Rankings · one cross-league scale

The Big Board

Every qualifying player in ten leagues on one number. The Statz Score is the only figure on this site that crosses pools: around 54 is the middle of the whole field, the mid-90s are genuinely world-class, and a 90 means the same for a centre-back as for a forward. Six players tie at 99, so the order at the very top is a declared tiebreak, not a verdict.

3,027 players 10 leagues ≥900 minutes 2025/26, back to 2023/24 (Big-5 only) Understat × FotMob × Sofascore

The board

Ranked by the Statz Score. Narrow it, then read down.

Pool

Season

Position

Minimum minutes

900

Max age

Any

Max market value

Any

Contract

Leagues

Ranked by Statz Score

Where the quality sits

The board answers who. This answers where — by club, by league and by position, counted over the whole season rather than your filters. Pick a row to take it up to the board.

Group by

Reading the number

One number crosses ten leagues. Nothing else on this page does.

Only the Statz Score crosses pools. Big-5 and the challenger leagues are ranked separately everywhere else on this site — percentiles, archetype fits, role scores are all computed within a pool and never mixed. The score on this board lifts every player onto a shared axis by his league's anchored coefficient plus a club-strength lift (applied above the scale's floor of 20), so a PSV forward and a Manchester City forward are directly comparable here. That is also why no percentile appears anywhere on this page: a percentile cannot make that trip, and a board that showed one beside a cross-pool rank would be quoting two different populations in one row.

It is team-adjusted, and that is a choice you can see. The published score is round(softcap(20 + (rating − 20) × team_adj)): the part of the rating above the scale's replacement floor is scaled by team_adj — the league coefficient plus a club-Elo lift spreading 0.16 across a league, phased in with minutes played so a thin sample is never amplified by its badge. A squad player at a big club therefore outranks an equally-rated starter at a small one by construction. At the very top a smooth compression (identity below 96, ceiling 99) keeps 99 scarce and the order real. Sort by club and the effect is visible in a column rather than hidden in a coefficient.

The league anchoring is doing most of the work at the top. Median score by league runs from the Premier League at 61 down to the Scottish Premiership at 40; the challenger pool peaks at 82 and the Championship at 64, against 99 in the Big-5 — a deliberately steep exchange rate, calibrated on the players rated in both pools in the same season, so the best challenger seasons read as good Big-5 starters rather than Big-5 elites. So the first screens of a ten-league board are almost entirely Big-5. That is the coefficient, not a discovery, and not a sign the challenger data is broken — filter to one challenger league to read it on its own terms.

How it's computed — and what it is not

The population. Every player with at least 900 league minutes in a season the site carries. Big-5 goes back to 2020/21 — the full extent of the complete-board era — and the challenger pool (Eredivisie, Liga Portugal, Belgian Pro League, Scottish Premiership, Championship) spans the same window, so every board is ten leagues. A winter mover between pools appears once, on his Big-5 row.

Sided centre-backs are measured, not asserted. An LCB/RCB label comes from the mean pitch-width of every defensive action the Opta event corpus holds for the player this season (hover it for the number) — only a decisive lean over 80+ actions gets the letter, a central profile or a season without event coverage stays plain CB, and full-backs carry FotMob's own LB/RB/LWB/RWB labels. Never a guess from a squad sheet.

The rank. Rank is the player's position in the whole season's ranking and never renumbers — filter to Championship centre-backs and the numbers stay 1, 4, 17, 240, because how far down the world someone sits is the question this page exists to answer. Sorting by another column leaves those ranks in place on purpose. The Statz Score caps at 99 and the cap ties every season, so ties break on the in-pool rating, then minutes, then name, then id — the same chain the Annual uses to pick its Player of the Year.

Coverage gaps are marked, not smoothed. A dagger marks a player the coverage guard fired on. Those ratings are hard-capped and regressed harder toward the positional mean, and the build refuses to crown them an archetype. Ranking them silently beside a fully measured rating would make a claim the model itself declines to make. The guard's remit is wider than what it actually catches: it is written to fire on centre backs, full backs and both kinds of central midfielder who are missing their defending category, as well as on keepers with no save-board data. Every figure that follows is counted off the board you are looking at, because the answer changes with the season.

Two currencies in one output column. Goal contributions per 90 read npxG + xA where Understat covers the player and goals + xA where it does not — the whole challenger pool, plus a handful of Big-5 players who never merged. Each cell names its own currency, because sorting a mixed board on that column compares two different things. Keepers ride at zero: their value is goals prevented, which the defensive model prices separately.

Market value stands in for a fee — a scisports estimate, not a fee, and never a valuation. Wages, agent costs and add-ons are not in this corpus at all. Value and contract data cover the current season only, because they are as-of-now facts; on 2023/24 and 2024/25 those two columns have nothing to read and are switched off rather than blanked. A contract date says when a deal ends and nothing more: option years and unannounced extensions are invisible to the source, and a player with no date on file cannot be shown to expire.

Notes & caveats

This is a ranking of measured output in a season, not a scouting verdict. It knows nothing about injuries, contracts you cannot see, temperament, or whether a player would suit your team — the Fit Board and the Forge exist for that.

Minutes are league minutes only. A player who spent the season winning a cup is ranked on his league work.

Ranking is not the same as valuing. The cheapest player on a filtered board is not thereby the best buy; the Window prices a squad change in goals, and the Forge ranks on metrics you choose rather than on the site's own weighting.

The Big Board

01 The board

Ranked by Statz Score

The value frontier

02 Where the quality sits

03 Reading the number

04 Notes & caveats

The board

Where the quality sits

Reading the number

Notes & caveats