audience.

And a band is a probability, never a promise. Roughly one post in five hundred from the bottom two bands still clears its author's own p90, and plenty of top-band posts land in the middle of the pack. The band tells you where the odds sit before you press publish. It does not tell you what any single post will do, and any tool that claims otherwise is selling you a coin toss with a confident voice.