UFC 334 predictions

Radtke and Thompson meet as a clear contrast in styles. Thompson is the volume striker, landing 4.01 significant strikes a minute to Radtke's 2.66, and he has a 3-inch height edge and 3 inches of reach. Radtke is the more accurate striker (52% to 46%) and brings the only real grappling threat, with 1.85 takedowns per 15 minutes against Thompson's 0.28 and 47% control time against 6%. Thompson's takedown defence is a modest 63%. Both men have been knocked out twice, and both absorb strikes at a similar rate on paper, though Thompson's 3.01 per minute comes against a much heavier output. Radtke's finish rate is 39% to Thompson's 27%.

Our model likes Radtke at 63%, mainly because of his wrestling and control advantage, his better accuracy, and his recent form (a +2 streak against Thompson's -3). Thompson is 43.8 and has been out 490 days, which the model treats as a mild negative. The market disagrees by 18 points and prefers Thompson at 55%. The likeliest reason is that the model only sees UFC numbers and may undervalue Thompson's 22-fight experience and the quality of opposition behind his volume, along with anything about his layoff or camp that we can't see.

Key question: can Radtke get Thompson to the mat often enough to blunt the volume striking?

Chris Duncan and Drew Dober are almost mirror images on the feet. Both land around 4.4 significant strikes a minute, both absorb a little over 4, and both defend 53% of incoming strikes. Duncan is the more accurate striker at 46% against 41%, while Dober carries more knockdown power at 0.65 per 15 minutes to Duncan's 0.40. The grappling gap is clearer. Duncan attempts 2.5 takedowns per 15 minutes and holds control for 22% of fight time, while Dober attempts just 0.65. Both defend takedowns at 59%. Duncan is also younger by nearly five years, and he has a slight edge in height and reach.

Our model leans Duncan at 67% because of three things: his accuracy edge on the feet, his wrestling volume, and his age and rating advantage (1580 Elo to 1567, with a 70% smoothed win rate against 57%). Dober's durability is a concern for him, since he has been knocked out three times and submitted four. The market is closer at 56-44, an 11-point gap. The likeliest reason is that the model leans heavily on Duncan's efficiency numbers and may undervalue Dober's 26 fights of experience and his current two-fight win streak, while Duncan comes in on a loss. Dober's finishing power also gives him a live path that averages can understate.

Key question: can Duncan use his takedowns to avoid Dober's power, or does the fight stay standing, where Dober's knockdown threat matters most?

Jim Miller brings a veteran's all-around game, while Terrance McKinney is the far busier and more explosive fighter. McKinney lands 6.54 significant strikes a minute to Miller's 2.93, and he hurts people, with 0.99 knockdowns per 15 minutes against Miller's 0.20. His defence is a weakness, though: he absorbs 4.32 a minute, has been knocked out four times, and has a 44% striking defence. Miller is the steadier defender at 57% but gives up more than he lands. On the mat, McKinney shoots more often and holds up better (69% takedown defence to Miller's 49%), while Miller is a constant submission threat at 1.68 attempts per 15 minutes. McKinney also has a 3-inch reach edge and is 11 years younger.

Our model gives McKinney 60%, driven by his volume gap, his knockdown power and his better takedown defence, which should make it harder for Miller to drag him into grappling exchanges. His 1532 Elo is below Miller's 1616, which keeps the lean modest. The market is more confident at 70%, a 10-point gap from us. The likely reason is age and durability: Miller is 43, and the market may weigh that more heavily than our UFC-only numbers do.

Key question: can Miller survive McKinney's early pressure long enough to find a submission, or does McKinney finish him first?

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 he is accurate at 58%. But his defence is shaky: he only avoids 45% of strikes, and he has been knocked out once. Johnson is the better defender at 60% and gives up fewer shots. The bigger gap is on the mat. Johnson attempts 1.70 takedowns per 15 minutes, stops 72% of attempts against him, and holds control for 26% of his fight time, against Almeida's 8%. Johnson also finishes half his fights and is nine years younger.

Our model likes Johnson at 77% because of that wrestling and control gap, his better striking defence, and his higher Elo (1569 to 1511) with a three-fight win streak. Almeida is 38 and has lost his last two. Caveat: Johnson has only three UFC fights, so his numbers are a small sample and are blended toward league averages. The market is much closer, at 58% for Johnson, a 19-point gap. The likeliest reason is that our model only sees UFC fights, so it can't weigh Almeida's size, his striking accuracy, or anything from outside the UFC that might make the market trust Johnson less.

Key question: can Johnson get Almeida to the mat and hold him there before Almeida's accurate, longer striking lands something big?

Bia Mesquita brings a grappling-leaning game with real output on the feet. She lands 3.86 significant strikes a minute while absorbing 2.37, and she hunts takedowns at 2.34 per 15 minutes with 1.90 submission attempts in the same span. Macy Chiasson is the bigger frame at 71 inches tall with a 72-inch reach, and she is far more experienced with 14 UFC fights. But she lands 3.22 a minute, absorbs 2.81, and defends only 46% of strikes against Mesquita's 63%. Chiasson's takedown defence of 63% is decent, and she has been finished once by KO and once by submission.

The model likes Mesquita at 65% mainly because of the striking gap: better accuracy (52% to 46%), better defence, and a clear edge in differential per minute. Her 70% finish rate and submission volume against Chiasson's 34% finish rate add a second route to victory. Chiasson is also on a three-fight losing streak. The market is almost identical at 66%, so there is no real disagreement to explain. One caution: Mesquita has only 3 UFC fights, so her rates are a small sample and partly blended toward league averages.

Key question: can Chiasson use her length to keep Mesquita at range and stop the takedown and submission game from taking over?

Gordon brings the higher output, landing 5.72 significant strikes a minute to Costa's 4.17, and he mixes in about two takedowns per 15 minutes with decent defence at 60%. Costa is the more dangerous finisher: a 38% finish rate against Gordon's 18%, plus 0.80 submission attempts per 15 minutes, which is a real threat on the mat. Costa also has the physical edge, with a 3-inch reach advantage and about eight years of age on a 38-year-old opponent. Both men are coming off losses.

Our model likes Costa at 72% mainly because of his cleaner defensive profile and his experience-adjusted rating. He absorbs only 3.20 strikes a minute against Gordon's 4.50, and his Elo is 1585 to Gordon's 1525. Gordon's durability is a worry too, since he has been knocked out four times and submitted twice in 17 UFC fights. The market is a bit more confident at 78%, but the gap is only six points, so the two views are broadly aligned. Gordon's volume and wrestling give him a path if he can keep the pace high.

Key question: can Gordon's volume and takedowns outwork Costa, or does Costa's finishing threat and longer frame end it first?