Ask a room of MMA fans when a fighter peaks and you will get a fairly narrow answer. Late twenties, probably. Twenty-eight to thirty-two if they want to be precise. Old enough to have learned the craft, young enough to still own the body. Commentators reach for the phrase "in his prime" a dozen times a card and nobody ever asks them to define it.
It can be defined. The MMAHIT archive holds 1,049,038 recorded bouts, and in 436,867 of them we know both the date of the fight and the birth date of one of the fighters in it. That is enough to ask a very simple question: how does winning change with age?
The answer is not the late twenties. It is not close to the late twenties.
The curve
| Age band | Bouts | Win rate | Of those wins, finished |
|---|---|---|---|
| 18–23 | 124,143 | 60.6% | 74.1% |
| 24–29 | 181,972 | 58.3% | 70.7% |
| 30–34 | 86,095 | 50.6% | 67.7% |
| 35–39 | 29,186 | 43.3% | 67.4% |
| 40+ | 8,200 | 38.7% | 72.1% |
Measured year by year, the single strongest age in the entire archive is 22, at 61.4 percent. By 30 it is 53.1. By 35, 44.9. By 40, 39.3.
There is no plateau at 28. There is no second bump at 31. From the middle of the twenties the line moves in one direction and keeps moving.
Why that table does not prove it on its own
Here I have to stop and point at a hole in the table big enough to drive a promotion through.
Young fighters are not matched the way old ones are. A 21-year-old prospect is handed opponents chosen to make him look good, because prospects are investments and nobody wants to lose one early. A 38-year-old veteran gets booked against whoever the promotion actually wants tested, and increasingly he is the test. Some of that 60.6 percent is youth. Some of it is matchmaking. The raw curve cannot separate the two.
So we ran the measurement that can.
The clean test: put both ages inside the same fight
Rather than compare age groups across different bouts, compare the two fighters standing in one bout. Same night, same rules, same promotion, and — by definition — the same opponent quality, because they are each other's opponent. The only variable left is which of the two has been alive longer.
- Older by any margin — won 43.1% of 144,334 bouts
- Older by 3+ years — won 40.4% of 87,126 bouts
- Older by 8+ years — won 35.5% of 27,683 bouts
That is a dose-response curve, and dose-response is what separates a real effect from an artefact of how fights get made. The wider the age gap, the worse the older fighter does, in a straight line, across 144,334 bouts that controlled for everything else by construction.
Put plainly: being the older man in an MMA fight is worth about seven percentage points off a coin flip. Being eight years older is worth nearly fifteen.
And then it gets strange
Go back to the first table and read the last column downward.
74.1 → 70.7 → 67.7 → 67.4 → 72.1
The finish rate declines with age exactly as you would expect, and then at 40 it turns around and climbs most of the way back. Break the finishes into types and the oldest group in the archive owns the highest knockout share of any age band in the sport:
| Age band | KO/TKO | Submission | Decision |
|---|---|---|---|
| 18–23 | 37.2% | 36.9% | 23.7% |
| 24–29 | 37.3% | 33.4% | 27.4% |
| 30–34 | 37.2% | 30.5% | 30.2% |
| 35–39 | 37.5% | 30.0% | 30.0% |
| 40+ | 39.8% | 32.3% | 26.2% |
So the fighters least likely to win are, in the fights they do win, the most likely to end things violently. That should not be possible. Forty-year-old athletes do not hit harder than twenty-two-year-olds.
The funnel explains it
They don't. What changed is not the fighters — it is the population being measured.
Count how many distinct fighters actually compete at each age:
- 25 — 16,873 fighters
- 30 — 12,802
- 35 — 5,492
- 40 — 1,431
- 45 — 315
- 50 — 59
Two hundred and eighty-six men compete at 25 for every one still competing at 50. That is not a career arc, it is a sieve. And the sieve does not select at random — it keeps the only two kinds of fighter a promoter will still book past forty: the ones who are genuinely still dangerous, and the ones who will still take the fight.
The archive holds both, and they are easy to tell apart. Dan Severn recorded 94 bouts after his fortieth birthday and won 78 of them. Tony Lopez has 54 post-40 bouts and 32 wins, the most recent in March 2026. Jeff Monson managed 35 and won 20. Those are the dangerous ones.
The same query also returns Tetsuo Kondo — 38 bouts after 40, five wins — and Carlos Eduardo Viana Gomes, 38 bouts and five wins. Those are the willing ones.
The knockout rate of the 40-plus band is simply the first group speaking louder than the second. The men who still finish people are the men who still get offered fights; everyone else stops appearing in the data entirely. Nobody defies age in this archive. They just leave it at different times.
What this is actually good for
"In his prime" is doing no work. When a broadcast tells you a 31-year-old is entering his prime, the archive says he left it around six years earlier. The phrase marks a commercial window — name recognition, title contention, pay — not a physical one.
Careers begin almost exactly where they peak. The most common debut ages in the archive are 21 (6,269 fighters), 22 (6,222) and 20 (6,144). Which means the average fighter's statistically best year arrives within a season or two of his first professional bout — usually years before anyone outside his own gym has heard of him.
The age gap is a real handicap and almost nobody mentions it. Height, reach and record all get read out before a fight. A six-year age difference moves the odds more than most of them, and it rarely gets a sentence.
A 40-year-old still finishing opponents is not beating age. He is what remains after age removed everybody else.
Method: figures drawn from the MMAHIT fight archive. Age is computed as the interval between a fighter's recorded birth date and the recorded date of the bout, giving 436,867 fight-records across 73,592 distinct fighters, for events dated April 1980 to May 2026. Draws and no-contests are excluded from win-rate calculations. One limitation is worth stating plainly: only 75,295 of the archive's 347,296 fighters carry a known birth date, and better-documented fighters tend to be better-known ones, so the age sample leans toward fighters who reached a level where somebody recorded their details. The head-to-head age-gap test is included precisely because it is immune to that bias — both fighters in every one of those 144,334 bouts are drawn from the same sample and matched against each other.
