UFC Fight Night: Nurmagomedov vs Song — the promotion’s own billing for this card
Song Yadong is 12-4-1 across 17 UFC appearances and has alternated results over his last five, beating Chris Gutierrez, losing to Petr Yan, beating Henry Cejudo, losing to Sean O'Malley and then submitting Deiveson Figueiredo in the second round in May. Umar Nurmagomedov is 8-1 in nine UFC fights and every one of his last five went to the judges, including a five-round unanimous-decision loss to Merab Dvalishvili in the January 2025 bantamweight title bout. Both men carry a recent win over Figueiredo, Song's by submission and Nurmagomedov's on the cards. The model has Nurmagomedov LOCK.
LOCK is the top tier and it is unusual to see it on a five-rounder, where the extra 10 minutes normally widen the range of outcomes rather than narrow it. Nurmagomedov's path is the one his record already describes: win the exchanges, hold position and take the scorecards, which is how all five of his most recent UFC fights were decided. Song's path is the fight he had in May, when he found the finish instead of the decision, and he needs the scramble rather than the long, managed round. The tell is that the model still calls decision the likeliest method, so it reads this as a fight Nurmagomedov wins across 25 minutes rather than takes away early.
Aoriqileng is 4-5 through 10 UFC fights, a run that swings between a first-round KO of Cody Gibson last October and a second-round KO loss to Cody Haddon in May. Kai Asakura is 1-2 in the UFC, both losses by submission, to Alexandre Pantoja in a flyweight title bout and to Tim Elliott, with a first-round KO of Cameron Smotherman in May at bantamweight. The model leans Asakura without much conviction behind it.
Both men fought on the same May card and walked off it going opposite directions, and that is most of what the model is working with. Asakura's path is the Smotherman fight repeated, a fast finish before anything settles. Aoriqileng's path is the grappling that ended both of Asakura's UFC losses, or the volume to take three rounds to the judges. Decision is the single likeliest outcome, though five of the eight most recent results between them ended inside the distance.
Levi Rodrigues Jr. is 0-1 in the UFC, stopped by Felipe Franco in the second round on July 18 in a light heavyweight bout, and he is back about six weeks later. Liu Ce makes his UFC debut.
The read is soft and the material behind it is thin: one UFC fight between the two men, and it ended in a knockout. Liu's path is the one the model's most likely method describes, landing something that finishes it, because nothing on record says how he handles a full three rounds in the UFC. Rodrigues owns the only octagon experience here, and his path is using it while making the July stoppage look like an outlier. The turnaround is the part the record can't speak to, and it may be the most important fact in the fight.
Both men make their UFC debuts. Bilal Hasan and Nilson Rojas have no UFC bout behind them, which leaves nothing in the recent-form column for either side. The model has Hasan HIGH and calls KO/TKO the most likely method.
A HIGH tier resting on no UFC history on either side deserves more scepticism than the label alone suggests. The confidence number is doing work that no UFC form is backing up. Hasan's path, per the model, is the finish, which is unusual on a card where decision is the top method almost everywhere else. Rojas's path is the same as Hasan's, that a debut can look nothing like the projection, and the honest range here is wider than the tier implies.
Andre Lima is 4-1 over five UFC fights, a run built on a disqualification win over Igor Severino, a split decision over Mitch Raposo, a unanimous decision over Felipe dos Santos and a third-round submission of Daniel Barez. He lost a unanimous decision to Kevin Borjas in June and returns 10 weeks later. Namsrai Batbayar has no prior UFC fight in our data, so there is no octagon record to read on him at all. The model has Lima at its highest tier.
The highest tier is the strongest read we publish, and this one rests on an information gap as much as on form: five UFC results on one side and none on the other. Lima's path is the one he walked four times before June, staying in the exchanges long enough to bank rounds and taking the submission if it appears, as it did against Barez. Batbayar's path is the one every debutant has, which is that a first octagon appearance can look nothing like the projection, and there is nothing on file to say which way it breaks. The tell is that the model still calls decision the likeliest method, so its strongest read here expects Lima to earn 15 minutes rather than end them.
Kevin Borjas is 2-4 through six UFC flyweight bouts, and his last five split cleanly by method: both wins came by unanimous decision, both stoppage losses came by knockout in the second round. Rei Tsuruya is 2-1 in three UFC appearances, beaten only on the scorecards, by Joshua Van in March 2025, and coming off a first-round submission of Luis Gurule on May 30 in a bantamweight bout. Borjas has the bigger sample; Tsuruya has the cleaner one. The model has Tsuruya at 83.2%, a LOCK and the highest number on the card.
An 83.2% LOCK is the strongest read the model has for Saturday, and the most likely method is a decision, which puts the confidence on who controls 15 minutes rather than on who lands the biggest shot. Tsuruya's path is the one his record already sketches: bank rounds, and take the finish if it appears the way it did in May. Borjas gets there by turning this into the fight he won against Ronaldo Rodriguez in March 2025 and Andre Lima in June, a full three rounds he can win on the cards, though the same stretch shows him stopped twice inside two rounds. The tell is that the model is weighting Tsuruya's short, clean record more heavily than the twice as many fights on the other side.
Sean Woodson is 7-2-1 in the UFC and won four in a row, including a first-round KO of Fernando Padilla, before Dan Ige stopped him in the third in April 2025. That is his last fight in our data, a layoff of more than a year. Jack Jenkins is 4-2 and has alternated results across his last five, most recently a unanimous decision over Ramon Taveras last September. The model leans Woodson, and not by much.
The lean rests on the better record and the longer win run rather than on anything recent. Woodson's path is the three-round decision that most of his UFC work is made of, taking rounds off a fighter whose own results keep flipping. Jenkins' path is the time Woodson has spent away, plus the fact that Woodson's most recent appearance ended in a knockout loss. The tell is that the model is barely discounting the inactivity, reading a 10-fight UFC sample against Jenkins' six.
Francesco Nuzzi has no prior UFC fight in our data, so there is nothing on his octagon record to read. Xiao Long is 1-3 through four UFC fights, with a third-round KO/TKO of Quang Le in November 2024 as the win and three losses since and before it, a split decision to ChangHo Lee and unanimous decisions to SuYoung You and Cody Garbrandt in March. The model splits this one almost exactly down the middle.
This is the closest read on the card and the model is barely off a coin flip, which is what happens when a debutant with no data meets an opponent whose data says competitive but short. Nuzzi's path is that a first UFC appearance carries no established ceiling and nothing on file to game-plan against. Xiao Long's path is the Quang Le fight, the one time in four the power landed and ended it, and he has the only octagon minutes between them. The tell is that the model still calls decision the likeliest method despite that finish, so it is reading a full 15 minutes with neither man clearly favored to close it.
Julia Polastri is 2-3 through five UFC fights and her results have alternated the whole way, a split-decision win over Cory McKenna, a third-round KO/TKO of Karolina Kowalkiewicz last October, and unanimous-decision losses to Josefine Knutsson, Loopy Godinez and, in April, Talita Alencar. Xiong Jingnan has one fight in our UFC data, a unanimous-decision loss to Angela Hill in May. That is the whole file on this matchup. The model leans Polastri and does not lean far.
This is close to even and the copy should say so rather than dress it up. Polastri's path is the one that has worked twice, keeping the fight competitive on volume and getting to the judges, with the Kowalkiewicz finish as evidence the power is there if the opening comes. Xiong's path is straightforward on paper because the only UFC data point against her is a decision she lost, and a fighter with one octagon result carries the wider range of outcomes. What the model is weighting is thin on both sides, which is why the lean is small.