Curtis Blaydes has a career takedown defence figure of 35%, according to UFCStats, the promotion’s own public record of every fight it has ever run. Read that cold and you would assume he is one of the easiest men in the division to put on his back.
He is also one of the most accomplished offensive wrestlers heavyweight has ever had. Both of those sentences are true at the same time, and the reason they can be is the single most useful thing to understand about MMA statistics.
What the percentage actually counts
Takedown defence is attempts stopped divided by attempts faced. The trouble is entirely in the denominator, because it counts what opponents chose to try — not what a fighter is capable of stopping.
Blaydes is nearly always the one shooting. Anybody who has watched five minutes of his tape knows that wrestling him is a losing plan, so most of his opponents never attempt it. What is left in his denominator is a small pile of attempts from the handful of people who fancied their chances anyway — other high-level wrestlers, usually in the one phase where they had something. A tiny sample made up exclusively of the hardest cases produces an ugly number, and that number tells you nothing about what happens when an ordinary opponent shoots a single leg.
The arithmetic runs the other way too, and more often. A fighter nobody bothers wrestling, who faces one lazy telegraphed attempt a fight, can carry an excellent figure for a whole career without ever being properly tested.
When the number does mean something
The figures worth quoting come with volume behind them. Ilia Topuria sits at 95% and Islam Makhachev at 91%, and both have been shot on repeatedly by opponents who needed the takedown to have any path to winning. That is the right test, applied often.
Makhachev is the more instructive of the two, because he also lands 3.22 takedowns of his own per fifteen minutes while absorbing 1.42 significant strikes a minute. Elite in both directions, while barely getting hit, is a completely different profile from a specialist with one gaudy column — and it is the reason his defensive number is credible rather than merely high.
At the other end, Merab Dvalishvili averages 5.97 takedowns per fifteen minutes. That is an offensive rate, not a defensive one, and reading the two next to each other as if they measure the same skill is the mistake almost everyone makes with this data.
Three questions to ask before you quote any takedown percentage
Which direction does this fighter normally work in? An offensive wrestler's defensive figure is built on a strange, self-selected sample. A striker's is built on a sample of people specifically trying to exploit him, which is at least the test the statistic was designed for.
How many attempts is it built on? Fighters with one or two UFC appearances routinely show 0% or 100%, and both are noise. We exclude those fighters from the statistical rounds of our own Higher or Lower game for exactly this reason — a per-minute rate drawn from a single fight punishes the players who actually know the sport.
Did the takedown accomplish anything? Neither figure records what happened next. A fighter who stands back up in three seconds costs the same on the stat sheet as three minutes of control on the fence, and only one of those wins a round.
How this changes reading a matchup
If you are trying to work out whether a wrestler will impose himself on a striker, the striker's defensive percentage is genuinely useful — that is the exact scenario it was measured from. If you are trying to work out which of two wrestlers ends up on top, both numbers are close to worthless and the tape is the only honest answer.
That asymmetry is why our fight simulator resolves grappling as an exchange between several attributes rather than one number against another. Rolling a takedown rating against a takedown-defence rating produces results that look reasonable on average and are wrong in precisely the fights people care about.
The habit generalises
Takedown defence is the clearest case, but every rate statistic in this sport carries a hidden assumption about what the fighter was trying to do. Striking accuracy punishes anyone whose game is feints and volume from range — Dominick Cruz records 33%, the lowest on our roster, and he is obviously not a bad striker. Strikes absorbed per minute flatters anyone whose fights end early. Submission attempts per fifteen minutes is often highest for fighters who spend a lot of time on their back.
None of this is an argument against the data. It is very good data, it is public, and it is the best record the sport has of what actually happened. It is an argument for asking what the denominator was before you quote the top line.
Sources & method
Every figure in this article comes from ufcstats.com, the UFC’s official public fight database, and reflects career totals as recorded there on 28 August 2026. Career rates move with every appearance, so a fighter who has since competed will read slightly differently. Takedown defence is calculated by UFCStats as the share of opponents’ takedown attempts stopped; takedowns and submission attempts are per fifteen minutes of fight time, and significant strikes are per minute. We hold the same figures for 292 fighters and use them in Higher or Lower, excluding anyone with too few UFC appearances for a per-minute rate to mean anything. Fighter ratings elsewhere on this site are our own model, not UFCStats data.