Fight Stats · Predictions

Can the data pick a fight?

Before every card we freeze a set of probabilities, and after it we score them. The record below is the answer, accumulating over years. It is published for analysis and curiosity only — never as betting advice, and we will never publish odds or compare against a bookmaker’s line.

The card calendar

8 cards
Sat 26 SeptlockedUFC Fight NightRaul Rosas Jr. v Raoni Barcelos6 of 12 called · 6 left aloneSee the predictions →Sat 3 OctlockedUFC 332Natalia Silva v Wang Cong4 of 13 called · 9 left alone · title boutSee the predictions →Sat 10 OctlockedUFC Fight NightBrendan Allen v Christian Leroy Duncan1 of 1 calledSee the predictions →Sat 17 OctlockedUFC Fight NightJoaquin Buckley v Mike Malott6 of 12 called · 6 left aloneSee the predictions →Sat 24 OctlockedUFC 333Alexander Volkanovski v Movsar Evloev7 of 9 called · 2 left alone · 2 title boutsSee the predictions →Sat 31 OctlockedUFC Fight NightRenato Moicano v Tom Nolan0 of 3 called · 3 left aloneSee the predictions →Sat 7 NovlockedUFC Fight NightGabriel Bonfim v Sean Brady1 of 1 calledSee the predictions →Sat 14 NovlockedUFC 334Ciryl Gane v Josh Hokit9 of 12 called · 3 left alone · 2 title boutsSee the predictions →

The running record

46 scored predictions
Finish or distance?47.8%Brier 0.272 · the claim with real evidence behind it
Who wins?56.5%Brier 0.252 · barely beats a one-line rule, and may not at all
Round one?34.8%Brier 0.273 · of the 23 that finished
Reading it

A Brier score of 0.250 is what you get for saying “50%” to everything. Lower is better. We publish it because a hit rate alone cannot tell a confident wrong model from a hedging right one.

Settled

most recent 12
Petr YanvMerab DvalishviliUFC 333
We said69% distance
It wasa decision
Montel JacksonvRicky SimonUFC Fight Night
We said70% distance
It wasa decision
Joshua VanvAlexandre PantojaCrypto.com UFC 331
We said50% early
It wasa finish
Arman TsarukyanvMauricio RuffyCrypto.com UFC 331
We said51% distance
It wasa finish
Patricio FreirevDooho ChoiCrypto.com UFC 331
We said52% distance
It wasa finish
Renato MoicanovBrian OrtegaCrypto.com UFC 331
We said58% early
It wasa finish
Alonzo MenifieldvIwo BaraniewskiCrypto.com UFC 331
We said66% early
It wasa decision
Marlon VeravCharles JourdainCrypto.com UFC 331
We said64% distance
It wasa finish
Tai TuivasavRobelis DespaigneCrypto.com UFC 331
We said55% early
It wasa decision
Edmen ShahbazyanvBrunno FerreiraCrypto.com UFC 331
We said67% early
It wasa decision
Casey O'NeillvEduarda MouraCrypto.com UFC 331
We said64% distance
It wasa finish
Giga ChikadzevJoanderson BritoCrypto.com UFC 331
We said63% distance
It wasa finish

How this works, and what it cannot do

Every fighter carries a profile built only from fights that had already happened when the prediction was made — never their career averages, which would contain the result being predicted. The models are refitted for each card against everything before that date.

“Will it finish?” is the claim worth reading. Tested on a held-out window it reached 60.7% against a 52% baseline, and its probabilities are well calibrated: fights called at 60–70% finished 68.5% of the time.

“Who wins?” is weak, and we would rather say so. On the same test it reached 62.6% — which is not statistically distinguishable from simply picking the fighter with the better UFC win rate. It is published for the record, not because it is clever.

Knockout or submission, we say only sometimes. Across all finishes the model is barely better than each division’s own knockout share and is worse on accuracy, so a probability printed on every bout would be meaningless. Above 75% confidence it earns its place: on the finishes it flags, the real knockout rate ran 8 to 14 points clear of what the division alone predicts, in every one of five walk-forward test windows, covering about a quarter of finishes. The call appears on those bouts and nowhere else.

It almost never calls a submission, and that is the finding rather than a gap. Not one bout in the held-out window reached 75% that way. A knockout is written into two fighters’ careers long before the cage door shuts; a submission tends to come out of a scramble no record predicts. Every bout still shows its division’s historical mix, which is context and not a claim.

It has no concept of who a fighter has beaten. A dominant champion and a fighter with the same record against far weaker opposition look identical to it, which is the single biggest thing missing and the next thing we will add.

Where a fighter is making their UFC debut, is a late replacement, or has too few fights with statistics, we publish no prediction at all rather than guess. Those bouts are shown as abstentions.

This exists to answer a question about data, not to help anyone gamble. We publish no odds, make no comparison to bookmakers’ lines, and none of it should be used to stake money. The insight stories explain the analysis these models are built from.