WNBA Championship Odds: The Lynx Were Too Good And Too Expensive

Eight AI models that never saw a price put the Minnesota Lynx under the market twice: 30% against 42 cents in July, 37.4% against 49 cents in August. The contract paid nothing. The New York Liberty, the 8 seed, swept the Lynx out of the 2026 WNBA playoffs, 91-75 in Minneapolis on Sunday, September 27, and 87-71 in Brooklyn on Tuesday, September 29. Kalshi graded the Minnesota title contract NO at 11:16 p.m. ET that night.

That is the grade, and it needs saying precisely. The panel had the Lynx first on every seat, so it did not see the sweep coming any more than the market did. What it would not do was pay 49 cents for a team that still had to win three series, and 49 cents paid zero. The full record, nine of the 15 contracts settled, is in The Record below, with the six open contracts repriced as of Thursday night.

Here is what the board looked like when the panel scored it. The Lynx were 31-7 with the best point differential in the WNBA, and they had built almost all of it while the best player on the roster was recovering from surgery on both ankles. Napheesa Collier had operations on her right ankle in January and her left ankle on March 24, sat out the first ten weeks of the season, and came back against Seattle in late July for 22 minutes, her first competitive basketball since October. Minnesota went 22-6 in the 28 games she missed and 9-1 in the 10 she had played.

So the team that already had the league's best record had just got its most dangerous player back, and the 2026 WNBA championship odds on Kalshi had the Lynx at 49 cents on the ask, close to a coin flip in a 13-team field. We handed eight AI models the same fetched picture of this season, hid every market price from them, and asked each one to split 100 points across the whole field. All eight came back with Minnesota first. The highest final number any of them gave the Lynx was 44.

The Quick Answer

The Minnesota Lynx, 49¢ on Kalshi and 37.4% on the panel in August, were swept in the first round by the 8-seed New York Liberty and settled NO on September 29. Nine of the 15 contracts are graded, all NO. Seven of them were on the board the panel scored, and on the only one where the two sides were more than a point apart, Minnesota, the panel's number was under the market. Six teams remain. Shortly before 9 p.m. ET on Thursday, October 1, the Atlanta Dream traded at 30¢ on the ask, Las Vegas and New York at 22¢, the Golden State Valkyries at 19¢, Indiana at 10¢ and Dallas at 2¢. The panel's August read on those six, in the same order, was 9.9%, 13.6%, 8.9%, 12.9%, 12.1% and 2.0%.

The August board, the Valkyries case where the panel and the market disagreed most, the reason a one-cent spread destroyed most of the value at the bottom of that board, and the graded record are all below.

New to Kalshi event markets? Our guide to how these basketball contracts settle and trade covers the mechanics before you look at a single price.

Every price in the board below was read from Kalshi's book as of August 22, 2026, and the panel scored the board the same day without seeing any of them. The board and the seat quotes are frozen as scored. The Record, further down, carries the settlements and the October 1 pull on the six contracts still trading.

The Board

The August 22 board first, narrow and built to read on a phone. "Ask" is what a buyer actually pays for one contract, in cents on the dollar. "Panel" is the blend: take the middle of the eight model numbers for each team after a revision round in which each model read the others' reasoning, then rescale the whole column so the underlying numbers add to 100, which is what a set of teams competing for one title has to do. The printed figures are rounded to one decimal, so they land a couple of tenths short of 100.

Team	Record	Ask	Panel
Minnesota Lynx	31-7	49¢	37.4%
Las Vegas Aces	25-13	14¢	13.6%
Golden State Valkyries	25-11	8¢	12.9%
Indiana Fever	24-13	14¢	12.1%
Atlanta Dream	23-13	10¢	9.9%
New York Liberty	22-15	7¢	8.9%
Washington Mystics	21-15	2¢	2.3%
Dallas Wings	21-16	2¢	2.0%
Portland Fire	15-21	1¢	0.3%
Chicago Sky	15-22	1¢	0.3%
Phoenix Mercury	13-23	1¢	0.1%
Los Angeles Sparks	12-24	1¢	0.0%
Toronto Tempo	11-25	1¢	0.0%

Four of the 15 teams could no longer win this title on August 22. The Seattle Storm and the Connecticut Sun had already settled at zero and stopped trading. The Los Angeles Sparks and the Toronto Tempo were also eliminated by ESPN's standings, and the arithmetic was not close: eight teams had already reached 21 wins, Los Angeles could finish with at most 20 and Toronto with at most 19. Both contracts were still on the board at a penny, and Kalshi graded them NO two days later.

Every Seat's Number

The full per-model board. Each seat priced all 13 teams in one pass rather than one team at a time, because exactly one contract here can pay out and treating the teams as independent is the classic way to get a field like this wrong.

Team	Fable	Opus	Sonnet	GPT	Gemini	GLM	Kimi	DeepSeek	Panel
Minnesota Lynx	36	38	34	35	38	37.5	39	44	37.4
Las Vegas Aces	12.5	14	15	13.5	13	14.58	14	13	13.6
Golden State Valkyries	14	11.5	13	14	12	13.54	13	12	12.9
Indiana Fever	13	13.5	13	12	11	12.5	11	8.5	12.1
Atlanta Dream	10	10.5	10	10	10	7.29	8	8	9.9
New York Liberty	8.5	8.5	9	8.5	9	9.9	9	11	8.9
Washington Mystics	2.5	1.5	2.5	2.1	3	2.08	2.5	2	2.3
Dallas Wings	3	2	1.7	2.5	3	2.08	2	1	2.0
Portland Fire	0.2	0.2	1	0.9	0.4	0.31	0.8	0.1	0.3
Chicago Sky	0.2	0.2	0.5	0.8	0.4	0.21	0.4	0.1	0.3
Phoenix Mercury	0.1	0.1	0.3	0.7	0.2	0	0.3	0.1	0.1
Los Angeles Sparks	0	0	0	0	0	0	0	0.1	0.0
Toronto Tempo	0	0	0	0	0	0	0	0.1	0.0

Every seat on this panel is graded against real market settlements — records to date: Claude Fable 88% on 5,981 graded calls · Claude Opus 87% on 6,126 graded calls · Claude Sonnet 86% on 6,070 graded calls · GPT 85% on 13,424 graded calls · Gemini 86% on 12,259 graded calls · GLM 84% on 5,553 graded calls · Kimi 86% on 7,272 graded calls · DeepSeek 84% on 5,346 graded calls. Recomputed daily; the full scoreboard is public.

On sports contracts specifically, these eight seats carry 31,789 graded calls between them, and the three sharpest are DeepSeek at 87.2% on 2,995 sports calls, Fable at 87.0% on 3,896, and Opus at 86.5% on 3,972. Sports is about half of everything the panel has been graded on, the longest paper trail it has, which is exactly why a board like this one is worth reading closely. These are estimates, not predictions of fact and not financial advice. Every number in this piece gets graded in public once the market settles, and you can see the running tally on the full graded scoreboard.

More boards from the same panel: WNBA MVP 2026 · Caitlin Clark's next team · NBA 2027 champion.

The Market At A Glance

Market	2026 WNBA champion, one contract per team
Venue	Kalshi, a CFTC-regulated event-contract exchange. You must be 18+ and in an eligible state to trade.
Ticker	KXWNBA-26, 15 contracts listed, 6 still trading, 9 settled NO
Settles	"If Indiana wins the 2026 Women's Pro Basketball Championship, then the market resolves to Yes." Same sentence with a different team name on all 15, and each contract settles once a champion is declared.
Expires	Contracts close early once a champion is declared. Kalshi's payload lists an expected expiration of December 1, 2026 and a latest possible expiration of November 30, 2028.
Market Favorite	Minnesota Lynx, 49¢ on the ask on August 22, settled NO September 29. Atlanta Dream, 30¢ on the ask on October 1
Panel Favorite	Minnesota Lynx, 37.4% on August 22, graded NO
Prices As Of	Kalshi order book, August 22, 2026, for the board; October 1, 2026, shortly before 9 p.m. ET, for The Record
Verdicts Generated	August 22, 2026, with prices hidden from every model

Hottest Prediction Markets Right Now

• 2028 Democratic presidential nominee · $188M traded
• 2027 Pro Football Champion · $69M traded
• 2028 U.S. Presidential Election winner? · $61M traded
• 2028 Republican presidential nominee · $59M traded
• Pro Baseball Champion · $58M traded
• Bitcoin price at the end of 2026 · $32M 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.

Minnesota Was Better Than Its Record And Still Not Worth 49 Cents

The unusual thing about the panel's read on Minnesota was that all eight models argued the Lynx were stronger than their 31-7 record, and all eight still priced them below the market. Both halves of that were worth taking seriously, and both halves have now been graded.

The first half was an availability argument. Collier was averaging 19.0 points in the 10 games she had played, and Minnesota's league-best plus-8.8 point differential was accumulated without her for 28 of 38 games. Add a scorer of that caliber to a team that was 17-2 on the road and 9-1 in its last ten, and you were not looking at a team the standings flattered.

Minnesota is 31-7 at +8.8 with Napheesa Collier (19.0 ppg) in only 10 of 38 games, and 17-2 on the road — a team whose ledger understates it. Run the actual series math with Collier back: roughly 83% in a best-of-three over Dallas, ~72% in a best-of-five, ~70% in a best-of-seven, which multiplies to ~42%. — Opus

What the models were weighing on Minnesota, in three numbers:

• 22-6 Without Collier, 9-1 With Her. Minnesota outscored opponents by 8.7 a night in the 28 games she missed and by 9.4 a night in the 10 she had played.
• 17-2 On The Road. Home court through the bracket is worth less to a team that already wins anywhere.
• Three Rounds To Survive. Best-of-three, then best-of-five, then best-of-seven.

The second half of the panel's read was the bracket. The WNBA seeds its top eight by record regardless of conference, and the opening round is a best-of-three. Only the Finals got longer in 2025, when the league expanded them from best-of-five to best-of-seven. A three-game series is short enough that a hot shooting night and a bad whistle can end a 31-win season, and Minnesota had to survive that, then a best-of-five, then a best-of-seven. Opus multiplied those out to about 42% and then shaved it to 38 for two reasons it names below. Kimi ran the same chain and landed at 39.

The whole gap fit in one sentence. Forty-nine cents was close to what you would pay if the best regular season in the league bought you a title, and three elimination rounds were the price of finding out. Minnesota did not get through the first one.

Golden State Was The Quietest Argument On The Board

On August 22 the Valkyries were the second-best team in the WNBA by record at 25-11, they held a plus-5.9 point differential, and they cost 8 cents, which was the largest disagreement on the board once you scaled for price. Every seat but one put them between 12 and 14, with Opus just below at 11.5, and the case was almost entirely structural.

As the records stood that day, the projected bracket ran 1 Minnesota against 8 Dallas, 2 Golden State against 7 Washington, 3 Las Vegas against 6 New York, and 4 Indiana against 5 Atlanta. Golden State drew the weakest qualifier in the field, a Washington team that was 21-15 with a negative point differential, and sat on the opposite side of the bracket from Minnesota. Nobody else in the top four had a path that clean. The real bracket came out differently, and The Record below walks through it.

The counterargument was real and every model named it. Golden State was scoring 82.5 points a game, the lowest average of the top four seeds, and its leading scorer was Gabby Williams at 14.4 points a night. The worry was that no one on the roster could be handed the ball in a tie game against a set playoff defense.

Golden State gets the softest draw (Washington, -0.7 differential — a record that flatters it) and a clean semifinal path, but an 82.5-points-for offense with a 14.4 ppg leading scorer has the lowest ceiling among the top four. — Kimi

Fable and GPT were the high seats at 14, and Fable made the same trade explicitly, calling the offense "the kind that stalls against set playoff defenses" while paying for the path anyway. Golden State had also lost two in a row, at home to Minnesota and then in Chicago against a 15-22 team, which was the sort of week that keeps a price at 8 cents while a bracket quietly improves.

What A One-Cent Spread Actually Costs You

Here is the part of this board almost nobody prices correctly, and it has nothing to do with basketball.

Kalshi lists these 15 teams as 15 separate yes-or-no contracts, each with its own buy price and sell price. On August 22, buying every one of the 13 live contracts at the ask cost $1.11 to hold something that will pay exactly $1.00. Selling every one at the bid collected 97 cents. That 14-cent gap between the two sides was the cost of doing business on this board, and it was not spread evenly.

The last column is simply the panel's number minus the ask, in percentage points. A positive figure means the models valued the contract above what a buyer paid for it that day.

Team	Bid	Ask	Spread as share of the ask	Panel	Panel minus ask
Minnesota Lynx	48¢	49¢	2%	37.4%	-11.6 points
Las Vegas Aces	13¢	14¢	7%	13.6%	-0.4 points
Golden State Valkyries	7¢	8¢	13%	12.9%	+4.9 points
Indiana Fever	13¢	14¢	7%	12.1%	-1.9 points
Atlanta Dream	8¢	10¢	20%	9.9%	-0.1 points
New York Liberty	6¢	7¢	14%	8.9%	+1.9 points
Washington Mystics	1¢	2¢	50%	2.3%	+0.3 points
Dallas Wings	1¢	2¢	50%	2.0%	+0.0 points
Portland Fire	0¢	1¢	100%	0.3%	-0.7 points
Chicago Sky	0¢	1¢	100%	0.3%	-0.7 points
Phoenix Mercury	0¢	1¢	100%	0.1%	-0.9 points
Los Angeles Sparks	0¢	1¢	100%	0.0%	-1.0 points
Toronto Tempo	0¢	1¢	100%	0.0%	-1.0 points

The same one-cent gap was 2% of Minnesota's price and 100% of Toronto's, and Toronto could not win this title at any price. The bottom two rows of that table were the two eliminated teams still listed, where a penny bought a contract the panel scored at zero. So the further down the board you go, the bigger the percentage disagreements look and the less any of them survive contact with what you would actually pay. Washington was the clearest example. The panel priced the Mystics 58% above where the market had them, which sounded like the biggest disagreement on the page, and then you paid 2 cents for something the models valued at 2.3 and the entire argument was worth three tenths of a point. The Mystics went out in two games to Atlanta and the contract settled at zero, so the argument was worth exactly what it looked like.

Golden State was the one row on this board where the panel-market gap was still large after the spread was subtracted. Whether a gap survives its own transaction cost is a different question from how big the gap looks, and neither the price alone nor the model alone answers it. You have to hold the two numbers next to each other, at the same second, on the same screen.

Dark Horses The Panel Wouldn't Dismiss

Below the top four, three teams drew real numbers from at least one seat for a named reason. Here is what the seats said in August, and what happened.

• New York Liberty (8.9%, 7¢ Ask). The Liberty's 22-15 record was posted while Sabrina Ionescu missed 13 games, Satou Sabally missed 24, and Leonie Fiebich missed 23. They also beat Las Vegas 93-85 in the Commissioner's Cup final on June 30. Every seat agreed the record understated the roster and every seat only half-credited it, because the healthy version had not taken the floor. As of the league injury report on August 22, Sabally was out with a concussion, Jonquel Jones was day to day with an ankle, and Fiebich still carried an out tag for a foot injury, though both Jones and Fiebich said on Friday they expected to play the next day. DeepSeek was the high seat at 11, calling New York "the board's biggest wildcard." Fiebich was back for the first round, Sabally was not, still out with the concussion, and the 8 seed swept the 1 seed anyway. New York's ask went from 7 cents to 22.
• Washington Mystics (2.3%, 2¢ Ask). The weakest profile in the projected field and the panel still would not zero them. Washington was the only playoff qualifier with a negative point differential at the time, but the roster was fully healthy and three players averaged at least 15 points: Sonia Citron at 16.9, Shakira Austin at 16.6 with 9.5 rebounds, and Kiki Iriafen at 15.1. Gemini was the high seat at 3. Washington finished 28-16 as the 5 seed, lost 92-77 and 93-75 to Atlanta, and the contract settled NO.
• Dallas Wings (2.0%, 2¢ Ask). Paige Bueckers was averaging 20.3 points and 5.8 assists, Jessica Shepard was pulling 11.1 rebounds a night, and the Wings held the last playoff spot. They were also 3-7 in their last ten, and Azzi Fudd's rookie season was over after a procedure on her right knee. The projected reward for the 8 seed was a best-of-three against Minnesota. Fable and Gemini both paid 3. Dallas finished 27-17 as the 7 seed, drew Golden State instead, and is tied 1-1 after Arike Ogunbowale scored 45 in a 108-100 overtime win in Game 2.

Where The Panel Changed Its Mind

After the first round of scoring, each model read the other seven boards with the names stripped off and could revise. Several did, and the biggest revisions all ran in the same direction.

Sonnet opened at 55.6 on Minnesota, the highest number anyone gave any team, and came back at 34 after reading the room:

This is a one-favorite board with a bunched second tier, not five long-shots behind a king. I cap MIN at 34 rather than higher because three rounds, starting with a best-of-three, is real variance even for a great team. — Sonnet

GLM made the other big move, dropping Minnesota from 49 to 37.5 and nearly doubling Las Vegas from 8 to 14.6:

Vegas gets 14% on Wilson's dominance and championship pedigree, but their +3.5 differential signals a .600 team, not a juggernaut. — GLM

DeepSeek went the other way, raising Minnesota from 33 to 44 and finishing as the most bullish seat on the board:

The Lynx's 31-7 record and +8.8 point differential, built largely without Napheesa Collier (19.0 ppg in only 10 games), signal a historic team. Collier's current 'Rest' designation points to a healthy playoff return, making them even stronger. — DeepSeek

Opus went up too, from 31.1 to 38, but stopped short of the group and said exactly which part of its reasoning it would not buy:

I haircut to 38 for two things the group waved past: Collier missing 28 games is an injury history, not a rest plan, so I credit only ~70% of her upside; and Courtney Williams was listed Out (lower leg) on the fetch date, which costs Minnesota its secondary playmaker. — Opus

What Would Change The Panel's Mind

Every seat was asked what would move its number. Four triggers came back repeatedly, each one a checkable event rather than a feeling, and The Record grades all four.

• Collier's Rest Designation Becomes A Real Absence. The league injury feed listed her out for rest as of August 20. If that turns into a postseason problem, GLM said Minnesota drops toward 28 and DeepSeek said it loses 10 points or more, redistributed to Golden State, Las Vegas and New York.
• Sabally Clears Concussion Protocol And Fiebich Returns Before The Bracket Locks. All eight seats tied New York's number to this directly. Opus put the healthy Liberty at 12 to 13 and said the money comes mostly out of Las Vegas, because the 3-versus-6 series flips toward a coin flip.
• Caitlin Clark Or Aliyah Boston Misses Time Beyond Day To Day. Both were day to day on August 22, Clark with a back issue and Boston with a lower-leg issue. Kimi said Indiana falls toward 7 and Atlanta rises, since the two meet in a 4-versus-5 series where their point differentials are separated by a tenth of a point.
• The Seeding Order Changes Over The Last Two Weeks. Atlanta had eight games left and Indiana seven. If Atlanta passes Indiana, or Washington falls out of the field, the collisions change. Opus said Golden State drops under 10 if it draws a healthier 7-seed, and that separating Indiana from Atlanta lifts both by two or three points.

When This Settles

Regular Season	Ended Thursday, September 24, 2026. Final top eight by record: Minnesota 33-11, Golden State 32-12, Las Vegas 31-13, Atlanta 30-14, Washington 28-16 (tiebreaker over Indiana), Indiana 28-16, Dallas 27-17, New York 26-18.
First Round	Best-of-three. New York over Minnesota 2-0 and Atlanta over Washington 2-0 are done. Las Vegas-Indiana and Golden State-Dallas are 1-1, with Game 3 on Thursday, October 1 in Las Vegas and Friday, October 2 in San Francisco.
Semifinals	Best-of-five. The bracket sends the Minnesota-New York winner against the Atlanta-Washington winner, so Atlanta and New York meet in one semifinal. The other is Golden State or Dallas against Las Vegas or Indiana.
Finals	Best-of-seven, and the contract pays out once the title is decided
Contract Expiration	Expected December 1, 2026; latest possible November 30, 2028
As Of	October 1, 2026, shortly before 9 p.m. ET

This page is re-read as the bracket moves, and The Record is where the grades land.

The Record

Nine of the 15 contracts on this board are graded, all NO, and only one of them was a real disagreement: Minnesota, 49 cents on the ask against 37.4% from the panel, settled at zero. The timeline in one sentence: Seattle and Connecticut were graded on August 17, before the panel scored the field; Chicago, Los Angeles, Phoenix and Toronto together on August 24 at 10:15 a.m. ET, two days after the pull above; Portland on August 26; then nothing for a month until the first round took Minnesota at 11:16 p.m. ET on Tuesday, September 29, and Washington at 10:27 p.m. ET on Wednesday, September 30. Seattle and Connecticut settled before this board existed and have no ask or panel number to compare, so they sit outside the table.

Team	Aug. 22 ask	Panel	Settled
Minnesota Lynx	49¢	37.4%	NO, September 29
Washington Mystics	2¢	2.3%	NO, September 30
Portland Fire	1¢	0.3%	NO, August 26
Chicago Sky	1¢	0.3%	NO, August 24
Phoenix Mercury	1¢	0.1%	NO, August 24
Los Angeles Sparks	1¢	0.0%	NO, August 24
Toronto Tempo	1¢	0.0%	NO, August 24

Six of those seven rows say almost nothing about who reads basketball better. On every contract from Washington down, the market and the panel were never more than a point apart, so a NO shows only that both sides could read a standings page. The honest way to weigh a graded record is by how far the two sides disagreed, and the Minnesota row, 49 cents on the ask, about 47% once the board is rescaled, against 37.4%, was the only double-digit gap on the page.

Be precise about what that grade says. The panel did not call the Lynx a loser. It had them first on all eight seats, and a team the panel gives 37% loses the title 63% of the time on the panel's own terms. What the panel called was the price: 49 cents for a team that still had to win a best-of-three, a best-of-five and a best-of-seven. The ledger also holds an earlier read of the same contract, from July 26, when the market was 42 cents and the panel's middle number was 30. Twice, with prices hidden, eight models put the best team in the league well under what traders were paying, and the contract paid nothing. That is a grade in the panel's favor, not proof of its number.

The team that beat them is the one this page half-credited. New York finished 26-18, the 8 seed, after the season the Dark Horses section describes, with Ionescu, Sabally and Fiebich missing 60 games between them by August 22. Every seat said the record understated the roster, and every seat stopped at 8.9% because the healthy version had not played together. It still had not when the series started, because Sabally remained out with the concussion. The rest of it was enough. Breanna Stewart scored 34 with 12 rebounds in Game 1, a 91-75 win in Minneapolis on Sunday, September 27. Jonquel Jones went for 21 and 13 in Game 2, an 87-71 win in Brooklyn on Tuesday, and Ionescu and Fiebich played 28 minutes or more in both games. Napheesa Collier played 33 minutes in each and scored 12 and 14. The first trigger the seats named, Collier's rest designation turning into a real absence, never fired. The second, written as Sabally clearing protocol and Fiebich returning, half-fired: Fiebich played 29 minutes or more in both games, Sabally sat both, and the Liberty swept anyway.

The Lynx contract traded at 44 cents at tip-off of Game 1 and 23 cents at tip-off of Game 2, and settled at zero that Tuesday night. New York's last trade before Game 1 printed at 5 cents. It is 21 bid, 22 ask tonight.

Washington's grade is the small one. The Mystics finished 28-16, the 5 seed, and lost 92-77 and 93-75 to Atlanta with Sonia Citron scoring 25 in both games. The panel's 2.3% against a 2-cent ask was the row this page used to show how a big-looking percentage gap shrinks to almost nothing once you pay the ask. It did.

The other two triggers: Aliyah Boston scored 18 in Indiana's Game 1 and Caitlin Clark 27 in its Game 2 win over Las Vegas, so neither absence materialized. The seeding did change, and by more than the seats imagined. Atlanta passed Indiana and finished 30-14 as the 4 seed, Washington climbed from the projected 7 to the 5, and New York and Dallas, projected 6 and 8 on the August records, finished 8 and 7. Golden State's softest draw turned into Dallas instead of Washington, and that series is tied.

Here is the open board, pulled shortly before 9 p.m. ET on Thursday, October 1, before Game 3 of Las Vegas-Indiana tipped.

Team	Aug. 22 ask	Panel (Aug. 22)	Oct. 1 bid / ask	First round
Atlanta Dream	10¢	9.9%	29¢ / 30¢	Beat Washington 2-0
Las Vegas Aces	14¢	13.6%	21¢ / 22¢	Tied 1-1 with Indiana, Game 3 Thursday
New York Liberty	7¢	8.9%	21¢ / 22¢	Beat Minnesota 2-0
Golden State Valkyries	8¢	12.9%	15¢ / 19¢	Tied 1-1 with Dallas, Game 3 Friday
Indiana Fever	14¢	12.1%	8¢ / 10¢	Tied 1-1 with Las Vegas, Game 3 Thursday
Dallas Wings	2¢	2.0%	1¢ / 2¢	Tied 1-1 with Golden State, Game 3 Friday

The new market favorite is a team neither side had in its top three. Atlanta was 10 cents in August against 9.9% from the panel, within a tenth of a point of its price, and trades at 30 cents on the ask now with the cleanest path left: two games against Washington, done, and a semifinal against New York, already set. Those two asks add to 52 cents, and the four teams on the other side of the bracket add to 53, so the market has the two halves of the bracket at a coin flip. Golden State is the row to watch for this page's sake. It was the panel's largest disagreement with the market after the spread, 8 cents against 12.9%, and it trades at 19 cents on the ask tonight with its series tied and Game 3 at home on Friday. The ask has gone through the panel's number, mostly because seven teams have left the field since August, and the grade waits on Friday night.

How To Read These Prices

• A Price Is Not A Probability Until You Rescale It. Because each team is its own contract with its own spread, the midpoint prices on the August board added up to 104 cents rather than 100, and the asks added up to 111. Minnesota's 48.5-cent midpoint rescaled to about 47%, which was the number worth comparing to the panel's 37.4%. The six open asks tonight add to 105 cents, so the same haircut applies to the bracket halves above.
• The Models Never Saw A Price. Each seat worked from the same fetched card of standings, point differentials, scoring leaders, availability and the league injury feed, then split 100 points across the field. Where the panel and the market disagreed, the gap was a question worth investigating, not an answer, and The Record is where the questions get answered one settlement at a time.

The Bottom Line

The live question is Golden State at 19 cents. The panel's 12.9% was scored against a 13-team field, and seven of those teams are gone, so the August number is not a live read on a six-team bracket. It is a record of where the disagreement was. What Friday's Game 3 at Chase Center decides is whether that disagreement ever gets graded on the Valkyries' terms or on Dallas's. Atlanta is the other half of the story: 30 cents on the ask with its semifinal against New York already set, and the two halves of the bracket priced at 52 and 53 cents, which is the market saying it cannot tell you which side the title comes from.

Minnesota was the best team in the WNBA, the panel said so on all eight seats, and the market paid 49 cents for a regular season that still had to survive three elimination rounds. It survived none of them. That grade is in. The other six land here when the contracts settle, and the panel re-scores the bracket once the semifinals are set.

Kalshi event contracts trade on a CFTC-regulated exchange. Kalshi admits traders at 18 and older in eligible states, and prices move.

Prices on this page are Kalshi's book. If you also trade on Polymarket, code OS4 gets new users a $50 trading bonus on a $10 deposit — affiliate link; terms as stated by Polymarket; 18+, availability varies by state.

Hero illustration: OddsShopper, in the house collage style. WNBA arena photo by SusanLesch, licensed CC0; photos cropped, toned, and composited.

More on this: College Football Championship Odds: Texas Took Over Late · CeeDee Lamb Receiving Yards Tonight: Don't Buy a Line · Buccaneers vs Cowboys Polymarket Picks: 0-4 Tampa Is Live · NFL Week 5 Anytime TD Picks: One Didn't Survive The Sims · Odds To Win Champions League 2027: The Models Won't Pay Barcelona