flat-pool-v1

The original algorithm:

score = round(100 × passed / applicable)

applicable is the flat pool of every check firing (site-wide checks plus per-page checks across every page audited) that didn’t return na. Simple and intuitive at small scales, but site-wide signals get diluted as page count grows: a 500-page audit pushes the site-wide checks down to ~0.1% of the denominator, while a 1-page audit puts them at ~40%. Different --max-pages settings produce different effective weightings for the same site’s site-wide signals.

See all scoring methodologies or browse scorecards to see which methodology each version pins.