How Julian Erosa beats JeongYeong Lee: the third-round wrestler

Everybody already knows the short version of Julian Erosa. He comes forward, he throws a lot, and he gets hit. Seven of his nine UFC losses came by knockout, five of those in the first round, and he has been dropped 10 times in 18 UFC fights. He is 37, he is on a two-fight skid, and on Saturday he walks out as the underdog against JeongYeong Lee on the prelims of UFC Fight Night: Moicano vs Nolan at the Meta Apex.

All of that is true. It is also only the first five minutes of the story. The part of Erosa's game that gets skipped is what he does once a fight survives into the third round, and it happens to line up almost perfectly with the one thing Lee has struggled with in the UFC.

So let's take a look at how the upset actually gets built.

The board, briefly

The betting consensus opened with Erosa at +151 and Lee at -183 on 13 books. It drifted further from Erosa at first, out to +154, and has since come back to +142 against -172 with 18 books pricing it. That +142 works out to an implied 41.3 percent.

Our model gives Erosa 51 percent, which is a coin flip. The model's own breakdown is honest about why it is not higher: age is the single largest input pulling against Erosa (he is 37, Lee is 30). The grappling inputs are what pull it back toward him. Hold onto that, because the grappling is where this whole piece lives.

The third-round wrestler

Erosa is not usually described as a wrestler, and over a full fight his numbers do not scream it either: 17 takedowns on 39 attempts across his UFC and Contender Series bouts, 44 percent.

Split those attempts by round and a different fighter shows up. In the eight Erosa fights that reached a third round, he went 5 for 18 on takedowns across the first two rounds. In the third rounds of those same fights he went 10 for 17.

That is not a coincidence of a single night. Look at the fights:

• Against Sean Woodson, Erosa was 0 for 2 on takedowns through ten minutes, then went 3 for 4 in the third round and finished him by submission at 2:44 of it.
• Against Charles Jourdain, 0 for 3 in the first round, nothing in the second, then 2 for 2 in the third with nearly two minutes of control and a submission at 2:56.
• Against Steven Peterson, he won a split decision, and in the third round he landed 62 significant strikes on 92 attempts plus both of his takedown tries.

I'm not going to pretend I have frame counts for this, but the order in the numbers is consistent: two rounds of volume first, then the takedowns start landing. In Erosa's best fights the takedown is the payoff for the striking, not a separate plan.

Lee's control problem

Now put that next to Lee's UFC record, which is 2-2.

Lee has spent 46 minutes and 5 seconds in the Octagon. Opponents have controlled him for 19 minutes and 43 seconds of it. Per 15 minutes, that is about six and a half minutes on his back or against the fence under someone else's control. Opponents have tried 42 takedowns against him; 39 of those came in just two fights, which tells you that the people who decided to wrestle him really committed to it.

The Daniel Santos fight is the clearest picture. Lee had the first round's striking numbers, 35 significant strikes to 33, and Santos went 0 for 5 on takedowns. Then Santos kept shooting. He went 3 for 5 in the second round and 3 for 8 in the third, with close to two and a half minutes of control in each. Lee's output fell from 35 landed to 20 to 13. Santos took the unanimous decision. It is not a one-off shape, either: in all three of Lee's UFC decisions, he landed fewer significant strikes in the third round than in the second.

Even Lee's wins carry the warning. Yizha, in Lee's Road to UFC final, attempted 21 takedowns, landed five and controlled him for 8:19. Lee still got the split decision, because Yizha landed only 24 significant strikes in the whole fight. That says the control did little damage. It says nothing about whether judges will forgive that much bottom time a second time. Against Blake Bilder, Lee conceded 2:23 of control in the first round without Bilder ever completing a takedown.

There is a detail in Lee's UFC numbers that matters here: zero submission attempts in four fights. Whatever he does from underneath, it has not included threatening a finish, which means an opponent pays very little for staying on top of him.

The pace gap

Erosa lands 6.06 significant strikes per minute in the UFC. Lee lands 3.17.

That gap is the engine for everything above. Lee has never landed more than 68 significant strikes in a UFC fight. Erosa has topped 90 four times: 155 against Peterson, 103 against Woodson, 103 against Jourdain, 92 against Hakeem Dawodu. The third-round takedowns in those fights did not come from nowhere. They came after 10 minutes of the other man having to deal with a striker throwing at that rate.

A fair caveat belongs right here, because Erosa's most recent decision went the other way. Against Melquizael Costa in May 2025, he got his takedowns early (3 of 10, 3:34 of control) and still lost every card, because Costa outlanded him 102 to 69. Wrestling without the volume underneath it is just a guy getting up and getting hit. The path only works in the order it worked against Woodson and Jourdain.

The scramble finish

Erosa has 14 submission wins as a pro, more than his 12 knockouts. The most instructive recent one is the fight he looked like he was losing. Ricardo Ramos took him down twice in three tries and held him for 87 seconds of a fight that lasted 2:15. Erosa landed four significant strikes. He also won by submission.

That is the second path, and Lee's numbers invite it: six and a half minutes of bottom time per 15 and no submission threat of his own. Bottom time ends in a scramble sooner or later, and Erosa has already finished a fight in which the other man had both the takedowns and the control. The thin method-of-victory market agrees, for what two books are worth: a submission is the shortest price on any Erosa method, at +360, well ahead of his knockout and decision prices.

What Lee brings: the first five minutes

None of this works if Erosa does not reach the second half of the fight, and the danger is not subtle.

Five of Erosa's seven UFC knockout losses came in round one. The three most recent, against Alex Caceres, Fernando Padilla and Lerryan Douglas, all ended inside 3:33 of the opening round, and Douglas dropped him twice in March. The books lean the same way: under 1.5 rounds sits at -134 on 11 books, and the same thin two-book method market has Lee by knockout at +107.

Lee's case for that finish lives mostly outside the UFC. He has four stoppage wins by strikes as a pro, including a 10-second finish on the Road FC circuit and a 42-second finish on the Road to UFC tournament. In four UFC fights he has scored zero knockdowns. The power is on his record; it has not yet shown up against this level.

One thing cuts clearly his way: he is more than six years younger. The frame is close to a wash (Lee stands 5 feet 10 to Erosa's 6 feet 1, but gives up only an inch of reach), and Erosa is the southpaw against Lee's orthodox stance. Against him: Lee has not fought since May 10, 2025, which is 539 days before Saturday, a booking against Gaston Bolaños in March did not happen, and his last win came in February 2024.

How it plays

Erosa's version of this fight is plain. Survive the first round without trading in the pocket. Make the second round a volume fight, because Lee has never reached 70 significant strikes in a UFC bout and Erosa has gone past 90 four times. Then start shooting in the third, when Lee's numbers say he slows and when Erosa's say he turns into someone else entirely. The submission is there in any scramble along the way.

Lee's version is shorter. He needs one clean exchange in the first five minutes against a man who has been dropped 10 times in the UFC.

Our model calls it close to even. The market calls it a Lee fight. What I'd watch is the clock: if Erosa is still on his feet when the bell ends round one, every number above starts working for him, one round at a time.