How Often Do UFC Judges Disagree? What Scorecards Show

Quick answer

UFC judges give the same exact score in about three of every four rounds. A 2011-2020 analysis found complete agreement in 4,029 of 5,238 scored UFC rounds, or 76.9%. That leaves 23.1% where at least one judge differed. Yet disagreement rarely changes the winner. A separate study tested 4,129 MMA decisions. Standard and round-by-round consensus scoring produced the same result 97.53% of the time. Judges disagree often enough to matter, but not every difference is a bad call.

Data snapshot: 5,238 UFC rounds (2011-2020), 4,129 MMA decisions (2003-2023), and 409 rounds with at least one 10-8 (2001-2019).

How often do UFC judges disagree?

The best direct UFC sample puts exact-score disagreement at 23.1%. Nate Latshaw merged official UFC Stats with MMA Decisions scorecards for fights from 2011 through 2020. After cleaning, the dataset held 5,238 rounds from 1,670 decisions.

All three judges entered the same score in 4,029 rounds. That is 76.9% agreement. The remaining 1,209 rounds had at least one different score.

That number needs a label. It measures the exact score, such as 10-9 versus 10-8, not just the round winner. Two judges can both pick the same fighter while disagreeing about the margin.

These rows ask different questions with different datasets. Round scores can vary while the majority still picks the same round winner. Several differences can also disappear in the final cards.

What does “judges disagree” actually mean?

There are three useful levels of UFC judging agreement. Mixing them produces bad claims.

Same exact score

All judges write 10-9 for Fighter A, for example. A 10-9, 10-9 and 10-8 set counts as disagreement despite the shared winner.

Same round winner

All judges choose the same fighter, but they may differ between 10-9 and 10-8. Winner agreement does not prove margin agreement.

Same fight winner

The majority of completed scorecards picks the same winner. Judges can submit different totals and dispute rounds while all backing one fighter.

So a “unanimous decision” does not mean a unanimous reading of the fight. It means all three final cards name the same winner.

How can a unanimous UFC decision hide disagreement?

Islam Makhachev's first fight with Alexander Volkanovski at UFC 284 is a clean example. Makhachev won by unanimous decision with cards of 48-47, 48-47 and 49-46.

All three judges gave Makhachev rounds one, two and four. Ben Cartlidge and David Lethaby gave round three to Volkanovski. Derek Cleary gave that round to Makhachev. Two judges also gave Volkanovski round five.

The final verdict was unanimous, but round three was not. Cleary's extra round explains his 49-46 total. The other cards finished 48-47.

This distinction is useful when fans compare a model with the official result. A model can match the winner yet disagree on two rounds. Calling that full agreement would hide most of the useful information.

Why do 10-8 scores create another kind of disagreement?

A 10-8 asks two questions. Who won the round? Was the gap wide enough to double the usual one-point advantage?

Paul Gift studied 409 rounds where at least one judge awarded a 10-8. Only one included disagreement about the round winner. The harder call was the margin.

Gift's earlier audit also found that judges unanimously agreed on a 10-8 in only 8.7% of relevant rounds from 2001-2012. That figure rose to 26.1% by 2017 after the criteria became less restrictive. These percentages concern rounds involving 10-8 scoring, not all MMA rounds.

Conor McGregor against Nate Diaz at UFC 202 shows why one margin call matters. Derek Cleary and Jeff Mullen scored the bout 48-47 for McGregor. Glenn Trowbridge scored it 47-47 because he gave Diaz a 10-8 third round. All three judges gave Diaz that round. Only one made it wide.

McGregor still won by majority decision. The disputed margin changed one final card from a McGregor win to a draw, but it did not change the official winner.

What are UFC judges supposed to value?

The Association of Boxing Commissions clarified MMA scoring in July 2025. At least three judges score every round and favor the more effective fighter.

Damage sits first. Legal techniques must reduce an opponent's ability or will to compete. Knockdowns, heavy strikes, damaging submissions and visible injury can matter.

Successful striking and grappling are equal in principle. Each is judged by its result. A takedown with no useful offense does not automatically beat effective strikes.

Aggressiveness and fighting-area control are fallback criteria. Judges reach them only when they cannot find even a small edge in effective striking or grappling. Defense receives no score by itself.

The clarification says a 10-8 must include major damage. Dominance and duration may support it. Position alone is not enough.

Clear written priorities reduce ambiguity. They cannot give three people the same viewing angle or force identical judgments about impact.

Why can reasonable judges see the same round differently?

Ringside position matters. The cage, referee and fighters can block part of the action. A clean strike may look glancing from another seat.

Impact is also hard to count. UFC Stats records actions after review, but judging turns on their effect. Ten light strikes do not automatically outweigh one that badly hurts a fighter.

Sequence matters too. Round totals remove timing. A fighter may recover after early trouble, or land the hardest blow late. Two judges can value the same five minutes differently without either ignoring the rules.

Individual preferences add noise. A 2024 peer-reviewed study found that MMA judges place different weights on fighting actions. Those preferences can decide a bout.

That does not excuse any score. It explains why raw disagreement is not a complete measure of judging quality.

Would consensus scoring fix UFC judging?

Standard MMA scoring works in two steps. Each judge adds their own round scores. The majority of those three completed cards then sets the result.

Consensus scoring reverses the order. It takes the score supported by at least two judges in each round, then adds those scores.

Vincent Berthet tested both methods on 4,129 MMA decisions from September 2003 through October 2023. They returned the same result in 4,027 bouts, or 97.53%. They differed in 102 bouts, or 2.47%.

In those 102 cases, the consensus result matched MMA Decisions fan scoring 48.96% of the time. Standard scoring matched it 43.75% of the time.

That gap is not proof that consensus is better. Fans are not trained judges. People who dislike a result may submit cards more often.

Consensus scoring can cancel one outlying judge. It cannot fix a mistake shared by two. It changed only one of MMA Decisions' ten most disputed 2023 results.

What does judge disagreement mean for prediction models?

A pre-fight model forecasts the bout. A scoring model estimates how judges interpret completed rounds. They solve different problems.

For a bout likely to reach the cards, disagreement adds variance. A narrow 55%-45% prediction is not a forecast of every scorecard.

Post-fight models have another limit. Latshaw's JudgeAI used 31 differences in round statistics and matched the majority round winner in 83.3% of evaluated rounds from 2012-2020. It also matched the exact majority score in 80.2%.

Aggregate stats cannot fully capture damage, viewing angle or exchange order. A model trained on consensus learns historical judging behavior, not an objective hidden score.

How solid is this data?

The 76.9% figure comes from a transparent merge of UFC Stats and MMA Decisions. It covers 1,670 of the 2,125 UFC fights decided by judges from 2011-2020. That is 78.6% of eligible decision fights, not the whole period.

Cleaning removed missing cards, unusual scores and likely deductions. That improves comparison but excludes messy cases where disagreement may matter most.

Berthet's wider study includes UFC, Bellator and other promotions. Its cleaned set has 13,024 rounds and 4,129 bouts, with 10,350 UFC rounds in the raw 19,883-round collection. The 97.53% result is therefore MMA-wide, not UFC-only.

Gift's 10-8 evidence covers UFC, WEC and Strikeforce scorecards from 2001-2019. Its 10,926 figure counts judge-round observations, not separate rounds.

Rules and judging practice changed across eras and jurisdictions. The ABC issues a common framework, but local commissions choose what they adopt. None of these samples proves that today's judge agreement rate is exactly 76.9%.

The strongest conclusion is narrower. Exact round scores differ often enough to inspect, while most differences do not reverse the final winner.

How should you read a UFC scorecard?

Start with the round winners, then inspect the margins. This avoids treating a 10-8 dispute like a disagreement over who won.

Use four checks:

• Did all three judges pick the same fight winner?
• Which rounds had different winners?
• Did any judge use 10-8, 10-10 or a point deduction?
• Did one unusual round change a final card or the official result?

Makhachev-Volkanovski answers the second question. McGregor-Diaz answers the third. Both show why the final label alone is incomplete.

FAQ

How often do UFC judges disagree on a round?

In a cleaned sample of 5,238 UFC rounds from 2011-2020, all three judges gave the same exact score in 76.9%. At least one judge differed in 23.1%. Some differences concerned only 10-9 versus 10-8, not the round winner.

Can a unanimous UFC decision have disputed rounds?

Yes. A unanimous decision means all three completed cards picked the same fight winner. Judges can disagree on individual rounds and submit different totals, as they did in Makhachev versus Volkanovski at UFC 284.

How often does judging disagreement change the MMA winner?

A study of 4,129 MMA decisions found that standard scoring and round-by-round consensus scoring produced different results in 102 bouts, or 2.47%. This measures scoring-system sensitivity, not the number of wrong decisions.

Why do UFC judges disagree on 10-8 rounds?

Judges may agree on who won yet differ on how much damage and dominance justify a 10-8. In 409 studied rounds with at least one 10-8, only one had disagreement about the round winner.

Are UFC judges or UFC Stats more accurate?

They measure different things. Judges assess effective fighting and damage from live viewing angles. UFC Stats counts recorded actions. Stats help audit patterns, but totals alone cannot decide every close round under the scoring rules.