September 2, 2026
UFC finish rates by division: why the same number means different things
A 60 percent finish rate is exceptional at strawweight and below average at light heavyweight. Finish rates for all eleven active UFC weight classes, why the women's divisions cannot be ranked, and why heavyweight's number is falling.
Say a fighter ends 60 percent of his bouts inside the distance.
At strawweight that is remarkable. Barely a third of strawweight fights end early, so he is running at nearly twice the rate of the division around him.
At light heavyweight that same 60 percent is slightly below average. Light heavyweight finishes 63 percent of its fights on its own. He would be doing a little less than the weight class does without him.
Same number. Opposite verdict. Nothing about the fighter changed between those two sentences, only the thing we held him up against.
That is the problem with a finish rate quoted on its own, and it is worth knowing what the backdrop actually looks like before you read another one.
What each division actually does
Some weight classes end fights early far more often than others, and the gap between the extremes is wide.
| Division | Finish rate | Fights |
|---|---|---|
| Light Heavyweight | 63.3% | 390 |
| Heavyweight | 58.5% | 388 |
| Middleweight | 56.4% | 590 |
| Lightweight | 53.1% | 717 |
| Welterweight | 50.3% | 688 |
| Flyweight | 47.2% | 316 |
| Featherweight | 46.6% | 594 |
| Bantamweight | 44.0% | 571 |
Light heavyweight finishes 63 percent of its fights, across 390 of them. Bantamweight finishes 44 percent, across 571. Same sport, same rules, same three or five rounds, and a 19 point gap in how often you see a decision read out. Both ends of that range rest on hundreds of fights, and they are nowhere near close enough to be a fluke.
The middle of the table is softer than the ends. Heavyweight, middleweight, lightweight, and welterweight land in order, but several of those neighbours sit within a few points of each other on samples of a few hundred fights. Read the range and its endpoints. Do not read too much into any single position.
Featherweight and flyweight are the clearest case. They finish 47 percent and 47 percent, close enough that which one sits higher is not a real question. Treat them as tied.
The women's divisions do not separate
The three active women's divisions come out at 37 percent, 37 percent, and 34 percent.
| Division | Finish rate | Fights |
|---|---|---|
| Flyweight | 36.7% | 278 |
| Bantamweight | 36.5% | 189 |
| Strawweight | 33.5% | 334 |
Replay the last ten years of women's MMA with slightly different luck and these three divisions could come out in any order. That is not a hedge. It is what the numbers say.
Flyweight and bantamweight are separated by less than a fifth of one percentage point, which is why they round to the same figure above. Meanwhile women's bantamweight has 189 fights on record, and a sample that small carries enough wobble that its real rate could sit anywhere from about 30 percent to about 43 percent. The gap we are trying to measure is far smaller than the error bars around it.
So we are not going to rank them.
The one thing that does stand out is the level. All three women's divisions sit below all eight men's divisions. That pattern holds across every comparison, though the closest pair, men's bantamweight at 44 percent and women's flyweight at 37 percent, are near enough that we would not lean hard on those two specifically.
Heavyweight is not what it was
Look up heavyweight's all time finish rate and you will find a number in the mid sixties. That figure is correct, and it is also averaging together two very different eras.
Heavyweight's most recent 100 fights finished at 54 percent. At its peak, that same rolling figure reached 78 percent.
| Year | Heavyweight |
|---|---|
| 2003 | 73% |
| 2004 | 73% |
| 2005 | 74% |
| 2006 | 76% |
| 2007 | 74% |
| 2008 | 74% |
| 2009 | 73% |
| 2010 | 72% |
| 2011 | 75% |
| 2012 | 75% |
| 2013 | 75% |
| 2014 | 75% |
| 2015 | 72% |
| 2016 | 67% |
| 2017 | 64% |
| 2018 | 63% |
| 2019 | 60% |
| 2020 | 62% |
| 2021 | 60% |
| 2022 | 62% |
| 2023 | 56% |
| 2024 | 57% |
| 2025 | 52% |
| 2026, to 29 August | 54% |
Broken into five year blocks, every block through 2014 sat between 72 and 77 percent. Then 63 percent for 2015 to 2019, 58 percent for 2020 to 2024, and 54 percent so far this decade, though that last figure rests on only 68 fights and will move.
What we cannot tell you is when the slide started or whether it was a turn or a drift. We tested both shapes against the data and they fit equally well. A gradual decline and a sudden break are indistinguishable here. What is clear is that heavyweight left the seventies and has not been back.
It would be easy to assume this is just what the sport does now, and that everything has drifted toward the judges. It is not.
Light heavyweight, the division directly below, moves around about as much as heavyweight does. It drops, climbs back, drops again. But it never goes anywhere. Its recent stretch sits close to where it was two decades ago, and when we tested 21 possible turning points across its history, not one of them held up. Light heavyweight is noisy. It is not trending.
That is the distinction worth holding onto. Both divisions bounce around. Heavyweight's bouncing has a direction and light heavyweight's does not.
Middleweight is a third case again. It fell to the mid fifties around 2009 and has held there since. It moved, then stopped.
Three divisions next to each other on the scale, doing three different things. Whatever is happening at heavyweight is specific to heavyweight, at least across the three we tested. We cannot tell you what it is. Better defensive striking, deeper talent, more complete athletes at the top of the division. All plausible, none of them provable from fight results alone.
What to do with a finish rate
A finish rate is not a fact about a fighter. It is a fighter measured against a backdrop, and the backdrop moves.
Compare two fighters from different weight classes on their raw finish rates and you are holding them up against two different backdrops while calling it a fair comparison. That is the whole thing, and it costs you nothing to fix. Look up what the division does first, then read the fighter against it. Our fighter stats pages carry the underlying records if you want to work one out yourself.
What it will not do is tell you who wins on Saturday. About half of all UFC fights end early, and knowing the weight class moves that to somewhere between 34 and 63 percent. Real, useful, and nowhere near a prediction. Any single fight is still mostly unpredictable, and no amount of base rate work changes that.
Worth carrying:
- Compare a fighter's finish rate to their own division, not to whoever else is on the card
- Check when those fights happened, because heavyweight's number ten years ago is not heavyweight's number today
- Give short records a wide berth. A fighter with a dozen bouts has a rate that a couple of results either way would move noticeably
- Be clear what is being counted. A fighter's own finishing record and the share of his bouts that ended early are different numbers, and the second one includes the times he was the one being stopped
None of that makes a number more accurate. It makes it mean something. A finish rate on its own is just a fraction, and it stays a fraction until you know what the fights around it were doing.
The full table
All eleven active divisions, June 2016 to August 2026. A finish is a knockout, a TKO, or a submission.
| Division | Finish rate | Fights | KO/TKO | Submission | Decision |
|---|---|---|---|---|---|
| Light Heavyweight | 63% | 390 | 182 | 65 | 136 |
| Heavyweight | 59% | 388 | 173 | 54 | 150 |
| Middleweight | 56% | 590 | 223 | 110 | 246 |
| Lightweight | 53% | 717 | 253 | 128 | 330 |
| Welterweight | 50% | 688 | 235 | 111 | 333 |
| Featherweight | 47% | 594 | 183 | 94 | 314 |
| Flyweight | 47% | 316 | 76 | 73 | 163 |
| Bantamweight | 44% | 571 | 152 | 99 | 311 |
| Women's Flyweight | 37% | 278 | 47 | 55 | 176 |
| Women's Bantamweight | 37% | 189 | 36 | 33 | 119 |
| Women's Strawweight | 34% | 334 | 49 | 63 | 220 |
Rows do not always add to the fight count, because a small number of bouts end in disqualifications or are recorded with no result. Those stay in the fight count but are counted as neither a finish nor a decision.
How these numbers were built
Everything here covers UFC fights from 18 June 2016 to 29 August 2026. Fights after that date are not included, and every figure here will move as new fights land.
That start date is not a rule change or a turning point in the sport. It is the date of the first women's flyweight fight, which makes it the earliest point at which all eleven current divisions existed at the same time. Start any earlier and you are putting one division's entire history against another division's recent years, which is not a fair comparison.
A few other things worth knowing:
- A finish means a knockout, a TKO, or a submission. Knockouts and TKOs are recorded together in the public data and cannot be separated. Fights stopped by a doctor are recorded separately at source and folded in here, which matches how the official record counts them
- These are bout level rates. A division's finish rate is the share of its fights that ended early, not anyone's individual finishing record
- Draws are counted as fights that went to the judges, because they did
- Which divisions count as active was taken from the UFC's published rankings, rather than inferred from which divisions had recent fights. Those two methods disagree: women's featherweight had a bout inside the last two years but has no current rankings presence, so a recency rule would have included a division the promotion no longer runs
- Catch weight bouts are excluded. A catch weight is not a division and has no base rate to compare against
- Divisions the UFC no longer runs are excluded: open weight, super heavyweight, and women's featherweight
- Contender Series and Road to UFC fights are excluded. They sit in the same public data source as UFC events and are filtered out by event name, which works but depends on those events being named recognisably
- The rolling window figures depend slightly on how fights from the same night are ordered, which the public data does not record. Where that matters, it moves a number by about one point
Where a number here is shakier than it looks, we have said so. The women's divisions are the clearest case: three rates that appear ranked, and not enough fights on record to rank them.