Great players.
Different verdicts.
Football’s biggest prize, through a different lens. Explore the evidence, rewrite the rankings, and put your own jury to the test.
No names. No reputation. Just the evidence.
Harry Kanehunter78.0- Jev certainty
- 69%
- Minutes seen
- 4,074
- Evidence
- 2026 nominees
- Editions replayed
- 12
- Winners matched
- 7/12
- Scoring criteria
- 3
- Nominees judged
- 338
2026, reranked.
Choose what matters. Watch the field respond.
Your #1Harry Kane78.0Harry Kane leads 2026 with 78.0. 30 matching players.
The track record.
Replay cards use fixed 60 / 30 / 10 weights and open with the balanced jury. The two evidence tracks are reported separately. ρ is Spearman’s rank correlation against the official finishing order: 1.00 would reproduce the jury exactly, 0.00 is no better than shuffling the nominees. Retrospective agreement only—not predictive accuracy.
What does this comparison actually test?
We replay the officially ranked nominees for every edition we can reach—12 of them, 338 judged players. Jev receives anonymous evidence: no names, no clubs, no award year, no voting results. It answers four typed questions per player, and the weighted score is 60% individual impact, 30% team achievement and 10% fair play.
What Jev is shown. Per-90 metrics and absolute season totals, because the award rewards volume as well as rate. Percentiles against the whole nominee field—not against the player’s own position, which rated a full-back among full-backs and kept lifting defenders over forwards. A strength tier for every competition, so thirty goals in a smaller league are not read as thirty in a top-five one. Titles weighed by prestige rather than counted, with minutes and start share attached, so a squad player on a treble-winning side is not the same case as that side’s mainstay. Yellow and red cards against the positional norm. And the player’s international tournament: goals, how far their country went, and whether it won.
Two evidence tracks, not comparable. 2022–2026 uses PitchAPI match data—domestic leagues, all three UEFA competitions, around thirty per-90 metrics. 2013–2021 predates that provider entirely and is built from Kaggle snapshots: Transfermarkt for appearances, cards and cup runs, FBref for advanced metrics from 2017-18, Understat for league xG from 2014-15. Those editions are judged on the calendar year, January to December, because that is the window the award itself used before moving to an August–July season in 2022. Older editions rest on thinner evidence, which each card reports.
What is still missing. Domestic cups in the 2022–2026 track, the Nations League and the Club World Cup. Nothing measures how prominent a player was in the press, and the jury is made of journalists. Unavailable metrics stay missing rather than becoming zero, and advanced metrics are divided by the minutes of the matches they were actually observed in.
This is a retrospective comparison, not a Ballon d’Or simulation and not a claim of predictive accuracy. Jev agrees with the voters roughly as often as a careful reading of the numbers would, and it still puts players in its top five that the jury left outside its top ten. The 2026 snapshot has no result to be measured against at all.
Every source, in full.
Official finishing orders are ground truth only. They are never shown to Jev, in any track.