Which stats actually predict fights — and which just repeat the betting line. Every number is measured across the full UFC history, with the sample size attached. The factor table below is refreshed when we review and publish a new measurement, not on a timer.
Some stats look like they predict fights when really they just point at the betting favourite. So we scored each one across — fights, then scored it again on just the — where both men were priced close to even. If it still wins there, it knows something the odds don't.
“UFC record” here is not the record our pick engine uses. This page tests the plain UFC win–loss gap you can read off a fight card, counted only from fights that had already happened. The engine’s own record factor uses each fighter’s whole professional career, adjusted so a 2–0 fighter isn’t treated as a flawless 100%. Those are different measurements, and they do not agree: tested the same way, the engine’s version lands on a coin flip. So a green light here is not a green light for the pick engine’s record factor, which we have not shown to work on its own.
| Division | Fights | Finish rate | KO/TKO rate | Submission rate |
|---|---|---|---|---|
| Loading… | ||||