How Andre Fili wins: the takedowns that beat the strike count
Andre Fili vs Kai Kamaka III at UFC Fight Night Allen vs Duncan: how Fili's takedowns and control win split decisions he loses on the strike count.
- Author
- Mate Bersenadze
- Published
- Oct 9, 2026
- Read time
- 5 min read
- Tags
- UFCUFC Fight Night 2026-10-11Andre FiliKai Kamaka IIIDataPredictions
There is a version of Andre Fili that lives in the highlight reels: the long featherweight with the hard right hand, a Team Alpha Male product who has knocked out Jeremy Larsen, Gabriel Benitez, Sheymon Moraes and Lucas Almeida in the UFC. That Fili is real. He is just not the one who wins most of his fights.
Saturday at the Meta Apex, Fili (25-14, 1 NC) meets Kai Kamaka III (18-8-1) on the main card of UFC Fight Night: Allen vs Duncan. Our model gives Fili 53.1 percent. The sportsbook consensus has him at -146, about 57 percent with the margin out, after opening at -154 on October 1 and dipping to -138 two days later. Seven of the ten people on our pick'em took him.
I think the case for Fili is not about the right hand at all. It is about what happens on the scorecards when he loses the striking.
Winning fights you are losing
Fili has had 27 fights in the UFC since 2013 and is 13-13 with a no contest. Nine of the 13 wins went to the judges, and five of those nine were split decisions. Look at the significant-strike counts in four of the five splits:
| Fight | Fili landed | Opponent landed | Fili takedowns | Fili control |
|---|---|---|---|---|
| Dennis Bermudez, 2018 | 80 | 103 | 4 of 12 | 0:50 |
| Charles Jourdain, 2020 | 44 | 62 | 5 of 6 | 1:46 |
| Bill Algeo, 2022 | 59 | 81 | 1 of 3 | 4:18 |
| Christian Rodriguez, 2025 | 64 | 83 | 3 of 5 | 1:42 |
He lost the strike count in all four, by 18 to 23 strikes each time, and he won all four. The column that explains it is the one to the right. Judges score effective striking first, but a takedown is a moment a judge can see and remember, and control time is a minute of the other man not doing whatever he was winning with.
The pattern holds across the whole UFC career. In his nine decision wins Fili landed 26 takedowns, almost three a fight, and averaged 2 minutes and 52 seconds of control. In his six decision losses he landed nine takedowns, one and a half a fight, and averaged 1 minute and 39 seconds of control. The strike counts in those six losses run Calvin Kattar 75 to 73 (with no takedowns at all from Fili), Michael Johnson 65 to 37, Sodiq Yusuff 73 to 49, Bryce Mitchell 46 to 27, Nathaniel Wood 84 to 58 and Jose Delgado 94 to 62.
So the first rule of Fili fights is simple. When the strike gap is around 20 and he is the one shooting, he has usually got the split. When the gap has been 30 or more, the takedowns have not rescued him.
Why Kamaka is the right kind of problem
Kai Kamaka III is 31 and wrestled in college at Midland University in Nebraska, and by his own account his first kickboxing fight came at 8 years old. In the UFC he fights at a pace that wins strike counts. He landed 114 significant strikes on Tony Kelley in his 2020 debut, a Fight of the Night, and 105 on Dakota Hope in April. His career UFC rate is 5.49 significant strikes landed a minute.
The detail that matters for Saturday is where they land. Of Kamaka's 398 significant strikes in six UFC fights, 164 went to the body, 41 percent. Against Kelley it was 64 of 114; against Hope, 47 of 105. Body work does not show up in a highlight, but it shows up in the count, and it takes the legs out of a man who wants to shoot in the third round.
The other half of Kamaka's record is what comes back. He absorbs 6.03 significant strikes a minute in the UFC, the most of anyone in this bout by a distance, and he avoids 55 percent. Kelley landed 114 on him and Hope 106. Luke Riley knocked him out in 3:03 at UFC 329 in July, 91 days before Saturday. His UFC takedown defense is 55 percent too: opponents have landed nine of 20 attempts, and Jonathan Pearce landed five of seven and kept him down for 3 minutes and 26 seconds.
The shot off the exchange
None of this asks Fili to become a wrestler. His UFC takedown average is 2.24 per 15 minutes at 46 percent accuracy, and he defends 72 percent of the shots taken at him. What the record says is that he does his best scoring work as a punctuation mark. Kamaka will want to win the exchanges with volume and the body. A level change at the end of one of those exchanges, when Kamaka is set and throwing, is the cleanest entry Fili has, and a takedown that ends an exchange Fili was losing turns the judges' memory of that minute around.
Fili also has the frame for it. He stands 5-foot-11 with a 74-inch reach against Kamaka's 5-foot-7 and 69, and the reach is one of the model's reasons for him. A longer man who keeps the jab in Kamaka's face makes Kamaka come forward to find the body, and a man coming forward is a man whose hips are close enough to reach.
The danger is the Delgado fight in March. Fili landed four of six takedowns there and still lost a split, because he was out-landed 94 to 62. The fight after that, Vinicius Oliveira landed 105 on him and stopped him with four seconds left in the second round. Opponents have landed 83, 94 and 105 significant strikes on Fili in his last three fights. That number is the one moving in the wrong direction, and at 36 it does not usually turn around.
What the model sees
The model's two biggest reasons go against Fili: level of competition and age, 36 to 31. It comes back to him on his record before the UFC, 92 percent wins to Kamaka's 67, on 25 career wins to 14, and on the reach. Net, it lands a fraction above even, a little short of the market.
That feels about right to me, and I am on the same side. If the strike count stays within 20 and Fili is shooting in the second half of rounds, it goes to the cards and his record on the cards is better than anyone gives him credit for. If Kamaka's volume opens the gap to 30, the takedowns will not save him. They did not in March.
Prediction: Andre Fili by split decision
The full card and start times are on the event page, and the pick'em is open at the predict page until the first bell.