How the statistical ranking works

Overall = 0.30 × Peak performance + 0.25 × Career performance + 0.20 × Individual accolades + 0.15 × Team accomplishments + 0.10 × Longevity

• Each category reports full, limited or unavailable evidence. An unavailable category contributes nothing and is shown as unavailable.
• A calculated rank requires evidence in all 5 categories. Players with at least 3 are shown as rank pending with a clearly labelled provisional estimate; they are never interleaved with ranked players.
• Rates need minimum samples: 100 league matches or 20 caps.
• Evidence coverage (how many categories have full evidence and what share of inputs are directly sourced) is reported separately from the score. It is a completeness indicator, not a calibrated probability.
• The list is not padded to 100: only players meeting the evidence rules are ranked.

Position adjustment on contemporary award points: goalkeeper ×1.15, defender ×1.15, midfielder ×1, forward ×1 — an editorial assumption that major awards have historically favoured attacking players.

Normalisation uses fixed, published anchor points rather than the current player pool, so adding or removing players never silently re-scores everyone else.

• Ties are broken by overall score, then peak, accolades, team, career and longevity (each descending), then name and slug alphabetically.
• Every recalculation runs twice and compares an input hash; a non-deterministic result blocks publication.
• Published snapshots are immutable — enforced by the database itself. Data or methodology changes produce a new version with recorded rank changes.
• The personal ranking is a separate record and is never read or modified by recalculation.

• Official appearance and goal: a senior competitive match for a first-team club or a full international (cap). Youth, reserve, Olympic, friendly and unofficial matches are excluded from totals.
• Club figures from Wikidata follow the Wikipedia infobox convention — domestic league matches only — and are labelled as such. Historic totals differ between sources depending on friendlies and unofficial matches; when sources disagree on the same spell, the conservative claim is used and the record is flagged.
• Team titles are either stated directly, derived from tournament participation plus the edition winner, or (provisionally) inferred from a club stint covering a title-winning season.

• Peak is measured through recognition, not on-pitch performance data, which double-counts some honors with the accolade and team categories.
• Defenders and goalkeepers are assessed on selection rate and honors only; no cross-era defensive metric exists in the current sources.
• Era factors for scoring environment and international fixture density are neutral (uncalibrated), which may understate players from lower-scoring eras and favour players from eras with more internationals.
• Source coverage is uneven: some players have undated or unrecorded awards on Wikidata, and competitions outside the catalog (domestic cups, state championships) are not counted.
• Team titles inferred from stint dates may credit a player for a season in which they barely played.
• Team success is an imperfect proxy for individual quality; its weight is deliberately modest.