Kyrie Irving is the last thing in Dallas that has not changed.
The executive who traded for him was fired in November. The head coach who ran his offense parted ways with the club in May. A president arrived in his place who has not seen him play a minute for the team he now runs. Four days ago, on the same afternoon, the assistant coach who worked most closely with him agreed to join Philadelphia and the teammate he was supposed to share a backcourt with was bought out and headed for Miami. Irving himself has not played an NBA minute since March 2025.
Kalshi runs a thirty-outcome market on what his next team is before October 21, with one of those outcomes covering him staying put or retiring. For most of August it barely traded at all. Then, on the Sunday after that double departure, roughly five thousand contracts changed hands in a day, and one outcome briefly quadrupled.
By Tuesday almost all of it had come back. We put the full thirty-outcome board in front of eight AI models that were never shown a single price, then let them read each other's reasoning and revise. The number they came back with is not the one a week like that would suggest, and the place they part from the crowd is a team that sat at a penny all summer and is now the second-most expensive outcome on the board.
New to this kind of market? Start here with how Kalshi's yes-or-no contracts work.
The Eight Weeks That Actually Matter
Almost everything written about Kyrie Irving's trade market this summer is about a date this market does not reach.
The contract language is unambiguous. Each of the 30 outcomes reads the same way, with the team name swapped: "If Kyrie Irving's next team is Boston before Oct 21, 2026, then the market resolves to Yes." The 2026-27 NBA regular season starts October 20. So the window is 56 days from today, it ends the morning after opening night, and its back end is training camp and preseason.
That matters because the freshest reporting on Irving points somewhere else entirely. Marc Stein wrote on August 25 that while Dallas has rebuffed inquiries all offseason, rival teams are watching to see whether Dallas changes its stance "before February's trade deadline". February is months past this market's expiry. A trade at the in-season deadline settles nothing here. Every model on our panel seized on that sentence, and several named it the single most important fact on the board.
One more piece of the settlement language does real work. The favorite is not labeled "Dallas." It is labeled "Stays with Dallas or Retires," and it pays on either. That bundling quietly converts the scariest thing that could happen to a 34-year-old rebuilding a knee into a result that still settles the favorite YES.
Stay ahead of the markets.
Daily insights and expert picks on Kalshi, Polymarket, and what's moving markets.
Free forever. Unsubscribe anytime.
Everybody Who Brought Him Here Is Gone
Irving arrived in Dallas on February 6, 2023, when Brooklyn traded him and Markieff Morris for Dorian Finney-Smith, Spencer Dinwiddie, an unprotected 2029 first-round pick and two second-rounders. He has re-signed there twice since, most recently on July 6, 2025, for three years and $119 million, a deal he agreed to four months after tearing his ACL against Sacramento. He is owed $39.5 million this season with a $42.4 million player option for 2027-28.
Everything around that contract has been replaced. Nico Harrison, the general manager who made the trade, was fired on November 11, 2025. Jason Kidd and the club mutually agreed to part ways on May 19, 2026. Masai Ujiri, the architect of Toronto's 2018-19 championship team, was hired as team president and alternate governor on May 4 and now runs basketball operations. Dusty May was hired off Michigan's bench on June 23. Mike Schmitz is the new general manager. Luka Doncic was traded to the Lakers in February 2025 and Anthony Davis, who came back in that deal, now plays in Washington.
Then came Friday. The Mavericks bought out Klay Thompson, who signed a two-year deal with Miami after clearing waivers. On the same day, assistant coach Phil Handy, the last holdover from Kidd's staff and the coach closest to Irving, agreed to join Nick Nurse in Philadelphia.
The counterweight is what Dallas keeps saying out loud.
"There is only one Kyrie walking around in the world."— Masai Ujiri, Mavericks president, relaying a line from Kevin Durant
Ujiri added that the club wants to see how Irving fits next to Cooper Flagg, the No. 1 pick who won Rookie of the Year on a 26-56 team. On July 28, ESPN's Shams Charania reported Irving fully recovered and "fully entrenched in the Mavericks' plans," having already met with May, Ujiri and Schmitz. ESPN's own player file lists his return date as October 1.
There is one more thing on Irving's record that belongs here, because it is the reason anybody prices this board above a rounding error at all. He has changed teams three times, and each move had a named cause. He asked out of Cleveland in July 2017 and was traded to Boston on August 22, nine years ago last Saturday. He left Boston in free agency in 2019. He asked out of Brooklyn on February 4, 2023, after contract talks soured, three months after the club suspended him for at least five games after he "refused to unequivocally say he has no antisemitic beliefs, nor acknowledge specific hateful material in the film." When Kyrie Irving moves, he moves fast, and he moves because he asked to. Nobody has reported that he has asked.
The Sunday The Board Woke Up
For most of August this market was asleep. The Dallas contract traded 24 contracts on August 18 and about 22 on August 20. Miami did not trade at all on the 20th or the 21st. Twenty-seven of the 30 outcomes carried no bid at all.
Then the Klay and Handy news landed on Friday, and the weekend happened. On Sunday August 23, by Kalshi's own daily candles, the Dallas contract traded 1,638 contracts and dipped as low as 78 cents intraday. Miami traded 1,083 and printed as high as 15 cents, from a two-cent close five days earlier. Philadelphia, home of both Handy and LeBron James, traded 554 and touched 16 cents. Houston traded 903 and touched 8. Detroit, Minnesota and Cleveland traded 476, 243 and 195. On a board that had been moving in double digits a day, roughly 5,100 contracts changed hands in one Sunday.
Then it unwound. As of 2:16 p.m. ET on August 25, Philadelphia was back to a penny bid and three asked. Houston was back to no bid at all. Cleveland had no bid. Dallas was 90 bid and 91 asked, within a cent of where it sat before any of it started. Only Miami had kept anything: four bid, seven asked, the one outcome on the board besides Dallas with a bid above a single cent.
It did not stay unwound. In the three hours after that snapshot the board moved again, quietly and on almost nothing. By 5:11 p.m. ET the Dallas contract had slipped to 81 bid and 85 asked, a spread four times as wide as the one-cent market it had been quoting all afternoon, on 25 contracts traded. And Toronto, which had spent the summer at no bid and a one-cent ask, opened a genuine two-sided quote at three bid and 12 asked, on about two hundred contracts.
Neither of those is a wave, and it would be easy to oversell both. A nine-point drop on 25 contracts is a market maker widening out before the evening, not a verdict; two hundred contracts is a rounding error against the 36,000 that have traded on Dallas. But Toronto had not been quoted above a penny at any point this summer, and it is now the second-highest ask on the board. No reporting we can find connects Irving to Toronto. What we can say is what is adjacent and already on the record: Toronto is the franchise Dallas's new president built a championship for, and Toronto is separately the heavy favorite, at 87 cents, on Kalshi's Kawhi Leonard board. Neither of those is a cause. We are not going to invent one.
Two adjacent transactions, neither of them about Irving, produced one day of trading on a market that had been quiet for a month, and by the second business day most of the move had gone back where it came from. Two of the prices you will still see quoted here are ghosts: Charlotte's last trade printed at six cents and Denver's at four, and neither has traded once in the past three weeks. Both books are zero bid, one asked. A last trade is a memory, not a quote.
The Full Board: What Kalshi Charges And What The Panel Says
A price in cents on Kalshi reads roughly as a probability in percent, because every contract settles at either one dollar or nothing. Two cautions first. Only five of the 30 outcomes have both a buyer and a seller showing: Dallas, Toronto, Miami, Houston and Philadelphia. The other 25 have no bid at all, and every one of those 25 is parked at the one-cent minimum, so the price you see there is what a seller is asking with nobody underneath it. And the whole column of asking prices adds up to 138 cents rather than 100, which overstates every outcome. Take the midpoint of each quote instead, which totals 114.5 cents, and the favorite reads about 72 percent rather than 85.
The panel column is a seat median across eight models, taken after a revision round in which each model read the other seven's anonymized reasoning. None of them saw a price at any point.
| Outcome | Where it stands | Cost to buy | Panel blend |
|---|---|---|---|
| Stays In Dallas Or Retires | Signed through 2027-28, front office publicly rejecting calls, and retirement settles here too | 85¢ | 90.5% |
| Toronto | Quoted at the minimum all summer, then opened a two-sided market on Tuesday afternoon | 12¢ | 0.1% |
| Miami | Just signed Klay Thompson; a real bid on both sides | 8¢ | 1.0% |
| Houston | Spiked hard on Sunday, faded, now bid at the minimum | 5¢ | 0.1% |
| Philadelphia | Hired Phil Handy on Friday; also employs LeBron James | 3¢ | 1.2% |
| Detroit | Named in reporting since June; the panel's top non-Dallas outcome | 1¢ | 2.0% |
| Minnesota | The other team named in that same June report | 1¢ | 1.4% |
| The Other 23 Teams, Each | No reported interest and no bid; the four the panel ranked highest are narrated below | 1¢ | under 1% |
The 23 teams in that last row are Atlanta, Boston, Brooklyn, Charlotte, Chicago, Cleveland, Denver, Golden State, Indiana, the LA Clippers, the LA Lakers, Memphis, Milwaukee, New Orleans, New York, Oklahoma City, Orlando, Phoenix, Portland, Sacramento, San Antonio, Utah and Washington. Every one of them is a listed, tradeable outcome; none carries a bid.
Every Seat's Number
| Seat | Stays with Dallas | Detroit | Minnesota | Philadelphia | Miami | Toronto |
|---|---|---|---|---|---|---|
| Claude Opus | 93.0% | 1.2% | 0.9% | 0.9% | 0.8% | 0.1% |
| DeepSeek V4 | 92.4% | 2.0% | 1.0% | 0.7% | 0.8% | 0.1% |
| Gemini Pro | 92.0% | 2.0% | 1.2% | 0.7% | 0.9% | 0.1% |
| Claude Sonnet | 91.0% | 2.3% | 1.5% | 1.1% | 0.9% | 0.1% |
| Kimi K3 | 90.1% | 2.2% | 1.6% | 1.2% | 1.4% | 0.1% |
| Claude Fable | 90.0% | 2.0% | 1.5% | 1.2% | 1.3% | 0.1% |
| GPT-5.5 | 90.0% | 1.8% | 1.3% | 1.3% | 1.1% | 0.1% |
| GLM 5.2 | 88.0% | 2.5% | 1.5% | 1.5% | 1.5% | 0.1% |
| Seat Median | 90.5% | 2.0% | 1.4% | 1.2% | 1.0% | 0.1% |
The Toronto column is not a formatting error. After the revision round, every seat has the board's second-most expensive outcome at a tenth of a percent or below. They did not all start there: DeepSeek opened at 0.5%, Sonnet and Fable at 0.18%, GPT-5.5 at 0.15%. Reading each other's reasoning moved those four down, while two seats that had been below a tenth drifted up to it. The panel converged on Toronto in the opposite direction from the market, on the same afternoon.
Per-seat figures and the seat median are rounded to one decimal. Model estimates generated August 25, 2026, price-blind, from facts fetched at run time. Market prices quoted from Kalshi at 5:11 p.m. ET on August 25, 2026. These are model estimates, not predictions of fact and not financial advice. Models are frequently wrong and the market price reflects real traders' money. Kalshi is a CFTC-regulated event-contract exchange; 18+, availability varies by state.
Every seat on this panel is graded against real market settlements — records to date: Gemini 86% on 4,760 graded calls · GLM 81% on 2,504 graded calls · Kimi 83% on 2,494 graded calls · DeepSeek 80% on 2,542 graded calls. Recomputed daily; the full scoreboard is public.
Every model estimate here gets graded in public once this market settles, on the full graded scoreboard.
More live boards from the same panel: Kawhi Leonard next team odds (Toronto 87¢ — the same city, and the star who actually is moving) · Trey Murphy III next team odds (Stays with New Orleans 91¢) · Steph Curry next team odds, a board with the same bundled-favorite shape as this one. Prices fetched 5:11 p.m. ET, August 25, 2026.
Settlement timeline
| Date | What happens |
|---|---|
| October 1, 2026 | Irving's return date as listed on his ESPN player file |
| Before Opening Night | Training camp and preseason, the last evaluation window inside this market; ESPN's note has Irving "a full participant by the time training camp comes around" |
| October 20, 2026 | 2026-27 regular season begins |
| October 21, 2026 | This market settles on whichever team he is on |
| February 2027 | The in-season trade deadline, which is outside this board entirely |
This page re-scores when the story moves, and prices on it refresh while a news wave is running. Prices as of 5:11 p.m. ET, August 25, 2026.
The rows worth narrating are Toronto and Detroit. The market's most expensive outcome after Dallas is now Toronto at 12 cents asked; the panel's is Detroit at a penny. Those two rankings are the whole story of this board, and it is worth being blunt about what they do and do not mean.
The only class of panel-versus-market disagreement we have measured is the big one, gaps of ten cents or more, and there the market has been right about two times in three. For most of the day nothing on this board came close to that. Toronto now does, but only depending on which number you use: against the twelve-cent ask the gap clears the bar, and against the seven-and-a-half-cent midpoint of a nine-cent spread it does not. A gap that only exists on one side of a spread that wide is not a finding. It is a reason to check again tomorrow. The models are context and a public record, never a signal.
Hottest Prediction Markets Right Now
- 2028 Democratic presidential nominee · $193M traded
- 2027 Pro Football Champion · $72M traded
- 2028 U.S. Presidential Election winner? · $62M traded
- Pro Baseball Champion · $60M traded
- 2028 Republican presidential nominee · $60M traded
- Bitcoin price at the end of 2026 · $33M 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.
The Longshots The Panel Won't Dismiss
Below Detroit and Minnesota, the panel does not flatten the board into 30 identical zeros. Four outcomes drew real reasoning from at least one seat.
Philadelphia (1.2%, one cent bid and three asked). Handy is there now, and so are LeBron James and Nick Nurse. Gemini Pro made the Sixers its top non-Dallas outcome in the first round on exactly that logic: "Irving's closest coaching confidant, Phil Handy, just joined the 76ers staff; if Irving forces a sudden training camp exit, Philadelphia has the star-chasing mandate to aggregate matching salary." GLM was less impressed on roster grounds: "Philadelphia has the Handy tie but already fields Maxey, Simons, Brown, and LeBron."
Miami (1.0%, five bid and eight asked). The market's favorite non-Dallas outcome is the panel's fourth. DeepSeek made the case in round one, before it read anyone else: Miami has "win-now roster construction with Giannis and Adebayo," a friend of Irving's newly signed, and a plausible salary-match path. It then revised the number down after the other seats pointed at the calendar.
Cleveland (0.5%, no bid). The homecoming row, and the third-highest-volume contract on the board at roughly 25,900 traded lifetime with about 10,500 still open. Every seat kept it a step above the flat tail, GLM highest at about one percent, even though Cleveland already has James Harden and Donovan Mitchell in the backcourt. The Lakers (0.5%, no bid) landed in the same place across the whole panel; DeepSeek listed them as a "perennial Kyrie destination."
None of these is a hidden favorite. They are the small, specific chances a board with a bundled favorite and 56 days left will sell for almost nothing. The genuinely flat part of the board starts below them: 23 teams the panel scored between a tenth and a third of a percent. That is the panel's ranking, not the market's, and the two lists come apart at the bottom. Twenty-one of those 23 ask a penny. The other two are the ones the market prices well above the floor: Houston at five cents and Toronto at 12.
Where The Panel Changed Its Mind
The revision round moved one seat a very long way.
DeepSeek opened at 74.6% on Dallas, the lowest number on the panel by more than ten points — and not because it had a trade thesis. Its round-one note argued the other way: Irving's past moves all came from an explicit request or from free agency, and "no such request exists here." The low number came from the shape of its tail. It spread 11.4 points across 23 teams it could name no mechanism for, and gave Miami 3.98% on the Thompson friendship, the Heat's win-now roster around Giannis and Adebayo, and a plausible salary-match path. Then it read the other seven and moved to 92.4%, one of the two highest numbers on the board. Its stated reason was the calendar, and specifically Stein: "The Aug 25 Stein item explicitly frames any change in Dallas's stance as a question for February's deadline, which is outside this market." It also reframed its own bear case: "The staff turnover (Handy's departure) is a comfort signal, not a trade mechanism."
Claude Opus barely moved, from 94.0% to 93.0%, and explained why it sat above the group on a structural point rather than a news read. "This is a no-move market wearing a destination market's costume," it wrote. "The DAL contract also absorbs retirement and any catastrophic camp setback, so the failure modes that would kill a trade mostly settle DAL YES anyway." On the trade window itself: "The late-August-to-opening-night corridor is the deadest trade window in the NBA year," it added, with rosters built, cap flexibility spent, and the deals that do happen in that stretch being salary dumps rather than $39.5 million stars.
GLM 5.2 did not move at all, printing 88.0% in both rounds and finishing as the panel's lowest number, and it was the one seat that refused to let the Friday news go. "The comfort structure around Irving is genuinely dismantled," it wrote, listing the coaching changes and the Thompson signing, "and Irving has forced exits from less." It kept its number where it was because the other half of its reasoning had not changed either: matching $39.5 million for a 34-year-old with no game tape since March 2025 is close to unbuildable before camp.
What Would Change The Panel's Mind
Every seat was asked what would move its number. Four triggers came back repeatedly, and each one is checkable.
A reported trade request. This preceded both of the trades that moved him, in 2017 and in 2023, and every seat named it first. Claude Fable put a size on it: a reported request or camp holdout "would cut DAL by 25+ points toward DET/MIA/PHI."
Dallas softening its language. Not a deal, just a change in posture. Opus: Ujiri going from "we want to see him with Flagg" to "we'll listen" before camp "cuts DAL to the high 80s and fattens DET/MIN/MIA." This is the trigger most likely to fire quietly, in a single beat-writer sentence.
A named framework, not "interest." Several seats drew a hard line between monitoring and offering. A report naming a specific team with specific players attached would move that team to five or ten percent on its own, per Opus. Nothing like that has been reported.
A knee setback before camp. The counterintuitive one. Every seat that raised it pointed the same way: a rehab setback pushes the favorite up, not down, because an Irving who cannot pass a physical is untradeable and because retirement settles the same contract YES. Sonnet was the lone hedge, allowing that a failed physical could also trigger a change of heart.
More on this: Klay Thompson Next Team Odds: He Chose Miami, And The Market Barely Blinked · Draymond Green Next Team Odds: Kalshi Odds · Jalen Duren Next Team Odds: The Kalshi Board, Scored · James Harden Next Team Odds: Kalshi's Cavs Verdict · Trey Murphy III Next Team Odds: One Suitor: Kalshi Odds
The Bottom Line
Strip away the week and this is a market about a calendar. Fifty-six days, a training camp and a preseason, a player who has not been on a floor in 17 months, a $39.5 million salary that is hard to move at the exact moment its owner is least legible, and a front office that has spent every month since May saying no in public. The people paid to know when Irving might move are pointing at February, and February is not on this board.
For most of Tuesday the panel and the crowd agreed almost exactly: 90.5% against a 91-cent ask, which is as close as these two ever get. By late afternoon the crowd had drifted to 81 bid and 85 asked, a midpoint of 83, leaving the panel's number seven and a half points high. That gap opened on 25 contracts, which is another way of saying it may not be a gap at all — only a wider spread going into the evening.
The sharper split is Toronto, and it is the one we would rather be honest about than clever with. Sellers are asking 12 cents for it, with buyers three cents underneath. All eight models, none of which ever saw a price, put it at a tenth of a percent, unanimously, and none of them moved when they read each other. We can tell you exactly what the panel knew when it said that: a $39.5 million salary, 56 days, no game tape since March 2025, and not one report anywhere connecting Irving to Toronto. We cannot tell you what the 12 cents knows. On a board where the crowd has usually been the one reading something, that asymmetry is worth more to a reader than a confident number would be.
The thing worth watching is not a trade rumor. It is a sentence. Dallas has been saying the same thing since May, and the day it starts saying something slightly different, this board will move before anyone files a story about a deal.
