UFCalendar
MMA Robbery Index
The most disputed decisions in MMA, ranked by the real record: fan scorecards, media scores and split verdicts stacked against the official result. No model, no narrative, just how hard the public record disagrees with the judges.
#1 · The worst verdict on record
- 75% of 87 fan cards scored it for Kevin Randleman
The worst verdicts
- 61Controversy#2
- 54% of 26 fan cards scored it for Ken Shamrock
- 54% of 26 fan cards scored it for Ken Shamrock
- 16Controversy#3
- 19% of 27 fan cards scored it for Mark Coleman
- 19% of 27 fan cards scored it for Mark Coleman
Most robbed fighters
Most robbed fighters
The public majority scored these decisions for them. The judges didn't.
- 1Kevin Randleman1 decision79
- 2Ken Shamrock1 decision61
Most gifted fighters
Handed the most wins the public scored against them.
- 1Bas Rutten1 decision79
- 2Dan Severn1 decision61
Robbery of the year
The most disputed decision of each year, by controversy score.
Who keeps scoring these fights?
How the index works
Each decision gets a controversy score out of 100: the share of fan scorecards that scored the fight for the official loser (55%), the share of media scores that picked the loser (35%), and whether the judges themselves split (10%). A fight needs at least 25 fan cards or 5 media scores to qualify, small samples don't rank.
Frequently asked questions
What is a robbery in MMA?
A robbery is a judged decision where the public record, fan scorecards and media scores, overwhelmingly favored the fighter who officially lost.
What is the biggest robbery in MMA history?
By controversy score, the biggest robbery on record is Bas Rutten def. Kevin Randleman at UFC 20: Battle For The Gold in 1999.
How is the controversy score calculated?
Each decision gets a controversy score out of 100: the share of fan scorecards that scored the fight for the official loser (55%), the share of media scores that picked the loser (35%), and whether the judges themselves split (10%). A fight needs at least 25 fan cards or 5 media scores to qualify, small samples don't rank.