El ColegiadoData on refereeing in the Liga

Updated on 

Data on refereeing in the Liga

26 seasons, 9,880 matches and 80 referees, compared with what the bookmakers' odds predicted. Every figure can be checked: sources cited, method published, tables downloadable.

Undiano MallencoReferee Real Madrid CF2000–19

Matches

52

Wins

31

Win %

60%

31 W · 8 D · 13 L

No clear difference once team strength is taken into account.

Examples chosen by number of matches, not by deviation.

Undiano Mallenco × Real Madrid CF

What the data says

Four results, each with its details and method.

Referee × club deviations after correction

0notable out of 1,336 pairs

With so many comparisons, a few gaps always appear by chance (261 out of 6,590). Once chance is accounted for, none remain.

Summary of verdicts →

The odds, a reliable benchmark

−1.3% to +1.8%

Bookmakers' odds predict results very well: in each of the 26 seasons, the whole Liga's actual points stay within a few % of their forecast. That is our benchmark.

Calibration by season →

Effect of VAR (2018-19)

+26%

Since VAR, more penalties are given: 0.29 per match before, 0.36 after, across the whole Liga. It arrived the same season the payments ended.

The VAR study →

Negreira study: before / during / after

−0.111pt/match

Between the payment years (2001-18) and afterwards (2018-26), Barça's points (relative to the odds) change in the same way as Atlético's, Valencia's and Sevilla's: no clear difference.

Read the study →

Common questions

Short answers, computed from the data; the detail is one click away.

Where to start

Every club, referee and season has its own page.

How it works

Three steps, all documented.

Methodology
  1. 1

    Open sources

    Scores, odds, referees, cards and penalties from 9,880 matches, gathered from public sources, normalised and cross-checked.

  2. 2

    Expected points from the odds

    Before each match, the market gives the chances of a win, draw and defeat. Points obtained are compared with these expected points.

  3. 3

    Corrected tests

    Every deviation is tested, then corrected for multiple comparisons: a deviation is only called notable if it survives the correction.

Why trust us

  • Sources and licencesfootball-data.co.uk, Zenodo, Wikipedia: each file carries its licence.Data
  • NeutralityNo ranking by deviation or by "suspicion", morally neutral colours, sample size always shown.About
  • Public correctionsFound an error? Report it: every correction is checked and recorded.Corrections
  • Liga onlyPrimera División, 2000-01 → 2025-26. Copa del Rey and Champions League: coming later.Methodology
  • Non-commercialNo advertising, no paid offer, no commercial ties.Legal notice