Botworthy Score v3

v3: reweighted on ninety studies

20 October 2026

Reweighted after reviewing about ninety studies published up to October 2026. Being retrieved matters more than how a page is formatted: in the largest 2026 datasets, retrieval rank explains roughly ten times more of who gets cited than page features do, so the score now says plainly that it measures how quotable a page is once an engine finds it. Within the page, the feature that holds up best is a sentence that directly answers the question in the words a searcher would use, so Answer-first sections (20 to 24) and Question-led headings (6 to 10) carry more weight and will be scored on match to the question from v3.1. Self-contained passages (12 to 14) and Recently updated (8 to 10) also rise, the latter now meaning a genuine update within 12 months rather than any change within 3. Lists and tables (14 to 10), Short sentences (7 to 4), Named sources (10 to 8) and Named expertise (6 to 4) fall, because rank-controlled studies find their effects small, null or unmeasured. Short sentences now applies only to section opening sentences. The Princeton GEO lifts that anchored v2 have not replicated on current models and are no longer cited as effect sizes. Ten criteria, still 100 points.

Previous: Before v3: where the score started. Next: v3.0.1: a page integrity gate and six more opening habits.