Big Brother 28 Winner Odds: The AI Model Verdict

Five weeks into Big Brother 28, the season subtitled "Time Trip," Kalshi has turned the house into 17 separate contracts, one per houseguest, and traders have sorted them into a clear pecking order. I asked six AI models to score the whole field without letting any of them see a single price. They agree with the market on one co-leader, split hard from it on the other, and back a houseguest the crowd has left for dead.

The Quick Answer

Kalshi traders make Survivor 45 winner Dee Valladares the favorite near 22 cents, with two-time Survivor player Rick Devens second near 18 cents. Our price-blind, six-model panel keeps Dee at the top but fades her to 13.2%, well below her de-vigged price, and it splits sharply from the market on the co-leader: it blends Kamu Kirk to 12.8% against a market price of just 8 cents, because Kamu just won the Week 5 Power of Veto to go with a Head of Household win and sits safely inside the ruling alliance. The panel also fades Rick Devens hard, down to 10%, reading his big Survivor résumé as a target rather than a shield.

• Market Favorite: Dee Valladares (~22¢), the Survivor winner inside the dominant Toolshed alliance
• The Panel's Co-Leaders: Dee Valladares (13.2%) and Kamu Kirk (12.8%), though Dee is faded below her price while Kamu is lifted far above his
• The Biggest Market-Vs-Model Gap: Kamu Kirk, the Week 5 Veto winner, blended to 12.8% against an 8-cent price
• The Panel's Clearest Fade: Rick Devens, the market's 18-cent second choice, scored down to 10% as a comp-beast target
• The Best Value The Models Keep Circling: Drew Campbell, a quiet Toolshed member with no comp wins, at 9.8% against 7 cents
• The Dead Rows: Ashley Trail, Rome Seymour, Jason De Puy, and now Lyric Medeiros, all evicted and scored at zero
• The Honest Caveat: it is only Week 5, roughly eight more weeks and close to a dozen evictions remain, and every model expects the eight-person alliance to eat itself in ways no one can time.

Free: The Weekly PM Market Brief — the 8-model panel's graded record, the week's biggest market-vs-model gaps, and what's spiking next. One email, Sundays. Sign up in the box at the end of this article.

New to Kalshi event markets? Start here → how prediction markets actually work.

The full 17-name board, the case for and against the names that carry the story, and a plain-English guide to reading these prices are all below.

The Market At A Glance

Venue	Kalshi, a CFTC-regulated event-contract exchange (18+; availability varies by state, as of August 2026)
The Event	"Big Brother Season 28 Winner?", 17 separate yes/no contracts (Kalshi event KXBIGBROTHER-26DEC31)
The Contract	Each market resolves YES only if that houseguest is officially declared the winner of Big Brother 28, per Kalshi's settlement rules
Decided	On the scheduled finale, October 1, 2026, when the jury votes; the market's expected expiration is a December 31, 2026 buffer
Prices Below As Of	midday August 10, 2026 (Week 5)

These are model estimates, not predictions of fact and not financial advice. Kalshi event contracts trade on a CFTC-regulated exchange; you must be 18+ and in an eligible state to participate. Prices on this page are refreshed during the live news wave, so the quotes below carry the dated "as of" stamp above rather than a frozen snapshot.

What Decides A Big Brother Winner

Before any houseguest's name, the shape of this game explains why the models built the board the way they did. Three forces drive almost every winner's path:

• You Have To Survive, Then You Have To Win The Vote. A Big Brother champion must do two separate things: last through ten-plus more evictions to reach the final two, and then win a jury vote made up of the players who were cut along the way. A brilliant early game counts for nothing if you are evicted in week nine, which is why a mid-season lead is a soft signal, not a result.
• Comp Wins Buy Safety But Paint A Target. Winning Head of Household or the Power of Veto keeps you safe for a week, but stacking them early marks you as a threat the house wants gone. The players who reach the end are often the ones managing relationships quietly, not the ones dominating competitions, so raw comp counts can mislead.
• Alliance Position Is The Real Currency This Season. For four straight weeks one alliance, the eight-person "Toolshed," won every Head of Household and controlled every vote. Sitting inside a majority bloc is the cleanest path to the endgame, but that same bloc cannot send all eight of its members to the final two, so its own fracture is the pivotal event nobody can schedule. In Week 5 that fracture finally started.

There is one more truth the panel kept circling: it is early. With about eight weeks left, the honest read is a wide field where a dozen live houseguests hold real probability, not a two-horse race. That is the single biggest gap between the models and a market that has already crowned a top two.

The Board Right Now

Here is the field, sorted by the model blend, in the phone-friendly version first. The four evicted houseguests are shown at the bottom at zero.

Houseguest	Status	Market	Model blend
Dee Valladares	In	22¢	13.2%
Kamu Kirk	In	8¢	12.8%
Rick Devens	In	18¢	10.0%
Drew Campbell	In	7¢	9.8%
Angela Murray	In	6¢	8.5%
Taylor Brown	In	9¢	8.0%
LaTrice Verrett	In	4¢	7.7%
Mallory Aurichio	In	7¢	6.3%
Melody Morris	In	5¢	5.2%
Chuk Anyanwu	In	1¢	5.0%
Barrett Pfeiffer	In	7¢	4.8%
Haley Thogmartin	In	1¢	4.8%
Yash Patel	In	7¢	4.5%
Lyric Medeiros	Evicted	1¢	0.0%
Ashley Trail	Evicted	1¢	0.0%
Rome Seymour	Evicted	1¢	0.0%
Jason De Puy	Evicted	1¢	0.0%

These are model estimates, not predictions of fact and not financial advice. Market figures are raw YES quotes from Kalshi and include the exchange's bid/ask spread, so the 17 listed prices sum to about 106 cents, over 100%. The de-vigged comparison is worked through below.

And here is every seat's number, so you can see where the panel agrees and where it splits.

Houseguest	Status	Market	Opus	Sonnet	Fable	GLM	Kimi	DeepSeek	Model blend
Dee Valladares	in	22¢	13	14	11	14	14	13	13.2%
Kamu Kirk	in	8¢	12	12	13	13	13	14	12.8%
Rick Devens	in	18¢	9	11	9	11	11	9	10.0%
Drew Campbell	in	7¢	10	8	10	10	11	10	9.8%
Angela Murray	in	6¢	9	8	8	9	9	8	8.5%
Taylor Brown	in	9¢	8	8	9	8	6	9	8.0%
LaTrice Verrett	in	4¢	7	7	8	8	7	9	7.7%
Mallory Aurichio	in	7¢	7	6	8	5	5	7	6.3%
Melody Morris	in	5¢	6	6	6	4	4	5	5.2%
Chuk Anyanwu	in	1¢	5	5	5	7	5	3	5.0%
Barrett Pfeiffer	in	7¢	5	4	4	7	5	4	4.8%
Haley Thogmartin	in	1¢	4	5	4	6	6	4	4.8%
Yash Patel	in	7¢	5	5	5	3	4	5	4.5%
Lyric Medeiros	evicted	1¢	0	0	0	0	0	0	0.0%
Ashley Trail	evicted	1¢	0	0	0	0	0	0	0.0%
Rome Seymour	evicted	1¢	0	0	0	0	0	0	0.0%
Jason De Puy	evicted	1¢	0	0	0	0	0	0	0.0%

These are model estimates, not predictions of fact and not financial advice. Every number here also lands in our full graded scoreboard, where past price-blind calls are graded against settled results so the method is measured over time, not just asserted.

A note on the panel. Our standard Model Verdict lineup runs eight seats. On this run the ChatGPT, Gemini, and Grok seats were unavailable on the machine that produced these verdicts, so the board runs on six models: Claude Opus, Claude Sonnet, Claude Fable, GLM, Kimi, and DeepSeek. All six scored the field on August 10, 2026, without seeing any prices, then read each other's anonymized reasoning in a revision round before finalizing.

The one number that matters most: the panel's ceiling is 13.2%. No houseguest, not the Survivor 45 winner and not the double comp winner, cleared 14% in the blend. That is the models saying, plainly, that Week 5 is still too early for anyone to be a true favorite.

With the field in view, four names carry the story: the two co-leaders the models respect, the Veto winner they lift far above his price, and the Survivor résumé they refuse to trust.

Kamu Kirk: The 8-Cent Contract The Models Make A Co-Favorite

This is where the panel departs hardest from the crowd. Kamu Kirk is priced at just 8 cents on Kalshi, the No. 4 contract but far back of the top two, yet the models blend him to 12.8%, a near-tie for the top of the board. The reason is the most concrete thing on it: Kamu just won the Week 5 Power of Veto, which means that with three Toolshed allies on the block, he personally controls this week's replacement nomination. That is leverage with every camp in the house at the exact moment the season's first fracture opens.

Three seats made him their outright top pick. As DeepSeek put it, he is "safe with the Week 5 Veto, owns an HOH and Veto win, and sits in the Toolshed's middle tier where he can survive the fracture without the biggest target." Fable was blunter about why that middle-tier profile beats the flashier names: Kamu is "the only player currently holding cross-alliance leverage at the season's first fracture point, without Dee's winner-reputation target." Two comp wins buy safety; a reputation that fits in a sentence keeps the house from cutting you for it.

Dee Valladares: The Favorite The Models Keep, But Fade

Every model still put Dee Valladares at or near the top, and she edges Kamu for the No. 1 blend at 13.2%. The case is real: a Survivor 45 winner who entered as the season's secret 17th houseguest, banked a Week 1 Head of Household, and has since lowered her profile inside the Toolshed. As Opus framed it, she brings "elite game IQ, secure inside the dominant Toolshed, one comp win, and the best jury-management instincts in the house."

But the models will not price her as a runaway. Kalshi's 22 cents de-vigs to about 20.8%; the panel shaves her to 13.2%, roughly a third lower. The reason is the same target logic that shapes the whole board. Kimi capped her deliberately, noting that "her Survivor-winner reputation stays a live target once the Toolshed cannibalizes," and applied a hard base rate on top: a Week-5 favorite "converts only about one-in-seven" of the time. Every seat believes Dee is one of the two best players in the house. None believes it is safe to pay 22 cents for a player this visible with the bulk of the season still to survive.

Rick Devens And The Survivor-Résumé Fade

If the models lift one Survivor name and merely trim the other, they actively fade the third. Rick Devens, a two-time Survivor player with a Week 2 Head of Household and the Week 2 Veto already banked, is the market's second choice at 18 cents (a de-vigged 17%). The panel writes him down to 10%, and the reasoning was close to unanimous. Fable called him "casual money — two-time Survivor fame plus an HOH/Veto week, but that exact profile is who a fracturing eight-person alliance cuts first." Opus, GLM, and Kimi all independently named him the field's most overrated houseguest for the same reason: in a house about to turn on itself, the most decorated résumé is the first one sacrificed.

The flip side of that fade is where the models find value. Every seat that named a best-value pick pointed at a quiet Toolshed member rather than a star. Drew Campbell, at 9.8% against a 7-cent price, drew the most votes. Opus described "a quiet Toolshed member with no comp-threat profile and no target on his back, the classic under-the-radar floater who drifts to the final two," and GLM, DeepSeek, and Fable said nearly the same thing in their own words. Angela Murray, the only houseguest with prior Big Brother experience (she played Big Brother 26), gets the same treatment at 8.5% against 6 cents: alliance safety without the heat that comes from winning everything.

Dark Horses The Panel Won't Dismiss

The market has already buried most of the house. The models have not.

• LaTrice Verrett (7.7% Vs A 4-Cent Price). At 58, "Lala" became the oldest houseguest ever to win the endurance Wall competition, seizing the Week 5 Head of Household and cracking the Toolshed's grip for the first time all season. Sonnet called that "an underrated winner-arc signal," and Kimi's read was sharper still: she is "dismissed as a 58-year-old one-week wonder, but she's safe now, owns the season's biggest move, and a Red Corner flip hands her a winning jury story." The caution, which every seat shared, is that by nominating three Toolshed members she has painted herself as the next retaliation target the moment her HOH ends.
• Taylor Brown (8.0% Vs 9 Cents). She saved herself off the block with a Week 4 Veto win and has now survived nominations on her own merit. The models like the fighter's résumé but cap her for a lack of alliance protection.
• Mallory Aurichio (6.3% Vs 7 Cents). A Week 1 Veto winner who has twice slipped off the block. Fable and DeepSeek kept her in the quiet-shield tier — real survival instincts, thin path to a jury win.

The Block, And The Dead Rows

Three names sit on the block this week, and the panel discounts all three despite their Toolshed membership. Haley Thogmartin, a Week 4 Head of Household, is down to 4.8%; Chuk Anyanwu is at 5.0%; and Barrett Pfeiffer, tied with Haley for the lowest of the three at 4.8% on a still-empty comp résumé, has fallen from a 12-cent market third choice a few weeks ago all the way to 7 cents as the house turned on him. As Sonnet flagged, Haley's name recognition is exactly the trap: "a proven HOH winner, but she's on the block this week facing a real elimination threat right now." One of the three most likely leaves within days, so the market's 1-cent prices on Haley and Chuk are, for once, close to the models' own read.

Then the board's dead rows. Ashley Trail (evicted Week 1), Rome Seymour (Week 2), Jason De Puy (Week 3), and Lyric Medeiros — blindsided in an 8-3 Week 4 vote — are out of the game and cannot win, so every model scored them at zero. Kalshi still lists each near 1 cent, the residual noise of a market that never quite settles a losing contract to zero.

Where The Panel Changed Its Mind

After the first round, every model read the others' reasoning anonymously and could revise. The movement was small but pointed in one direction: toward the Veto winner.

• Fable moved Kamu past Dee, writing that another seat's point about "Kamu's Week-5 Veto giving him sole control of the replacement nomination — leverage inside AND outside the fracturing Toolshed — lifted him past Dee, while the cannibalization logic trimmed both [Dee and Rick] a point." — Fable
• Kimi split the difference, taking the argument in but holding the ceiling: "Forecaster D's argument that Kamu controls this week's reshuffle while already holding two comp wins moved him up my board, while a base-rate warning that Week-5 favorites convert only about one-in-seven kept Dee capped at 14% rather than higher." — Kimi
• Sonnet held its ground and said why, the most useful kind of dissent: "I keep Dee over Kamu because a Survivor-winner's social/strategic profile is the more durable BB-winning archetype than a two-comp-win threat who is now accumulating target equity the game itself cuts before finale." — Sonnet

What Would Change The Panel's Mind

This board is built for a house that is about to move. Four concrete events would shift the numbers:

• This Week's Eviction (On Or Around August 14). If a Toolshed member (Haley, Chuk, or Barrett) is voted out, the fracture the models are pricing becomes real — and the quiet survivors, Kamu and Drew, get a lift. If the Veto reshuffle saves the target and the vote lands elsewhere, the alliance holds and Dee's number firms.
• What Kamu Does With The Veto. He controls the replacement nomination. Using it to protect an ally cements the cross-alliance leverage that made him a co-favorite; a passive week softens the whole case for his 12.8%.
• The Next Head Of Household. LaTrice cannot compete in the upcoming HOH. If a Toolshed member reclaims it, her 7.7% fades toward her price; if the rival "Red Corner" seizes power, her dark-horse jury story gets real, and so does the case for a non-Toolshed winner.
• Dee Or Rick Hitting The Block. The models' entire fade on the Survivor names rests on the theory that big résumés get cut. The first time one of them is nominated, that thesis is either confirmed — collapsing the top of the board — or refuted.

More live boards from the same panel: Oscars Best Picture 2027, where Christopher Nolan's The Odyssey leads at ~50¢; and Next James Bond, where Callum Turner leads at ~36¢. Prices fetched August 10, 2026.

How To Read A Kalshi Winner Price Like This

The models can tell you where they disagree with the crowd, but the number in front of you still has to be understood. Here is how to read a winner-market price like the ones in the board above, plus the honest friction that comes with a season-long market.

Work an example. Say Dee Valladares shows a 22-cent YES price. That is not a clean "22% chance." On Kalshi the YES and NO quotes carry a spread, and across a mutually exclusive field the individual YES prices collectively add up to more than 100 cents — an overround baked into the quotes. Here the 17 contracts sum to about 106 cents of implied probability. To get closer to a fair read, compare each name's price to that total and re-scale: Dee at 22 cents is really about 20.8% once you strip the overround out (22 divided by 106). That is why the panel's 13.2% blend on Dee reads as the models sitting clearly below the crowd, and why Kamu's 12.8% against a de-vigged 7.5% is the board's real disagreement.

Houseguest	Raw YES price	De-vigged (÷1.06)	Model blend
Dee Valladares	22¢	20.8%	13.2%
Rick Devens	18¢	17.0%	10.0%
Kamu Kirk	8¢	7.5%	12.8%
Drew Campbell	7¢	6.6%	9.8%
LaTrice Verrett	4¢	3.8%	7.7%

De-vigged column re-scales each raw price by the field's ~106¢ total to strip the overround. These are model estimates, not predictions of fact and not financial advice.

That habit of reading a fair, vig-free version of every number is the same discipline as line shopping applied to a winner board: checking what a spread-free version of a price looks like before you trust it, the same way you would convert Kalshi's cents into real odds for comparison. A few honest points before anyone acts on a number in the table above:

• Eligibility. Kalshi is a CFTC-regulated event-contract exchange, not a sportsbook. You must be 18+ and located in an eligible state, and availability varies by state and can change.
• Liquidity And Spread. Season-long winner contracts on the tail houseguests are thin. Many of these markets show a wide bid/ask gap, so the "price" on a 1-cent name is softer than a single number suggests.
• Time And Capital. This market does not settle until the fall, so any position ties up money for months with no interim payout.
• Not Advice. Everything here is analysis for information, not a recommendation to trade. These are model estimates, not predictions of fact and not financial advice.

Settlement Timeline

Settles	On the winner's declaration at the live finale, scheduled October 1, 2026 (Kalshi contracts carry a December 31, 2026 expiration buffer)
Next Catalysts	Week 5 eviction on or around August 14, 2026 · a new Head of Household the same week · weekly evictions every week through the late-September double-eviction stretch
Re-Scored	This board is re-scored whenever the story moves: a new eviction, a shift in alliance control, or a comp result that changes a top name's safety. As of August 10, 2026 (Week 5).

The Bottom Line

The market and the models agree on half of Big Brother 28: Dee Valladares, the Survivor 45 winner, belongs at the top of a house the Toolshed alliance ran for four straight weeks. Where the panel adds something is everywhere else. It fades Dee below her price and fades Rick Devens hard, because in a house about to cannibalize itself the biggest résumés are the first cuts. It makes an 8-cent contract, Week 5 Veto winner Kamu Kirk, a co-favorite. And it keeps backing the quiet Toolshed members — Drew Campbell, Angela Murray — and a 58-year-old who just cracked the alliance's grip. Weigh all of it against the caveat the models kept returning to: it is Week 5, close to a dozen evictions still have to happen, and an eight-person alliance never reaches the end intact. If this method is your thing, our Oscars Best Picture 2027 model verdict and next James Bond model verdict score every contender the same price-blind way.

FAQ

Who is the favorite to win Big Brother 28? Kalshi traders make Dee Valladares the favorite near 22 cents, with Rick Devens second near 18 cents. Our six-model AI panel co-leads Dee (13.2%) with Kamu Kirk (12.8%), the houseguest who just won the Week 5 Power of Veto but whom the market prices at only 8 cents. The panel fades the market's Survivor names because it is only Week 5 and a long game remains.

Why do the AI models rate Kamu Kirk so much higher than the market? Kamu Kirk sits in the dominant Toolshed alliance, owns two competition wins (the Week 3 HOH and the Week 5 Veto), is not on the block, and controls this week's replacement nomination. The panel reads that as co-favorite leverage without the target that comes with a big Survivor reputation, so it blends him to 12.8% against an 8-cent price.

Which Big Brother 28 houseguests have been evicted? Four are out and cannot win, so the models score them at zero: Ashley Trail (Week 1), Rome Seymour (Week 2), Jason De Puy (Week 3), and Lyric Medeiros, blindsided in a Week 4 vote. Kalshi still lists each near 1 cent as residual noise.

When is the Big Brother 28 finale? The season is scheduled to end with a live finale on October 1, 2026, when a jury of the evicted houseguests votes for the winner. The Kalshi winner contracts carry a December 31, 2026 expiration as a settlement buffer.

How were these Big Brother 28 model estimates produced? Six AI models scored the entire 17-name field without seeing any market prices, working only from the live game state fetched at runtime, the evictions, the weekly HOH and Veto winners, and each houseguest's alliance position, then read each other's anonymized reasoning in a revision round. These are model estimates, not predictions of fact and not financial advice.

Hottest Prediction Markets Right Now

• 2028 Democratic presidential nominee · $166M traded
• Rocket Classic Winner · $124M traded
• 2027 Pro Football Champion · $51M traded
• Pro Baseball Champion · $40M traded
• Bitcoin price at the end of 2026 · $30M traded
• 2027 Pro Basketball Champion · $19M traded

Every market above links to our full AI model verdict; browse them all on the OddsShopper prediction markets hub, and see how every settled call actually scored on the full graded scoreboard.

Verdicts generated August 10, 2026. Prices as of August 10, 2026 (Week 5). Panel: Claude Opus, Claude Sonnet, Claude Fable, GLM, Kimi, and DeepSeek, price-blind. These are model estimates, not predictions of fact and not financial advice.