Amanda Lemos is No. 8, and a 2-to-1 underdog to No. 12

Amanda Lemos has not fought since March 14. In the twenty weeks since, the official rankings have listed her eighth, then seventh, then sixth, then seventh, then eighth again. She did not win a fight to climb. She did not lose one to fall. The board moved around her while she stood still.

On Saturday she fights Alexia Thainara in the fourth bout of a five-fight main card at the Meta Apex, and Thainara is ranked four spots below her at No. 12. Twenty-five sportsbooks currently price Lemos at +206 and Thainara at -250. The lower-ranked fighter is a better than 2-to-1 favorite, and the line has drifted further in her direction almost every day since the full market opened.

The easy read is that the books are wrong, or that this is an upset waiting to happen. So I went and checked. When the market and the UFC rankings disagree about which fighter is better, how often does each one turn out to be right?

It is not close, and it has not been close for years.

603 fights where the market outranked the rankings

I pulled every completed UFC bout since the rankings began in February 2013 in which both fighters held a divisional ranking on fight day and closing odds exist in our database. That is 1,570 fights.

In 603 of them, 38.4 percent, the betting market made the lower-ranked fighter the favorite. Those favorites went 382-221. The market wins that argument 63.3 percent of the time.

Put the other way: in the 967 fights where the market and the rankings agreed on the pecking order, the favorite won 69.8 percent of the time. So the market's hit rate barely suffers when it contradicts the official board. It gives up about six percentage points of accuracy and keeps winning roughly two fights in three. The rankings, on the losing side of that split, are wrong nearly two times out of three whenever the money disagrees with them.

The disagreement is also getting more common, and quickly.

In 2013 the market took the lower-ranked fighter in about one ranked matchup in five. In 2025 it did so in more than half of them. The two systems used to broadly agree about who was better. Now they are closer to a coin flip on the question, and one of the two keeps being right.

Some of that gap is structural rather than scandalous. Rankings are voted by a media panel that updates weekly, and a ranking is a record of what a fighter has already done. Odds are a forecast, they move on injuries, camps, styles, weight cuts and money, and they are free to price a 3-0 newcomer above a veteran with a better résumé. Nobody should expect the two to match.

But a ranking that lags this badly stops being a lagging indicator and starts being a misleading one, because it is the number the promotion uses to sell the fight, and the number a fan uses to decide whether the fight matters.

The four-spot version, and the division it keeps happening in

Narrow it to matchups shaped like Saturday's, where the favorite is exactly four spots below the underdog, and there are 73 of them. The favorite won 45, or 61.6 percent.

Narrow it instead to women's strawweight, and it gets worse for the board. There have been 38 ranked strawweight fights where the market preferred the lower-ranked woman. The favorites went 28-10, a 73.7 percent hit rate, the strongest of any division-sized sample I looked at.

Three weeks ago the division produced the purest version of it. On July 18, No. 8 Tabatha Ricci was a +516 underdog against No. 11 Fatima Kline, a line so lopsided it implied the rankings had the two women in the wrong order by a distance. Kline won a unanimous decision.

And on March 14, in the last fight Lemos had, she was the No. 5 and a +198 underdog to No. 8 Gillian Robertson. Robertson won a unanimous decision. Lemos has now been on the wrong side of this exact trade twice in five months, and the second time the market has moved her from a two-spot flip to a four-spot one.

What the market is actually reading

None of this means the odds are magic. It means the odds are reading the tape, and the tape here is not subtle.

Since the start of 2023, 43 UFC strawweights have fought at least three times. Of those 43, the fighter who has surrendered the most control time per fight is Amanda Lemos, at 8 minutes 35 seconds a night. Over six fights she has spent 3,089 seconds on the wrong end of control, against 963 seconds on the right end.

The fighter who has surrendered the least, of the same 43, is Alexia Thainara. In three UFC fights she has given up 11 seconds of control. Not eleven minutes. Eleven seconds.

They are ranked first and forty-third on the same list, in the same weight class, and they fight on Saturday.

Look at how Lemos has lost lately and the mechanism is the same each time. Robertson took her down three times and held her for 9 minutes 17 seconds. Tatiana Suarez took her down three times and held her for 9 minutes 34 seconds. Virna Jandiroba submitted her in the second round after 6 minutes 27 seconds of control. Lemos is 1-3 in her last four, and all three losses were grappling losses in one form or another.

Now look at what Thainara does. In her three UFC wins she has landed 12 takedowns on 23 attempts and accumulated 23 minutes of control. Her opponents are a combined 0-for-2 on takedown attempts against her. Against Bruna Brasil in March she landed six takedowns and controlled the fight for 9 minutes 34 seconds while conceding none.

That is not a market ignoring the rankings out of ignorance. That is a market noticing that a 39-year-old whose last three defeats came from being held down is fighting a 30-year-old whose entire UFC identity is holding people down, and pricing it accordingly.

The case the number leaves out

Lemos deserves better than a spreadsheet, and the honest version of this argument has to include the parts that hurt it.

She is 15-6-1, and five of those six losses came to Jessica Andrade, Zhang Weili, Jandiroba, Suarez and Robertson. The sixth was a bantamweight fight against Leslie Smith in 2017, a weight class and nine years away from the fighter she is now. She has, in other words, essentially only ever been beaten by the top of her division. She owns two Fight of the Night bonuses and a Performance of the Night for submitting Michelle Waterson-Gomez. She stopped Marina Rodriguez 54 seconds into the third round. She out-struck Mackenzie Dern 41 to 21 and dropped her in a Fight of the Night. When she gets to stay standing she is one of the harder punchers in the weight class, and against Robertson in March she still landed 42 significant strikes to 12 while losing, which tells you that fight was decided by where it happened, not by who was better at fighting.

Thainara, meanwhile, has beaten Brasil, Loma Lookboonmee and Molly McCann. That is a real 3-0 start and it is also not a top-eight résumé. The rankings are not being absurd by having her at 12. They are being slow.

And 63.3 percent is a two-in-three proposition, not a certainty. Roughly one time in three, the fighter the board rates higher goes out and proves the board right. That is what Lemos has to do on Saturday, and it is a live enough possibility that anyone telling you this is a formality is selling something.

The part that should bother the promotion

Here is what I keep coming back to, though.

Between March 17 and August 4, Amanda Lemos went from eighth to seventh to sixth to seventh to eighth without setting foot in the cage. Five different official assessments of her standing, generated by other people's results and other people's votes, in a stretch where the only new information about Amanda Lemos was that there was no new information about Amanda Lemos.

A ranking system that moves a fighter two spots in either direction for doing nothing, and that the betting market can beat 63 percent of the time when they disagree, is not really measuring the thing it claims to measure. It is a weekly opinion poll with a number attached, and the number is what gets printed on the graphic.

The market has Saturday's fight at roughly 71 percent for the woman ranked four spots lower. The board has it the other way. One of them is going to be right, and we have 603 previous readings on which one usually is.

Rankings, odds, control time and fight histories in this piece come from UFCalendar's own database, current as of Wednesday morning. You can follow the full card, the live odds and the official rankings as the week goes.