UFC 332 predictions: model picks for every Silva vs Wang bout

On Tuesday morning, Roberto Soldic was -309 against Khaos Williams. That price had barely moved since the lines opened, and I said so in Tuesday's odds piece.

It has moved now. The consensus across 18 sportsbooks slid to -298 by early Wednesday, -271 by Wednesday night and -259 this evening. The fans have not budged an inch: of the 27 people who have picked this fight on our pick'em, 25 took Soldic. Our model sits between the two, on Soldic's side at 56.9 percent and a long way short of the market's 69.

UFC 332 is Saturday, October 3, at the Delta Center in Salt Lake City. The early prelims start at 4 p.m. ET (20:00 UTC), the prelims at 6 p.m. ET and the main card at 8 p.m. ET (00:00 UTC Sunday); every region is in the viewing guide. Below is the model's number for all 14 bouts, the fans' split, and my picks, which follow the model most of the way and leave it in two places. Lines are quoted as information only, not betting advice.

The card

The model column is the win probability for the model's pick. The fans column is the pick'em vote, the majority side first.

The model never sees a betting line or a pick'em vote, which is why the disagreements are worth reading.

Soldic: the line moved, the fans did not

Photo: PIXSELL / CC BY-SA 3.0

The line did not drift so much as fall off a step. It eased from -309 on Tuesday morning to -291 by Wednesday afternoon, then dropped 18 cents in three and a half hours that evening and has kept going since. Once the bookmaker margin comes out, the market has Soldic at 69.0 percent, down from 72.2 on Tuesday. Somebody has been backing Williams all week. It was not anyone on our pick'em.

The model's reading is the most interesting of the three. Its biggest reason on this fight runs against Soldic: level of competition, the ONE and KSW résumé measured against the UFC welterweight pool Williams has spent 11 fights in. It also counts Williams' four-inch reach edge, 77 inches to 73. What carries Soldic over the line anyway is Williams' defense, 5.3 significant strikes absorbed a minute with only 41 percent avoided, and a career win rate of 84 percent against 77. Soldic is 21-4 with one no contest, and Saturday is his first fight in 590 days. Williams is 16-5 and lost two straight before knocking out Nikolay Veretennikov in May.

So the model, the market and the fans all have the same winner and three different levels of nerve. I am with the 25.

Prediction: Roberto Soldic by KO/TKO

The heavy favorites, and the one gap left

Four men go in as heavy favorites, and the model picks all four. On three of them it sits within about six points of the market. Anthony Wint is the exception.

Anthony Wint vs Lucas Armand. The widest gap on the card, 23.6 points. Wint opened at -245, was -463 on Tuesday and is -547 now. Every one of the model's top four reasons favors him: level of competition, his 2-0 record in UFC-run cages against Armand's none, a layoff of only 42 days, and a pro record of 8-0 against 6-0. The trouble is what sits under them: everything Wint has shown in a UFC-run cage is two fights eleven days apart in August, and the model treats the first two of those reasons as thin samples. Armand is three inches taller at 6-foot-2 and their reach is identical, 78 inches each. Two unbeaten heavyweights, one with two fights of that tape and one with none, is where a model gets nervous and a market does not. Prediction: Anthony Wint by KO/TKO

Damian Pinas vs Andrey Pulyaev. A 6.3-point gap, and the line has gone further out since Tuesday, -493 to -562. The model's biggest reason is level of competition, and it likes Pinas' age, 24 against 29. The one reason pulling the other way is striking defense: Pinas absorbs 5.3 significant strikes a minute and avoids 45 percent. He also has two first-round knockouts in the UFC, and Pulyaev has lost two straight, one of them to Ateba Gautier. Prediction: Damian Pinas by KO/TKO in round one

Jacobe Smith vs Bruce Whitehead. Smith is 12-0 and Whitehead, 8-2, is making his UFC debut. Every reason the model gives points at Smith, led by level of competition and his unbeaten UFC run. It lands at 85.9 percent, less than a point short of a market that has had him near -990 all week. Prediction: Jacobe Smith by KO/TKO

Marcus McGhee vs Anthony Romero. Anthony Romero, "The Bully," 7-2 with one no contest, a 5-foot-5 southpaw making his UFC debut, replaced Benardo Sopaj on September 30, and McGhee moves up from bantamweight, where he is ranked 15th, to featherweight to meet him. The first price was -365 on Wednesday evening; it is -498 now. The model has McGhee at 73.4. Its biggest reason is level of competition, then experience, 11 pro wins to 7 and six UFC fights to none, then reach, 69 inches to 66. The fans cannot decide at all: 2-2. McGhee's last three fights all went the distance. Prediction: Marcus McGhee by decision

Where the fans and the model split

Natalia Silva vs Wang Cong. The pick'em was 18-18 when I wrote why the model calls it a coin flip on Tuesday; it is now 25-24 for Wang. The model leans the other way by the width of a hair, Silva at 50.3 percent. Its reasons for Silva are age, 29 against 34, and her 8-0 UFC record; its reasons for Wang are her unbeaten record before the UFC and her striking volume. The market has Silva at -210, about 65 percent. I said on Tuesday that four ranked wins, Alexa Grasso and Rose Namajunas among them, are worth more than a model that does not rank opponents knows. I still think so. Prediction: Natalia Silva by decision

King Green vs Esteban Ribovics. The fans' biggest act of defiance on the main card. Green turned 40 on September 9 and has finished three men this year, the last of them Terrance McKinney at 4:59 of the first round in July. The model has Ribovics at 60.2 percent, and its two biggest reasons are level of competition and age. Its main reason for Green is striking: he absorbs 3.5 significant strikes a minute and avoids 62 percent, where Ribovics absorbs 5.7 and avoids 54. The line has been coming toward Green since it opened on September 24, from +260 to +192. I am going with the 12. Prediction: King Green by KO/TKO

Imanol Rodriguez vs Alden Coria. Rodriguez is 7-0 and the fans have him 9-3; the market agrees at -151. The model takes Coria (13-3 with one no contest, 3-0 in the UFC) at 54.7 percent, and its biggest reason is defense: Coria avoids 70 percent of significant strikes, Rodriguez 58. That is a real number, but the model is working from a single UFC fight for Rodriguez, the second-round stoppage of Kevin Borjas in February. Coria's last two went to the cards. Prediction: Imanol Rodriguez by KO/TKO

Johnny Walker vs Mick Parkin. Fans 9-5 Walker, model 59.4 Parkin, and the line has eased toward Parkin since Tuesday, -142 to -133 on Walker. The model's biggest reasons are age, 34 against 31, and Parkin's unbeaten record before the UFC; Parkin's 560 days out is the main thing it holds against him. I went through Walker's slide with the early prelims. Prediction: Mick Parkin by decision

Marvin Vettori vs Ismail Naurdiev. The model and the market have the same winner, Naurdiev at 58.3 and 55.4 percent. The fans have Vettori 9-3 anyway. He has lost four straight, and the model's biggest reason is the damage: he absorbs 4.6 significant strikes a minute to Naurdiev's 1.9. Prediction: Ismail Naurdiev by decision

Further down, eight people have picked Rafael Dos Anjos vs Alexander Hernandez and all eight took Hernandez. The model agrees on the side but gives the former champion 41.1 percent, the most generous number anybody is putting on him, with his 49 pro fights as its biggest reason; the market has him at +222, about 30 percent, and drifting. Dos Anjos turns 42 on October 26, has lost three straight and has not fought in 707 days. A 41 percent fighter wins two fights in five, and somebody will have to explain the zero. Prediction: Alexander Hernandez by KO/TKO

The rest

Ateba Gautier vs Roman Kopylov. The model has Gautier at 69.9 percent, led by age (24 against 34), the unbeaten UFC run and reach, 81 inches to 75. The market has walked further out, -208 on Tuesday to -243 now. I made the case for Gautier on Wednesday. Prediction: Ateba Gautier by KO/TKO

Deiveson Figueiredo vs Payton Talbott. Figueiredo has drifted from +442 on Tuesday to +502; the model gives him 20.3 percent, and 8 of 23 fans are with him. The case for that price is its own piece. Prediction: Payton Talbott by decision

Court McGee vs Eric Nolan. The model gives McGee 41.2 percent, with age, 41.8 against 29.1, as its biggest reason. Nolan is 0-2 in the UFC. Prediction: Eric Nolan by decision

Saturday's ledger

Count it up and I have the model's side in 12 of 14 fights, leaving it only for Green and Rodriguez. I am with the fans' majority in 10, against it in three, and McGhee vs Romero is tied. On Wint, the model and the market agree on who wins and disagree by 23.6 points on how sure to be. If he wins, nobody will remember the 57.1. If he loses, the model will have been the only voice in the building that half saw it coming.

Every number here keeps updating on the event page, and the UFC 332 pick'em is open until the first bell. The 25 who took Soldic are waiting for company.

Correction, October 2: an earlier version of this article quoted model probabilities affected by a data issue that has since been fixed. The figures above are the corrected ones.