UFC 334 predictions

Radtke and Thompson are close to opposites in how they work. Thompson is the volume striker, landing 4.01 significant strikes a minute to Radtke's 2.66, and he has three inches of height and reach on him. Radtke is the steadier all-rounder. He absorbs less (2.62 a minute vs 3.01), shoots 1.85 takedowns per 15 minutes at 41%, and holds control for 47% of his fight time. Thompson rarely wrestles and has defended only 63% of takedowns against Radtke's 75%. Both men have been knocked out twice, and Radtke's finish rate is higher (39% vs 27%).

Our model leans Radtke at 55%, mostly because of wrestling and control. His takedown game and 75% takedown defence give him a way to neutralise Thompson's range, and his 67% smoothed win rate and two-fight win streak beat Thompson's three-fight skid. Thompson's higher Elo (1575 vs 1545) keeps it close. The market sees it the other way, with Thompson at 54%, a nine-point gap. The likely reason is that Thompson's 22 UFC fights against top competition carry weight the model may undervalue. The model includes layoff and age, but may underweight Thompson's 490 days out and age of 43.8, as well as Radtke's age of 36.4 and 217 days off.

Key question: can Radtke get Thompson to the mat often enough to blunt his volume, or does Thompson's range keep the fight standing?

Chris Duncan and Drew Dober throw at nearly the same pace, about 4.4 significant strikes a minute each, and both absorb a little over 4 and defend about 53%. The gap is in how they get there. Duncan lands 46% of his strikes to Dober's 41%, and he is the far more active wrestler, attempting 2.5 takedowns per 15 minutes against Dober's 0.65. Dober carries the bigger knockdown rate (0.65 vs 0.40) and a 45% finish rate, but he has been knocked out 3 times and submitted 4 in 26 UFC fights. Duncan has the size edge, with 2 inches of height and 1 inch of reach.

Our model gives Duncan 72%, and the market has him at 56%. That 16-point gap is large. The model leans on Duncan's better striking accuracy, his wrestling and control edge, and Dober's age of 38 plus his history of getting finished. The market may be giving more credit to Dober's 26-fight experience, his current two-fight win streak and his power, and the model may be underweighting his age and layoff, which are about 252 days for him and 224 for Duncan. Duncan is also coming off a loss.

Key question: can Duncan turn this into a grappling fight, or does Dober land his power first?

Jim Miller and Terrance McKinney are close to opposites in pace. McKinney throws at more than double Miller's rate, landing 6.54 significant strikes a minute to Miller's 2.93, with 55% accuracy and 0.99 knockdowns per 15 minutes. He also absorbs a lot (4.32 a minute, only 44% defence) and has been knocked out four times. Miller is the patient veteran with 46 UFC fights, a 43% finish rate and 1.68 submission attempts per 15 minutes, though he has been submitted three times. McKinney also wrestles more, with 2.52 takedowns per 15 minutes and 69% takedown defence against Miller's 49%. McKinney has the 2-inch height and reach edge.

Our model favours McKinney at 60%, mainly because of his output gap, his better takedown defence and the age difference of 43.2 versus 32.2. Miller's Elo of 1616 is higher than McKinney's 1532, which is why this is not more lopsided. The market is at 70%, 10 points more confident than we are. The model may underweight how age and Miller's 189-day layoff affect a 43-year-old, and it can't see camp or recent news.

Key question: Can Miller survive McKinney's early pressure long enough to drag him into a grappling fight?

Cesar Almeida is the taller man at 73 inches to Donte Johnson's 68, though their reach is identical at 74. Almeida lands a bit more on the feet (3.44 significant strikes a minute to 3.04) and is the more accurate striker at 58%, but he is also hit more often, with just 45% striking defence against Johnson's 60%. Johnson brings the wrestling: 1.70 takedowns per 15 minutes, 72% takedown defence and 26% control time, against Almeida's 52% takedown defence. Johnson's finish rate is 50% to Almeida's 31%.

Our model leans Johnson at 77% because of three things: his edge in striking defence, his clear grappling advantage, and a large age gap (27.8 versus 38.6), with Almeida also on a two-fight losing streak against Johnson's three straight wins. The market is far less sure, giving Johnson 58%, a 19-point gap. The likeliest reason is that the model only sees UFC fights, and Johnson has just three of them, so his numbers are a small sample blended toward league averages. The model does include age and layoff, but it may underweight how much Almeida's size and power can matter in a single exchange.

Key question: can Almeida keep the fight standing and use his height and accuracy, or does Johnson take him down and make age and grappling decide it?

Mesquita brings the sharper, more complete profile on paper. She lands about 3.9 significant strikes a minute and absorbs only 2.4, with 63% striking defence, while Chiasson lands 3.2, absorbs 2.8 and defends just 46%. Mesquita is also the more active grappler, with 2.3 takedowns and 1.9 submission attempts per 15 minutes, though Chiasson's 63% takedown defence is a bit better than hers. Chiasson owns the visible size edge, at 71 inches tall with a 72-inch reach, and has far more UFC experience with 14 fights. Mesquita's sample is only 3 fights, so her numbers are heavily blended toward league averages and shouldn't be read as a settled profile.

Our model likes Mesquita at 62% mainly because of the gap in striking defence, her better absorbed-to-landed ratio, and her finishing record (70% finish rate, never KO'd or submitted) against Chiasson's 34% and a current three-fight skid. The market is close at 66%, so there's no real disagreement. Chiasson's 259 days off and Mesquita's 147 are both in the model, though it may underweight them.

Key question: can Mesquita's volume and submission threat hold up against Chiasson's size, or does the reach edge make it a longer night than the numbers suggest?

Jared Gordon is the busier striker, landing 5.72 significant strikes a minute to Melquizael Costa's 4.17, and he leans on takedowns at nearly two per 15 minutes with 33% control time. Costa is the more dangerous finisher, with a 38% finish rate against Gordon's 18%, plus 0.80 submission attempts per 15 minutes and a bit more reach (71 inches to 68). Gordon has been knocked out four times and submitted twice in 17 UFC fights, which gives Costa clear paths to a finish. Both men are coming off losses.

Our model leans Costa at 68%, driven by three things: his higher Elo (1585 to 1525), his finishing threat, and the age gap of 30 against Gordon's 38. Gordon's volume and wrestling keep him live, and the model may underweight how much his experience matters. The market is more confident, at 78% for Costa. That 10-point gap likely reflects how much weight bettors put on Gordon's age and durability concerns, which our model only partly captures.

Key question: can Gordon's volume and wrestling control the pace for 15 minutes without walking into Costa's finishing chances?

Gane is the longer, cleaner striker here. He lands 5.24 significant strikes a minute and absorbs just 2.38, with 61% defence, and he has 3 inches of height and 8 inches of reach on Hokit. Hokit is the volume brawler: 7.95 landed a minute, but he absorbs 5.18 and defends only 49%. He also brings real power (1.08 knockdowns per 15 minutes) and a bigger wrestling game, with 1.43 takedowns per 15 and 26% control time against Gane's 10%. Hokit has only 4 UFC fights, so his rates are a small sample and partly blended toward league averages.

Our model leans Gane at 60% because of the gap in Elo (1696 to 1613), the large difference in strikes absorbed, and Gane's reach edge over 13 UFC fights. Hokit's youth (29 against 36.6) and his wrestling keep it from being more lopsided. The market is more confident at 67%, a 7-point gap that is under our threshold for calling a disagreement, so the two views are broadly aligned. Both fighters are coming off the same 154-day layoff, so that doesn't separate them.

Key question: can Gane keep Hokit at range and out of the clinch, or does Hokit close the distance and wrestle his way into a fight on his terms?

Nunes brings the far bigger UFC résumé: 18 fights, 4.46 significant strikes landed per minute against 2.50 absorbed, and real knockdown power at 0.56 per 15 minutes. Harrison's UFC sample is only three fights, so her numbers are heavily blended toward league averages and should be read as a small sample, not a full profile. What stands out is her control time share of 50% and her 1.81 strikes absorbed per minute. Nunes carries a 77% takedown defence, while Harrison's judo-based grappling is the obvious test of it. Both stand 68 inches, but Nunes has a 3-inch reach edge.

Our model leans Harrison at 59%, and the market is close at 63%, so there is no big disagreement. The drivers are Nunes's long layoff of 1,254 days, Harrison's recent form (a +3 streak against Nunes's +2), and Harrison's grappling control. Nunes's Elo of 1697 is higher than Harrison's 1587, so the model is not dismissing her. Age is similar, at 38.5 against 36.4. The model includes layoff and age, but it may underweight how much three-plus years away affects a fighter.

Key question: can Nunes keep the fight standing and use her power, or does Harrison's control take over?

Kevin Holland brings a lanky frame to this one: he stands 75 inches with an 81-inch reach, a full 10 inches of reach over Uros Medic. Both men land at a similar clip, about 4.5 to 4.8 significant strikes a minute, and neither wrestles much. Holland's takedown and control numbers are modest, and Medic's are lower still. The difference is power. Medic scores 2.43 knockdowns per 15 minutes against Holland's 0.34, and he finishes 65% of his fights to Holland's 38%. Holland has been submitted three times and knocked out twice, so he does give chances away.

Our model leans Medic at 57%, driven by three numbers: his sharper striking accuracy (56% to 48%), his knockdown rate, and his four-fight win streak. He is also fresher, with 106 days since his last fight against Holland's 218. Holland's 28 UFC fights give his profile a solid sample, while Medic's 11 are a smaller one. The betting market is close at 61% for Medic, so there is no big disagreement. Holland's age of 34 and longer layoff are in the model, though it may underweight them.

Key question: can Holland use his reach to keep Medic at range, or will Medic's power close the distance?

Bilal Hasan brings a finisher's profile into this one, though it rests on a single UFC fight, so his numbers are a small sample and heavily blended toward league averages. He lands 4.56 significant strikes a minute, absorbs 3.67 and throws knockdowns at a notable 0.83 per 15 minutes. Luis Gurule is the busier, more hittable fighter: 5.58 landed but 6.36 absorbed per minute, with almost no knockdown power and a 7% finish rate. Gurule is also shorter at 65 inches with a 64-inch reach, has been stopped once and submitted once, and offers little on the mat (20% takedown accuracy).

Our model leans Hasan at 77%, mainly because of the gap in strike differential (Hasan is slightly positive, Gurule is well in the red), Hasan's higher Elo (1528 vs 1434) and an eight-year age edge (25 vs 33). Gurule's 169-day layoff and recent losing streak don't help him. The market is higher on Hasan at 86%. A likely reason is that per reports he is 9-0 with eight finishes and a three-time CFFC champion, which the model can't see because it only knows UFC fights.

Key question: can Gurule's volume survive Hasan's power long enough to turn this into a grind, or does Hasan find an early finish?