UFC 330 odds: where our AI model splits from the books
UFC 330 odds vs our AI model: the books favor Dern, Turner and Fernando, the model disagrees, led by a 28 point gap on Tobias vs Fernando.
Our AI model and the sportsbooks disagree about UFC 330, and not politely. Across the twelve fights at Saturday's card in Philadelphia, five carry a gap of 17 points or more between the model's win probability and the consensus implied probability across the 25 books we track. That is a lot of daylight for one card, and it clusters in interesting places, including both title fights.
A housekeeping note on what these numbers are: the model's probabilities are the same ones rendered on the event page, and the market number is the average implied probability across every book in our odds feed. Where they disagree, one of them is wrong, and the fun of publishing both is that we find out Saturday. This is market-watching, not betting advice.
Rafael Tobias vs Lucas Fernando: the 28 point canyon
The biggest split on the card is on the early prelims. The market prices Lucas Fernando (13-3) around 70 percent over Rafael Tobias (14-2). The model flips it entirely: Tobias 58 percent. A 28 point gap.
What does the model see? Tobias is 23, six years younger, with the better pre-UFC resume, and Fernando's UFC numbers are ugly in exactly the ways the model weighs: he absorbs 4.73 significant strikes a minute, lands only 1.87, and has yet to stop a takedown in the UFC. Tobias lost his own debut on the cards to Diyar Nurgozhay in March, which is presumably what the market is looking at. The model thinks one bad debut against an unbeaten prospect is thinner evidence than Fernando's whole statistical profile.
Notably, the line has been moving the other way: Tobias opened around a 40 percent implied chance and has drifted below 30. The market is running away from the fighter our model rates as the card's biggest sleeper.
Mansur Abdul-Malik vs Dustin Stoltzfus: 87 against 60
The books make Mansur Abdul-Malik (9-1-1) about an 87 percent favorite over Dustin Stoltzfus (16-8), the steepest price on the card. The model agrees on the winner and not at all on the certainty: 60 percent. Abdul-Malik absorbs 4.9 significant strikes a minute with 40 percent striking defense, and his March fight ended with Yousri Belgaroui knocking him out in the third round. Fragile favorites can still be favorites. They should not be 1 to 8.
Mackenzie Dern vs Gillian Robertson: the model picks the challenger
Photo: Canal Faixa Preta / CC BY 3.0
The co-main is the headline disagreement. The market holds champion Mackenzie Dern (16-5) around 68 percent. The model makes Gillian Robertson (17-8) a 52.5 percent favorite, which means our model is on the other side of the fight from the books.
The drivers are almost all grappling, which is supposed to be the champion's department: Robertson averages 2.76 takedowns a fight to Dern's 0.93, completes 40 percent of her attempts to Dern's 18, and betters Dern's striking output and defense besides. She arrives on five straight wins, including a decision over Amanda Lemos, who beat Dern in 2024. I broke the full matchup down earlier this week.
Jalin Turner vs Kaue Fernandes: another side swap
Same shape, smaller gap: the market has Jalin Turner (15-9) around 60 percent, the model has Kaue Fernandes (11-2) at 58. The model's biggest single driver is striking defense: Turner absorbs 4.54 a minute and defends 41 percent, while Fernandes absorbs 1.42 and defends 62. Turner's December knockout of Edson Barboza is the recency the market remembers; the model remembers he is 2-3 in his last five.
Islam Makhachev vs Ian Machado Garry: same winner, very different confidence
The main event gap is 17 points: books near 78 percent for Islam Makhachev (28-1), model at 60.5. No side swap here, just a much less comfortable champion. The model credits the 16 fight win streak and a 97 percent career win rate, then subtracts for age (34.8 against 28.7, with Makhachev flagged past his statistical prime) and size (giving up four inches of reach and five of height to Ian Machado Garry). The market itself has been inching the same direction, from roughly 80 percent at the open to 78 now.
Five gaps, two of them outright side swaps, both of those in fights the books consider comfortable. Saturday grades all of it. Results land on the event page as they happen, and you can put your own picks against the model on the pick'em.