Islam's Turf, Garry's Moment, and a Card Full of Interesting Lies

UFC 330 is headlined by a light heavyweight champion defending his title in a sport where being the best in the world still doesn't make you a sure thing. Except here, Islam Makhachev kind of is. The model has him at 70%, the market has him at 74% — that's about as close to agreement as this version of our model gets with a sportsbook, and when two independent systems land in the same neighborhood, that's usually the market telling the truth.

The more interesting tension is everywhere else. Our biggest divergence on a fully-trusted bout is Edson Barboza, who the market has at 19% and we have at 58%. That's not a rounding error. Then there's Myktybek Orolbai, a wrestler the market loves at 86% who our model sees closer to a coin flip. And Vicente Luque, a guy the market apparently forgot about, moving up to middleweight at 47% when we have him at 77%. This card has opinions, and a few of them are loud ones.

Remember — our model is v0, unvalidated against betting markets. We've beaten a coin flip at 56.9% on holdout data. These are disagreements worth understanding, not instructions.

• Edson Barboza (model: 58%, market: 19%, gap: +38.9pp vs Esteban Ribovics): The market sees a 40-year-old on the wrong side of the hill. Our model sees his rate profile and still finds a fighter who generates 4.19 significant strikes per minute with a 50% career finish rate. Ribovics is genuinely dangerous — 6.14 sig strikes per minute, which is a lot of punches in the air — but his career finish rate sits at just 25%, meaning most of that volume isn't closing fights. The gap here is almost certainly age discount versus actual output. The market is pricing the story; the model is pricing the tape.
• Jeremiah Wells (model: 45%, market: 14%, gap: +31pp vs Myktybek Orolbai): This one comes with a caveat — Wells has only one UFC fight in our sample, so the model is working thin. That said, Orolbai at 86% for a three-round welterweight fight is a steep number. We'll get into the style matchup below, but the short version is: the market may be overcorrecting on Orolbai's grappling dominance.
• Vicente Luque (model: 77%, market: 47%, gap: +30.3pp vs Tresean Gore): Luque is moving to middleweight, and the market has apparently decided that makes him a pick-em. Our model, working off three fights of UFC sample for each fighter, doesn't see it. Luque brings 2.90 sig strikes per minute with 0.53 knockdowns per 15 — real power — and an 88% career finish rate. Gore finishes everyone he beats too, but the model isn't buying the coin-flip framing here.
• Neil Magny (model: 67%, market: 47%, gap: +20.5pp vs Ramiz Brahimaj): Magny is 39 and the market is almost certainly discounting him on age. But 3.18 sig strikes per minute, 2.52 takedowns per 15 minutes, and 37 UFC fights of history for the model to read — that's not a ghost of a fighter, that's a profile. Brahimaj's 1.55 sig strikes per minute is the second-lowest volume on this entire card.
• Gillian Robertson (model: 54%, market: 33%, gap: +20.3pp vs Mackenzie Dern): A five-round co-main and the market has Robertson as a significant underdog. The model sees something closer to a toss-up — 54/46. Robertson's 2.42 takedowns per 15 minutes against Dern's 1.29 is a meaningful grappling volume edge, which matters in a five-round fight. This one's worth watching closely.

These are the fighters the model believes in most, ranked by confidence. The model and market are sometimes aligned here, sometimes not — noted either way.

1. Vicente Luque (model: 77%, market: 47%): The biggest conviction play on a fully-trusted bout where the model and market genuinely disagree. Luque's finish rate is 88% across his career, and he generates legitimate knockdown threat at 0.53 per 15 minutes. The model strongly believes in this result even if the market doesn't.
2. Islam Makhachev (model: 70%, market: 74%): Essentially full agreement. The model likes Makhachev for straightforward reasons — 2.01 takedowns per 15 minutes, 65% career finish rate, and 18 UFC fights of history to read. When two systems this different land this close, you pay attention.
3. Joel Álvarez (model: 68%, market: 70%): Another near-consensus. Álvarez at 4.15 sig strikes per minute and an 88% career finish rate is a lot of fighter. The model and market are basically holding hands on this one.
4. Neil Magny (model: 67%, market: 47%): The model is more bullish here than the market, primarily on volume and grappling output. 2.52 takedowns per 15 against a fighter posting 1.55 sig strikes per minute is a favorable shape.
5. Edson Barboza (model: 58%, market: 19%): The model's biggest surprise on the card. Not a high-conviction win probability in absolute terms, but the disagreement with the market is the story. The model sees a live fighter; the market sees a name past its date.

A few favorites the model is cooler on than the market, and then the fights where we're all just guessing.

Model-vs-Market Caution Flags

• Myktybek Orolbai at 86% market, 55% model: That's a 31-point gap on a trusted bout. The model isn't calling an upset — it's saying this fight is closer to even money than the price suggests. Worth knowing before you get comfortable.
• Tresean Gore at 53% market, 23% model: The market sees a coin flip. The model sees a significant Luque advantage. One of them is quite wrong.
• Esteban Ribovics at 81% market, 42% model: The largest gap against a market favorite on a trusted bout. Nearly 39 points. The model is not buying Ribovics as a 4-to-1 favorite over Barboza.

Low-Data Bouts — Anyone Claiming an Edge Is Guessing, Including Us

Five fights on this card are flagged LOW DATA, meaning debutants or fighters with minimal UFC sample. The model is essentially blind here, and we're deferring entirely to the market price as the best available signal:

• Lucas Fernando vs. Rafael Tobias (Light Heavyweight): Fernando has zero stat-sample fights in our system. Hard pass on the model number.
• Kauê Fernandes vs. Jalin Turner (Lightweight): Limited UFC history on Fernandes. Turner is the more readable fighter, but the model's sample is thin for both.
• Jose Ochoa vs. Charles Johnson (Flyweight): The market has Ochoa at 77%, the model at 47% — a 30-point gap — but this is a low-data flag, so that divergence means very little. Don't read into it.
• Eric McConico vs. Donte Johnson (Middleweight): McConico has only two UFC fights in sample; Johnson has only two as well. Not enough to trust.
• Dustin Stoltzfus vs. Mansur Abdul-Malik (Middleweight): Abdul-Malik is four fights deep in our system and carries a 100% career finish rate, which is interesting, but the low-data flag keeps the model humble.

Islam Makhachev vs. Ian Machado Garry — Welterweight Championship (5 Rounds)

Islam Makhachev doesn't beat you with volume — 2.48 sig strikes per minute is well below average — he beats you by removing your freedom. At 2.01 takedowns per 15 minutes across 18 UFC fights of history, he's the most prolific wrestling threat on this card. Ian Machado Garry generates 4.34 sig strikes per minute and attempts 1.47 takedowns per 15 himself, so he's not passive. But Garry's career finish rate sits at 30% — he gets decisions. Makhachev finishes 65% of the time. The stylistic read: Garry needs to stay upright and outpoint a grappler for five rounds. That's not impossible — it's just a very specific and demanding task.

Chidi Njokuani vs. Joel Álvarez — Welterweight (3 Rounds)

Chidi Njokuani at 6.40 sig strikes per minute is the highest volume striker on this entire card — and he adds 0.56 knockdowns per 15 minutes, which is genuine pop. The problem is he's 37, 25-12, and facing Joel Álvarez, who throws 4.15 per minute with an 88% career finish rate. This is two active, dangerous strikers in a phone booth. Njokuani has the volume edge; Álvarez has the finishing edge. At 3 rounds it likely ends before the cards.

Vicente Luque vs. Tresean Gore — Middleweight (3 Rounds)

The most interesting setup on the undercard. Vicente Luque moving to middleweight, 2.90 sig strikes per minute, 0.53 knockdowns per 15, 88% career finish rate. Tresean Gore has a 100% career finish rate too — he doesn't go to decisions either. Gore generates 3.71 sig strikes per minute with some grappling at 0.97 takedowns per 15. Both men finish. The difference the model is leaning on is Luque's knockdown rate and his 25 UFC fights of experience versus Gore's 7. Experience matters when both guys are hunting the finish.

Gillian Robertson vs. Mackenzie Dern — Women's Strawweight (5 Rounds)

Two grapplers in a five-round co-main, which should tell you something. Gillian Robertson leads with 2.42 takedowns per 15 minutes against Mackenzie Dern's 1.29 — Robertson is the more active wrestler by that measure. Dern counters with 3.68 sig strikes per minute, noticeably higher than Robertson's 2.49, so Dern has the striking volume edge if the fight stays standing. Dern's career finish rate is 45%, Robertson's is 71%. Five rounds gives both grappling styles room to breathe, and Robertson's finish rate suggests she does something with her opportunities.

Myktybek Orolbai vs. Jeremiah Wells — Welterweight (3 Rounds)

Myktybek Orolbai's 11.43 takedowns per 15 minutes is not a typo, and it's the single most eye-catching rate number on this card. At 28, with a 60% career finish rate, the market has correctly identified that when he gets his hands on you, things go badly for you. Jeremiah Wells is 39, and his UFC sample is thin — just one fight in our system — so take the model's number with appropriate skepticism. What the model sees is a fight the market has made very one-sided, and Orolbai's 2.17 sig strikes per minute means if Wells can neutralize the wrestling, he's facing a lighter volume striker. That's the only path, and it's a narrow one.

Edson Barboza vs. Esteban Ribovics — Lightweight (3 Rounds)

The fight where the market and model disagree most on a trusted bout. Esteban Ribovics at 6.14 sig strikes per minute is genuinely high output, but his 25% career finish rate means most of it doesn't end fights — at 15-3 he wins decisions. Edson Barboza throws 4.19 per minute with a 50% finish rate, a six-inch reach advantage, and 32 UFC fights of history behind him. The market is pricing age heavily here. The model is less convinced the 40-year-old version of Barboza is as diminished as a 19% win probability implies.

A few profiles worth building around, framed by scoring potential rather than lineup construction.

• Myktybek Orolbai — 98.9 projection, 183.7 ceiling. That ceiling is the highest on the card, by a comfortable margin. A grappler posting 11.43 takedowns per 15 minutes will rack up control time, advances, and submission attempts if the fight goes to the mat — which it usually does. The projection reflects consistent floor; the ceiling reflects what happens when the wrestling is going well and the fight doesn't end quickly.
• Islam Makhachev — 93.7 projection, 144.2 ceiling. The highest floor among main-card fighters. Championship rounds, consistent takedowns, and a 65% finish rate give him multiple paths to a high-scoring night. If he finishes early, ceiling applies. If it goes long, control time piles up.
• Chidi Njokuani — 44.8 projection, 109.6 ceiling. Lower projection reflects the likelihood of a loss, but if Njokuani is the one landing — 6.40 sig strikes per minute with 0.56 knockdowns per 15 — his ceiling in a winning performance is real. A boom-or-bust profile in the truest sense.
• Gillian Robertson — 79.9 projection, 135.8 ceiling. Five rounds of grappling at 2.42 takedowns per 15 minutes adds up. Robertson's 71% finish rate means there's a real ceiling outcome here too. The projection is strong, the ceiling is strong, and the market has her as a significant underdog — which generally means she's lower-owned.
• Joel Álvarez — 78.7 projection, 121.1 ceiling. High floor, high ceiling, strong market consensus. The model and market agree he wins. At 4.15 sig strikes per minute and 88% finish rate, his scoring path is straightforward. The chalk option with a clean profile.
• Jalin Turner — 69.4 projection, 119.9 ceiling, 100% career finish rate. Every single one of Turner's wins has been a finish. If he wins, something dramatic happens. Low-data flag applies — the model is working thin here — but the career finish rate is a profile fact, not a model output.

Three things to take into fight night:

1. The main event is the one market consensus we mostly share. Makhachev at 70% model, 74% market — that's agreement, not rubber-stamping. When a grappler with 2.01 takedowns per 15 minutes faces a 30% finisher in a five-round title fight, the stylistic math makes sense. The market isn't lying here.
2. The card's biggest story might be Barboza. Our model has him at 58%, the market has him at 19%. That's 39 points of daylight on a trusted bout. The model is v0 and unvalidated, so this is not a recommendation — but it is the most significant disagreement between our numbers and the price on this entire card. If you're watching one underdog closely, you know which one the model is watching.
3. Five fights are low data, and no one has a real edge there. Fernando-Tobias, Fernandes-Turner, Ochoa-Johnson, McConico-Johnson, Stoltzfus-Abdul-Malik — the model is functionally blind, and anyone claiming strong conviction on those is working from vibes. The market price is the best available signal. Lean on it and move on.