How it works

Every score on this site starts from a verbatim Reddit excerpt with a link to its source. Mentions are extracted from threads where people ask for or give recommendations, resolved to one canonical entity, weighted by an author-trust model that discounts copy-paste rings and self-promotion, and time-decayed per entity type. Category rankings use a Bayesian consensus so a handful of mentions can never outrank hundreds.

Analytics are first-party and cookie-free; where-to-buy links sit below the evidence and never influence a score. See the spam report for the campaigns detected so far.