PFL Tampa Odds & Prediction: The Board Posted on 11 of 12 — and Listed the Co-Main Twice

FanDuel posted the PFL Tampa board, and it did two things worth reporting. It priced eleven of the twelve bouts on Cris Cyborg's retirement card — and it listed the co-main event twice, at two different prices, on a card that only has one of them. Here's the read on the PFL Tampa card. New to reading prices as probabilities? Start with our PFL prediction markets guide.

> Board out, card changed, one number unresolved. FanDuel priced 11 of 12 bouts. Gustavo Oliveira withdrew this morning and Nkosi Ndebele (13-3) stepped in against Marcirley Alves on two days' notice — the book had already repriced it before the card pages caught up. The co-main, Rabadanov vs. Reeder, appears on FanDuel's list at both -4500/+1200 and -2500/+900; we can't resolve which is live, so it reads "Market opens soon" on the card page rather than carrying a guess. Kalshi and Polymarket still have no PFL contracts listed (re-verified August 20).

Every bout on the card, one tale of the tape at a time — records, ranks and the labeled reference line, with the co-main held at "Market opens soon" until the board resolves. The full card lives on the PFL Tampa card page, and every event we cover is on the MMA hub.

The whole board, de-vigged

A sportsbook moneyline isn't a probability — it's a probability plus the house's cut. Strip the vig and the two sides sum to 100, which is the same language a prediction-market contract already speaks. Here is FanDuel's entire Tampa board converted into what it's actually claiming:

Two things jump off that table before you get to any individual fight. The board has three ranked fighters as underdogs — Magomedov, Schulte and Zendeli are all priced to lose to unranked opposition. And there is exactly one bout inside 55/45 on the entire card. This is a chalk board with a soft spot in it, and the soft spot isn't where you'd expect.

The number the book couldn't settle on

Start with the anomaly, because we've never seen it on a PFL board before and neither, we suspect, has anyone else covering this card.

FanDuel's PFL list carried Gadzhi Rabadanov vs. Tracy Reeder twice. Once at -4500/+1200, once at -2500/+900. Same two fighters, same weight class, same night, two prices roughly 800 basis points apart on the favorite. PFL Tampa has twelve bouts; the list had thirteen rows.

We could have picked one. The later slot sits in the co-main position and would have been the defensible guess. We didn't, and here's why: a price you can't source is not a price, it's a decoration. The rule on this site is that a de-vigged, attributed book line is real and a filled-in table cell is not, and that rule doesn't get suspended because a blank looks bad. The co-main reads "Market opens soon" on the card page until the board resolves, at which point it fills in on its own.

What it probably means is more interesting than the number anyway. A duplicated listing at two prices is what a pricing system looks like mid-argument with itself, and the timing fits: this card took a main-card fighter change on the Thursday of fight week. Whatever the true number is, both versions agree on the shape — Rabadanov, the #2-ranked lightweight at 27-5-2, is a prohibitive favorite over a 12-5 PFL debutant, somewhere north of 96%. There is no trade there in either direction. The interesting part is watching a book visibly not know its own mind on a card it just had to reprice.

Cyborg at 74%, and the case the board is making against her

-370 / +265. Seventy-four percent.

For a farewell fight, that's a restrained number. Cyborg is 29-2, hasn't lost since Amanda Nunes knocked her out in December 2018, holds titles in four organizations before this one, and is defending rather than vacating the belt on her way out. The lazy version of this line is -700 and a coronation. FanDuel is instead giving Ketlen Vieira better than one chance in four of being the woman who ends it.

Here's what's inside that 26%, because it isn't nothing:

• Vieira was cut off a win. The UFC removed her from the roster on May 22, 2026 — six days after she beat Jacqueline Cavalcanti by unanimous decision. She left with a 10-5 UFC record and wins over Miesha Tate, Holly Holm and Sara McMann. That's three former champions or title challengers, and it is a better list of scalps than most PFL signings arrive with.
• She's already fought at featherweight, and won. Her May 2025 bout with Macy Chiasson moved to featherweight on weigh-in day and she took a unanimous decision. This isn't a bantamweight guessing at a new division for the first time.
• The activity gap runs the wrong way for the champion. Vieira has fought three times in the last sixteen months, most recently in May. Cyborg has fought twice in the last twenty-three — Larissa Pacheco in October 2024, Sara Collins in December 2025. At 41, on a retirement night, in front of a home-country crowd, eight months of inactivity is a real input, not a talking point.

And the case for the 74%: nobody in this weight class has solved Cyborg's power in eight years, Vieira has never been a finisher, and five rounds is a long time to survive someone who ends fights. Both halves of that are true. Which one you weight more is the position — and if your read on Vieira is materially north of 26%, that's now a specific, priced disagreement instead of a vibe. Run it through the Probability Converter and then the EV Calculator to see what your number is actually worth against theirs.

The softest number on the card is a man who took the fight on Wednesday

This is the one we'd circle.

Nkosi Ndebele accepted a bantamweight bout with Marcirley Alves on two days' notice after Gustavo Oliveira withdrew — PFL gave no reason. Ndebele is 13-3 and a real fighter; we had him seeded on the scratched Washington DC card in July at 12-3. Alves is 15-5, ranked #8, and the reigning PFL bantamweight tournament winner, in his own division, on a full camp.

FanDuel is asking you to believe the two-days'-notice man wins this better than one time in three.

Set aside what you think of either fighter and just look at the structure of the ask. A short-notice replacement is giving up a weight cut he planned, a camp he never had, a gameplan built for someone else's opponent, and in this case a flight. The question isn't whether Ndebele is a 35% fighter against Alves with eight weeks to prepare — he might well be. It's whether he's a 35% fighter with forty-eight hours. That's an empirical question with a base rate attached to it, and our Base Rate tool exists precisely so you can put a number on the reference class before the narrative gets to you.

Our read: 35% is generous, and Alves at 65% is the most underpriced favorite on this board. We'll grade that publicly on the card page either way.

The tell in the vig: this board was modeled, not shaded

Now the part that only shows up if you do the arithmetic on all eleven bouts at once.

FanDuel's hold on this card, bout by bout: 6.12%, 6.06%, 6.13%, 6.08%, 6.02%, 6.08%, 5.95%, 6.13%, 5.85%, 5.90% — and one outlier, Ibragimov–Basharat at 5.32%.

Ten of eleven bouts are held inside a 28-basis-point band. That is not what a board looks like when a trader has opinions. When a book has a strong view — or is worried about one-way money — it shades the hold: tighter where it wants action, wider where it's exposed. A flat ~6% across ten consecutive fights is the signature of a model output with a uniform margin applied on top, which is exactly what you'd expect on a promotion whose bouts attract a fraction of UFC handle.

Why that matters to you: an unshaded board is a board that hasn't been corrected by money yet. Every price on this card is FanDuel's opinion, not the market's consensus after two days of sharp action. That cuts both ways — it means the errors haven't been arbitraged out, and it means there's no crowd to check your own read against. Ibragimov–Basharat, the single bout priced tighter than the rest, is the only fight on the card where the book behaved like it had a reason to.

Where our own ledger says this board is likely wrong

We grade every bout on every card we cover against the de-vigged line that preceded it. That's not a pick record — we have no MMA pick ledger and won't pretend otherwise — it's a market-calibration stat, and after seven graded 2026 cards it's starting to say something specific.

The 60–69% band is dead calibrated — 65% actual against 66% priced, across twenty bouts. There is no systematic edge there; you need a real opinion about the fight.

The 70–79% band is the one to stare at: 13-2, an 87% hit rate against a 74% average price. That's a +13-point gap. Sample size is fifteen, which is small enough that we're calling it a signal to watch and not a law of nature — but it points the same direction as the +6-point gap across the whole ledger, which is that this board runs conservative on strong favorites rather than gouging them.

Apply it to Tampa and it lands on three fights: Cyborg (74%), Trainer (76%) and Basharat (73%). Those are the bouts sitting in the band where the market has most consistently underpriced its own favorites this year.

And the other half of the finding is a warning: six of the eleven priced bouts on this card sit in the 60–69% band — Alves, Marcos, Santos, Bush, Zaynukov and van Steenis. That's the band where our data says the board is simply right. More than half of this card offers you nothing but a fair price, which is worth knowing before you go looking for edges that aren't there.

The niche inside the niche: the book that priced this card can't take a bet in this city

Every PFL card we cover carries a state-access angle. Tampa's is the sharpest one we've run into.

FanDuel is not live in Florida. The state's mobile sportsbook market runs through a single operator, Hard Rock Bet, under the Seminole Tribe's compact — no FanDuel, no DraftKings. So the board this entire article is built on, the one pricing a card at Benchmark International Arena, cannot accept a wager from anyone inside Tampa city limits. It's a reference price for a market its own author is locked out of.

That's not a knock on Hard Rock Bet, which is a real license-backed operator. It's a structural fact about this specific card: there's exactly one mobile door into a traditional book in this state. Kalshi doesn't have that constraint — it's a CFTC-regulated exchange that works identically in all 50 states, Florida included, so a position on Cyborg's last fight doesn't depend on which single operator holds the local compact. PFL hasn't listed on Kalshi or Polymarket yet, and when it does, the card page flips from the FanDuel reference to live contract prices with no manual update.

How to use this before Saturday

The board is out, which changes the assignment from "write your own number in a vacuum" to "find the places you disagree with a number that already exists." Three concrete ones from above: Vieira's 26%, Ndebele's 35%, and whichever of the two co-main prices turns out to be real.

The Probability Converter turns any American line into the percentage it's claiming, the EV Calculator turns your disagreement into an expected-value number after fees, and the Base Rate tool is the honest way to price a short-notice replacement before the story about him gets to you. Trade responsibly.

Last card: the read on PFL Charlotte's stacked middleweight night lives in our PFL Charlotte coverage. Next up: PFL x RIZIN Osaka on September 10, and the full slate is on the MMA hub.