UFC red corner bias: is it the judges? What 7,187 fights say
The red corner wins 57.8% of UFC fights against a no-vig closing average of 55.9%, the second angle in the UFC betting angles study. The easy explanation is the judges: the red corner goes to the champion, the ranked fighter or the one being promoted, and a close round could lean that way. So we split all 7,187 fights by how they ended. The edge is the same size when the fight ends by knockout or submission, biggest in unanimous decisions, and gone in the split and majority decisions the judges actually decide.
The short answer: the red corner against the no-vig closing line, by how the fight ended
- All fights, 7,187+1.9 ptsRed wins 57.8% against a 55.9% no-vig line, z 3.5: the angle in the study
- Decisions, 3,520+2.5 pts57.8% against 55.3%, z 3.2
- Finishes, 3,650+1.4 pts57.9% against 56.6%, z 1.8: the same edge where no judge is involved
- Unanimous decisions, 2,758+3.4 pts59.3% against 55.9%, z 3.9: the clear wins, not the close ones
- Split and majority decisions, 762−0.7 pts52.2% against 53.0%, z −0.4: where the judges decide, red gets nothing
- Women’s fights, 860+0.9 pts63% go the distance; 57.0% against 56.1%, z 0.6
The red corner by how the fight ended
A judging effect can only live in decisions, and mostly in the close ones. A knockout or a submission is scored by nobody. So the test is to split the 7,187 fights by their method and ask the same question in each piece: did the red corner win more often than its no-vig closing chance said?
| How the fight ended | Fights | Red corner won | No-vig | Points over the line | z | Favorites, won against the line |
|---|---|---|---|---|---|---|
| All fights | 7,187 | 57.8% | 55.9% | +1.9 | +3.5 | 66.9% against 66.9% |
| Decisions | 3,520 | 57.8% | 55.3% | +2.5 | +3.2 | 65.7% against 65.9% |
| Unanimous decisions | 2,758 | 59.3% | 55.9% | +3.4 | +3.9 | 69.0% against 66.5% |
| Split and majority decisions | 762 | 52.2% | 53.0% | −0.7 | −0.4 | 53.4% against 64.0% |
| Knockouts and submissions | 3,650 | 57.9% | 56.6% | +1.4 | +1.8 | 68.2% against 67.8% |
Filtering on how a fight ended changes what the line should expect from everyone, not only the red corner. Favorites finish fights; an underdog who wins more often needs the judges. So among knockouts and submissions favorites beat their line a little, among unanimous decisions by 2.5 points, and among split and majority decisions they win 53.4% against a 64.0% line, because a split decision is by definition a close fight. Read each red corner gap against the favorites’ gap in the same row: in decisions red is 2.5 points over a line the favorites match; in split decisions red is at the line while the favorites are 10.6 points under it.
| Rule | All fights | Decisions | Unanimous | Split and majority | Finishes |
|---|---|---|---|---|---|
| Red corner, any price | +1.9 (3.5) | +2.5 (3.2) | +3.4 (3.9) | −0.7 (−0.4) | +1.4 (1.8) |
| Blue corner underdog | −1.4 (−2.0) | −1.7 (−1.7) | −4.7 (−4.2) | +10.0 (4.4) | −1.3 (−1.3) |
| Red corner underdog | +2.4 (2.6) | +3.6 (2.7) | +1.1 (0.8) | +11.5 (4.2) | +1.0 (0.8) |
| Red favorite, 3+ years younger than the blue dog | +5.1 (4.3) | +3.6 (2.1) | +6.5 (3.3) | −9.1 (−2.1) | +6.6 (4.2) |
| Every favorite (control) | −0.0 (−0.1) | −0.3 (−0.4) | +2.5 (2.9) | −10.6 (−6.1) | +0.4 (0.5) |
| Every underdog (control) | +0.0 (0.1) | +0.3 (0.4) | −2.5 (−2.9) | +10.6 (6.1) | −0.4 (−0.5) |
The red favorite three or more years younger than the blue underdog, the study’s one rule that clears every bar, is 3.6 points over the line in decisions and 6.6 in finishes. Blue-corner underdogs are 1.7 points under in decisions and 1.3 under in finishes. Nothing about the corner rules changes when the judges are taken out.
The mirror test: a red favorite against a blue favorite at the same price
The filter moves every favorite’s win rate the same way whichever corner it stands in. So if the corner meant nothing, a red favorite and a blue favorite at the same closing price would win the same share of decisions, and the same share of finishes. They do not.
| Favorite’s no-vig chance | Decisions: red favorite | Decisions: blue favorite | Gap | Finishes: red favorite | Finishes: blue favorite | Gap |
|---|---|---|---|---|---|---|
| 50 to 60% | 57.1% (363 of 636) | 50.1% (280 of 559) | +7.0 | 56.4% (329 of 583) | 52.8% (278 of 527) | +3.7 |
| 60 to 70% | 65.8% (515 of 783) | 62.3% (311 of 499) | +3.4 | 63.7% (444 of 697) | 63.4% (300 of 473) | +0.3 |
| 70 to 80% | 77.4% (418 of 540) | 68.9% (166 of 241) | +8.5 | 79.3% (493 of 622) | 71.7% (190 of 265) | +7.6 |
| 80% and up | 87.6% (219 of 250) | 82.2% (60 of 73) | +5.4 | 87.8% (374 of 426) | 87.1% (101 of 116) | +0.7 |
In decisions the red favorite wins 3 to 9 points more than the blue favorite in every price band, z 2.4, 1.3, 2.5 and 1.2 from the shortest band up. In finishes the gaps are 3.7, 0.3, 7.6 and 0.7 points, z 1.2, 0.1, 2.4 and 0.2. Unanimous decisions carry the widest gaps, 6.6, 3.3, 10.0 and 6.9 points; split and majority decisions, with under 175 fights a side in any band, show 8.4, 2.2 and −0.9. The red favorite’s extra wins are not a judges’ gift: they come through knockouts and submissions too.
One number per subset
A logistic regression of “red wins” on the logit of red’s no-vig chance and the age gap gives the red corner one term per subset: its edge in log odds at a pick’em price, with the price and the ages held. Multiply by 25 for points of win chance at 50/50.
| How the fight ended | Fights | Red corner term (log odds) | z | Points at a pick’em |
|---|---|---|---|---|
| All fights | 7,187 | +0.120 ± 0.027 | +4.4 | +3.0 |
| Decisions | 3,520 | +0.136 ± 0.039 | +3.5 | +3.4 |
| Unanimous decisions | 2,758 | +0.145 ± 0.045 | +3.2 | +3.6 |
| Split and majority decisions | 762 | +0.083 ± 0.078 | +1.1 | +2.1 |
| Knockouts and submissions | 3,650 | +0.103 ± 0.039 | +2.6 | +2.6 |
Decisions give +0.136 and finishes +0.103: a difference of 0.033 with a standard error of 0.055, z 0.6. The point estimate leaves the judges less than a point of the red corner’s edge, and the data cannot tell that from zero. The split and majority decisions, where a lean would show first, carry the smallest term of all.
Who gets the nod in a split decision
In the 762 split and majority decisions since 2010 the red corner won 398, 52.2%, against a no-vig closing average of 53.0%, z −0.4, and the favorite won 53.0%. By era: 51.5% against 53.2% in 2010 to 2015, 54.2% against 53.3% in 2016 to 2020, and 51.0% against 52.4% since 2021. If the judges leaned toward the red corner, this is where it would show. It does not, in any era.
Winners by method tell the same story. Red corner wins came 39.4% by unanimous decision, 9.6% by split or majority decision and 51.0% by finish; blue corner wins 37.1%, 12.0% and 50.8%. With the price held, a winning blue favorite needed a split decision more often than a winning red favorite in three of the four price bands, 9.8% against 7.9% overall. If the judges handed red the close fights, red’s wins would include more split decisions, not fewer.
What it means
- Not a judging artifact. The red corner’s edge is the same size when no judge touches the fight, and absent in the fights the judges decide.
- The market, not the scorecards. The UFC’s choice of corner carries information about the fighters the closing line has not fully priced: about two to three points of win chance at a pick’em, in every era.
- “Red in fights likely to go the distance” is not a bet. Women’s fights go the distance 63% of the time and the red corner is 0.9 points over the line in them, z 0.6, −3.9% on $100 flat; five-round fights, 0.4 points. The edge does not grow with the chance of a decision.
- Nothing to change. The study’s rules stand as written: do not bet blue-corner underdogs without a reason, and the red favorite three or more years younger than the blue underdog is still the one rule that clears every bar.
Data and method
- The same 7,187 fights as the UFC betting angles study, March 2010 to October 2026: closing moneylines from BestFightOdds by way of the Ultimate UFC Dataset on Kaggle (CC BY-NC-SA 4.0), with results and methods from UFCStats. 17 disqualifications and overturned results sit in no method row. Decisions: 2,758 unanimous, 709 split, 53 majority. Finishes: 2,341 knockouts and technical knockouts, 58 of them doctor’s stoppages, and 1,309 submissions.
- Points over the line: a rule’s win rate minus the average no-vig closing chance of the sides it bet, in the subset. Its z-score is the surplus of wins divided by its standard error, as in the study. The mirror test’s z is a two-proportion test between the red favorites and the blue favorites in a band.
- The regression is a logistic model of “red wins” on the logit of red’s no-vig chance and the age gap (red minus blue, clipped at 15 years), one row per fight, fitted on each subset; its intercept is the red corner’s edge in log odds at a pick’em. The script is tools/ufc-closing-odds-study/corner_decisions.py in the vigged repository, and docs/ufc-red-corner-study.md holds every table, including the subsets by era.
Frequently asked questions
Do UFC judges favor the red corner?
Not in the data. In 762 split and majority decisions since 2010, the red corner won 52.2% against a 53.0% no-vig closing average, and the favorite won 53.0%. The red corner’s edge over the line shows up in unanimous decisions, 3.4 points, and in knockouts and submissions, 1.4 points, the fights the judges never score.
Does the red corner win more UFC decisions than the blue corner?
At the same price, yes, by about the margin it wins more finishes. A red favorite with a 50 to 60% closing chance won 57.1% of decisions and a blue favorite at the same price 50.1%; in finishes the figures are 56.4% and 52.8%. A regression puts the red corner at 3.4 points of win chance at a pick’em in decisions and 2.6 in finishes, a gap inside the noise.
Is betting the red corner in fights likely to go to a decision an edge?
No. In women’s fights, which go the distance 63% of the time, the red corner won 57.0% against a 56.1% line over 860 fights, z 0.6, and lost 3.9% on flat $100 bets. In five-round fights it was 0.4 points over the line. The red corner’s edge does not grow with the chance of a decision.
Why does filtering by decision change what the closing line expects?
Because the line prices a win by any method and favorites finish fights more often. Among split decisions favorites won only 53.4% against a 64.0% line, since a split decision is by definition a close fight; among unanimous decisions they beat their line by 2.5 points. Every subset is read against that shift, and the mirror test, a red favorite against a blue favorite at the same price, does not depend on it.
Live odds boards: NFL odds · NCAAF odds · NBA odds · NCAAB odds · NHL odds · MLB odds · WNBA odds · MMA odds · Cricket odds · EPL odds · MLS odds · UCL odds · Liga MX odds · La Liga odds · Serie A odds · Bundesliga odds · Ligue 1 odds · ATP odds · WTA odds
21+ | Gambling problem? Call 1-800-GAMBLER. Nothing here is betting advice, and no bet or strategy guarantees a profit. Sports betting is legal only in some US states; check the rules where you are. Responsible gambling.