How Often Do UFC Betting Favorites Win? What the Data Says

Quick answer

UFC betting favorites win about two-thirds of resolved fights; that headline is useful, but the price matters more than the label. A fighter priced near a 60% fair chance should win around six times in ten; an 80% favorite should win around eight. The largest published test cited here covered 6,190 UFC contests and found normalized odds tracked results, while average return stayed negative. Favorites are usually right; that doesn't mean their odds offer value.

Data snapshot: 6,190 UFC contests (2007-2022), 558 fights across 50 events (2019-2021), and 709 eligible bouts (2012-2013); all analyze pre-fight moneylines.

How often do UFC favorites win?

Two direct public audits land near the same answer, roughly 66% to 68%; a larger published study supports the market's broad calibration.

Pinnacle reviewed closing odds for 558 bouts across 50 events; favorites won 370, or 66.31%. A separate MMAOddsBreaker audit started with 737 bouts from 2012 and 2013, but two were unaired. After removing 13 no contests, three draws and 10 pick'ems, favorites won 483 of 709 eligible fights, or 68.12%.

That audit also summed each favorite's raw closing probability without removing vig, which projected 483.18 wins against 483 actual wins. That close total is descriptive, not a clean calibration test, because the inputs still include bookmaker margin.

The larger academic evidence points the same way: Thomas Robbins studied 6,190 UFC contests involving 2,021 fighters. His data ran from June 2007 through December 2022 and used odds collected from BestFightOdds. The paper found no systematic bias between normalized predicted probabilities and actual win rates across the distribution.

So the UFC favorite win percentage supports a broad rule: favorites win about two fights in three. But that average mixes slight favorites with huge ones.

What do implied-probability buckets show?

Pinnacle's smaller audit provides the clearest public breakdown by closing implied probability.

The pattern matters: UFC heavy favorites win more often than moderate favorites. Slight favorites lose often enough that calling them “the favorite” adds little without the actual line.

These buckets also sit close to their quoted ranges: the 70% group won 72.87%, while the 60% group won 64.53%. A perfect match shouldn't be expected across 558 fights, since random variation can shift a bucket several points when its count isn't published.

Robbins tested calibration on a much larger UFC sample. The 95% confidence interval for the slope relating predicted and actual win rates was 0.983 to 1.08. Because that interval contains 1, the paper found no systematic bias: grouped no-vig probabilities broadly rose with observed win rates.

Calibration doesn't mean every line is correct; it means similar prices behave sensibly when grouped across thousands of fights.

How do you turn UFC odds into implied probability?

American moneyline odds hide a simple calculation.

• For negative odds, divide the absolute price by that price plus 100.
• For positive odds, divide 100 by the price plus 100.

At -200, the raw implied probability is 200 divided by 300, or 66.67%; at +175, it is 100 divided by 275, or 36.36%.

Those numbers add to 103.03%, not 100%; the extra 3.03 percentage points are the overround, often called the vig or juice. Robbins found an average UFC booksum of 4.60% and an estimated bookmaker margin of 4.40% in his sample.

For a fair two-way estimate, divide each raw probability by the total; -200 becomes 64.71%, and +175 becomes 35.29%.

This step matters when you judge UFC odds accuracy, because raw probabilities make both sides look more likely than they can be. Use normalized, no-vig probabilities for calibration; use the offered price when calculating your break-even rate.

Why doesn't a high win rate mean a profitable bet?

A result can be likely and still be overpriced, so suppose favorites win 66% of fights. A flat bet at -200 needs a 66.67% win rate just to break even before tax or other costs. The same 66% record loses money at -250, where the break-even point is 71.43%.

The published data shows this gap: Robbins reported a -5.82% average return across UFC bets, versus a -4.40% average bookmaker hold. No odds bucket delivered a positive return that cleared the 5% statistical threshold.

Kaiana Miller's 2026 University of Nevada, Reno dissertation offers another check, analyzing 2,122 eligible numbered-card bouts from UFC 100 through UFC 293. Betting every favorite returned -4.68% per unit; betting every underdog returned -9.99%. Both flat strategies lost over the full July 2009 to September 2023 sample.

The later peer-reviewed paper by Miller and Mark Nichols describes MMA betting as largely efficient. It found no general favorite-longshot bias and few statistically significant positive results in out-of-sample tests. That is the practical warning: a higher hit rate isn't a shortcut to positive expected value, and this is measurement, not a betting recommendation.

What do Topuria, Gaethje, Holloway and McGregor teach us?

Ilia Topuria opened at -550 against Justin Gaethje for UFC Freedom 250, while Gaethje opened at +400. The raw line gave Topuria an 84.62% implied chance, but both sides totaled 104.62%.

After simple normalization, the pair was closer to 80.88% and 19.12%. Gaethje still won when Topuria's corner stopped the fight before the fifth round. That result wasn't proof that the line was foolish; a calibrated 81% favorite should lose about 19 times in 100 comparable fights.

The next month gave the opposite example: Max Holloway closed around -300 against Conor McGregor at UFC 329, while McGregor was about +250. Holloway won by first-round TKO after McGregor injured his leg, but that result didn't prove -300 was a good price.

These fights show why memorable upsets can distort your judgment, because an 80% favorite isn't a lock. An underdog win doesn't erase the closing line's information; you need many similarly priced fights before judging calibration.

Which odds should you record in your own UFC audit?

Use a fixed method before looking at results, since small choices can change a claimed UFC favorite win rate.

• Pick one odds source or a defined market average.
• Record the timestamp. Opening and closing odds answer different questions.
• Keep both sides of the moneyline so you can remove the vig.
• Treat pick'ems separately because neither fighter is a clear favorite.
• Exclude draws and no contests, or state exactly how refunds are handled.
• Group fights by no-vig probability, not the minus sign alone.
• Publish the number of fights in every bucket.

Closing odds make the cleanest market benchmark because they absorb late information; opening odds are better for studying line movement. Mixing them creates a noisy result, so also keep the unit of analysis straight.

A 500-fight dataset has 500 winners, but it has 1,000 fighter sides. Some papers report contests while others report betting opportunities, and confusing them can double a sample on paper.

How solid is this data?

The broad finding is solid: UFC favorites win near two-thirds of resolved bouts, and stronger favorites win more often, but the exact percentage isn't timeless.

Robbins supplies the largest published sample here, but it combines odds from multiple sportsbooks through BestFightOdds, and the data ends in 2022. Sportsbook competition, limits and bettor behavior can change.

Pinnacle's bucket table covers only 558 fights and 50 events, including the unusual 2020 schedule, so era effects may matter. The article doesn't publish each probability bucket's count, which blocks a proper uncertainty estimate for every row.

The 2012-2013 audit is transparent about exclusions and lands at 68.12%, but it is old and uses one closing-odds source. Miller's larger sample covers only numbered UFC events, not the full Fight Night schedule, and excludes pick'ems or bouts missing required fighter data.

There is selection bias too, because a closing line exists only after matchmaking has paired the athletes in a booked fight. These studies test the betting market's prices; they don't tell you which fighter traits caused each win.

The safest reading is narrow: pre-fight UFC moneylines have been broadly well calibrated in large historical samples. They still contain a margin, and no public study here supports blindly betting every favorite.

FAQ

What percentage of UFC favorites win?

UFC favorites win about 66% to 68% of resolved fights in the public audits reviewed here. Pinnacle found 66.31% across 558 bouts, while a 2012-2013 audit found 68.12% across 709 eligible bouts.

How often do UFC underdogs win?

UFC underdogs win roughly one fight in three; the 2012-2013 sample recorded 226 wins in 709 eligible fights, or 31.88%. The rate changes with the price, so slight underdogs win far more often than longshots.

What does -200 mean in UFC odds?

A -200 moneyline means you risk $200 to earn $100 profit, and its raw implied probability is 66.67%. That isn't a fair probability until you also convert the opponent's odds and remove the bookmaker's overround.

Are UFC closing odds accurate?

Across 6,190 UFC contests, normalized predicted probabilities tracked actual win rates without a detected systematic probability bias. That means the market was broadly calibrated, not that every closing line was correct.

Is betting every UFC favorite profitable?

Historical evidence says no: in a 2,122-bout numbered-event sample, flat betting every favorite returned -4.68% per unit. A high win percentage cannot overcome a price that demands an even higher break-even rate.