Updated August 21, 2026 · 15 min read · by Eric Lindquist
Dallas finished 7-9-1 last season and allowed more points than any team in the NFL. On Kalshi's NFC East board, that team trades within four cents of the reigning division champion. Washington, which won five games with its starting quarterback watching ten of them from the sideline, trades at exactly the same price as a New York Giants team that lost thirteen. We handed all four teams to eight AI models with every price covered up, and the numbers that came back argue with this board in both directions at once.
The Quick Answer
Kalshi prices the 2026 NFC East at Philadelphia 39.5 cents, Dallas 36, Washington 14.5 and the New York Giants 14.5, which reads as a two-team race between the Eagles and a Cowboys club that finished under .500. Our eight-model panel, which scored the division without being shown any of those prices, landed at Philadelphia 37, Dallas 31, Washington 20 and the Giants 12. The two disagreements point opposite ways, and the strongest argument for the Eagles turns out to be a 24-year pattern in this division that belongs to exactly one franchise.
New to Kalshi event contracts? Start with our plain-English walkthrough of how to bet the NFL on Kalshi.
Prices on this page were read from Kalshi on the evening of August 21, 2026, three weeks before kickoff, and they refresh as the story moves; the dated stamp on the market card below carries the evidence.
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.
The Board
A Yes contract pays one dollar if that team wins the NFC East, so 36 cents reads as roughly a 36 percent chance. "Panel" is the blended number the models produced: the median of eight independent estimates, taken after each model had read the other seven and had one chance to revise.
| Team | 2025 finish | Kalshi | Panel |
|---|---|---|---|
| Philadelphia Eagles | 11-6, won division | 39.5¢ | 37% |
| Dallas Cowboys | 7-9-1, second | 36¢ | 31% |
| Washington Commanders | 5-12, third | 14.5¢ | 20% |
| New York Giants | 4-13, fourth | 14.5¢ | 12% |
Start with Washington's row, because it is the one the board gets least defensible. Kalshi prices the Commanders and the Giants at the identical 14.5 cents. Those two teams did not have the same season, do not have the same quarterback situation, and did not have the same offseason. The panel puts eight points between the two, on a board that prices them the same to the cent.
One more thing to know before you read a single price as a probability: the four Kalshi numbers add up to 104.5 cents, not 100. That extra 4.5 cents is the spread, the built-in cost of running a two-sided market, and it means every price above is a shade rich by construction. Strip it out and both gaps survive.
Every Seat's Number
All eight seats scored this market. Each number is that model's own estimate that the team wins the division, taken after a revision round in which every model read the others' anonymized reasoning. "Edge" is the panel median minus the Kalshi price: positive means the models think the board is too low.
| Team | Kalshi | Fable | Opus | Sonnet | ChatGPT | Gemini | GLM | Kimi | DeepSeek | Panel | Edge |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Philadelphia Eagles | 39.5¢ | 37 | 41 | 37 | 34 | 37 | 33 | 38 | 36 | 37% | -2.5 |
| Dallas Cowboys | 36¢ | 31 | 27 | 30 | 36 | 32 | 36 | 31 | 30 | 31% | -5.0 |
| Washington Commanders | 14.5¢ | 20 | 19 | 20 | 20 | 19 | 19 | 20 | 22 | 20% | +5.5 |
| New York Giants | 14.5¢ | 12 | 13 | 13 | 10 | 12 | 12 | 11 | 12 | 12% | -2.5 |
Every seat on this panel is graded against real market settlements — records to date: Gemini 86% on 4,292 graded calls · GLM 81% on 2,367 graded calls · Kimi 83% on 2,345 graded calls · DeepSeek 80% on 2,392 graded calls. Recomputed daily; the full scoreboard is public.
These are model estimates, not predictions of fact and not financial advice. Every number in this piece gets graded in public once the market settles; the running tally lives on the full graded scoreboard.
More live boards from the same panel: NFC West odds (Rams 48.5¢, Seahawks 28.5¢) · NFL playoff odds (Eagles 59¢ to make the field) · Super Bowl 2027 odds (Eagles 4.5¢). Prices fetched August 21, 2026.
Read the Washington column left to right and you find the tightest agreement anywhere on this board. Every seat landed between 19 and 22. Eight models built from different training data, run through different companies' infrastructure, converged inside three points on a team the exchange has priced as an afterthought.
The Market
| Venue | Kalshi, a CFTC-regulated event-contract exchange (18+; availability varies by state, as of August 2026) |
| The Event | NFC East Division Winner, 2026-27 season. One yes/no contract per team, four in total, mutually exclusive |
| The Contract | Resolves Yes for the team that wins the NFC East, per Kalshi's rules text: "If Philadelphia wins the Pro Football NFC East Division, then the market resolves to Yes" |
| Decided By | The league's declared division title, which comes down to regular-season record and then the NFL's published tiebreaker sequence: head-to-head, division record, common games, conference record, and on down |
| Settlement Sources | Kalshi's achievements contract terms name the governing league plus the Associated Press, ESPN, The Wall Street Journal and Fox Sports, and specify that revisions made after expiration are not accounted for |
| Expected Expiration | January 11, 2027 at 10:00 AM ET, one day after the final Sunday of the regular season; the market can close early once a title holder is declared |
| Prices Above As Of | the evening of August 21, 2026. Quoted spreads were one cent on three teams and two on Dallas |
There is one settlement quirk worth naming, because the contract language allows something the sport does not. Kalshi's template covers every sports achievement it lists, and it says that if multiple participants are reported as the result, Yes holders split the dollar. In practice that clause never fires here: the NFL names exactly one division champion every season, and the tiebreaker sequence runs deep enough to resolve any tie in the standings. Four numbers, one winner, no split.
Dallas Cowboys: The Best Offense Behind The Worst Defense
The Cowboys are the reason this board looks strange, and the win-loss record is the last place to look for why.
Dallas scored 471 points in 2025, seventh-most in the NFL. Dallas allowed 511, the most of any team in the league, 30.1 per game. No other team in the NFL paired a top-10 scoring offense with a bottom-eight defense last season; Dallas was the only one. The next-worst defense belonged to the New York Jets, who allowed 503 while scoring only 300, the minus-203 of a team that was bad at both ends. Dallas's minus-40 point differential is the arithmetic of a very good offense being dragged behind a defense that could not get off the field, which is a different animal from a team that simply is not good.
Here is the same split across all four teams, and it is the table the standings column hides.
| Team | Points scored (rank) | Points allowed (rank) | Net |
|---|---|---|---|
| Dallas Cowboys | 471 (7th) | 511 (32nd) | -40 |
| New York Giants | 381 (17th) | 439 (26th) | -58 |
| Philadelphia Eagles | 379 (19th) | 325 (5th) | +54 |
| Washington Commanders | 356 (22nd) | 451 (27th) | -95 |
2025 final regular season, all ranks out of 32.
That distinction is the whole argument, and every seat on the panel made some version of it. A negative differential driven by defense is the fixable kind, because a defense can be rebuilt in one offseason in a way an offense usually cannot. Dallas spent the offseason doing exactly that: edge rusher Rashan Gary, safeties Jalen Thompson and P.J. Locke, cornerback Cobie Durant, first-round safety Caleb Downs at No. 11 and edge Malachi Lawrence at No. 23, with Christian Parker taking over as coordinator and the front moving toward a 3-4, per PFF's 2026 Cowboys preview. Add a three-year, $105.9 million extension for Quinnen Williams on August 10, per Ian Rapoport of NFL Network.
The core of the offense that produced those 471 points is still in place. Dak Prescott, CeeDee Lamb and George Pickens are all on the current roster, and Klayton Adams is the only offensive coordinator in the NFC East who kept his job this offseason. Brian Schottenheimer, in his second year as head coach, keeps play-calling. And Dallas already went 4-2 inside the division last season, the best intra-division mark of the four, which matters more here than anywhere else because six of the twelve games that decide this title are Cowboys games, and the first tiebreaker the NFL reaches for is head-to-head.
So why do the models still fade the price? Because the market is asking the defense to make a two-step jump. ChatGPT, which ended up as one of the panel's two highest seats on Dallas at 36, framed the ceiling honestly: "I do not need Dallas to become elite defensively; average-ish defense likely makes it a 10-11 win team." Opus, the panel's lowest at 27, priced the same fact as a compound bet rather than a single one, and its language is the sharpest on the board:
Asking a 32nd-ranked unit to reach average through a 3-4 conversion under a first-time coordinator and a year-two head coach is a two-step bet, not one. — Opus
Both readings are true, which is how the panel median lands at 31 with a seat range from 27 to 36: Dallas is the widest genuine disagreement inside the panel itself, which is a different thing from a disagreement with the market. When the models argue among themselves that much, the honest read is that the Cowboys are the highest-variance team on this board, not the safest fade.
Philadelphia Eagles: The Repeat Rate That Belongs To One Franchise
The case against the Eagles is a single number, and it is a good one. In the 24 NFC East seasons from 2002 through 2025, the defending champion repeated three times in 23 chances. Thirteen percent. All eight seats cited it in their first pass, and two of them, Gemini and DeepSeek, used it to put Dallas ahead of the Eagles outright.
Then, in the revision round, one model went and looked at whose repeats those were.
All three belong to Philadelphia. The Eagles went back-to-back in 2003 and 2004, then again in 2025. Decompose the same 24 years by franchise and Philadelphia is 3-for-10 defending its own division title, while Dallas is 0-for-7, the Giants 0-for-3 and Washington 0-for-3, a combined 0-for-13. The 13 percent is an average of two completely different histories: a franchise that sometimes repeats, and three that never have inside this window.
The 13% number is an average dragged down by 7-9 and 9-7 champs from the other three franchises. Applying it as a flat discount to Philadelphia ... misreads a franchise-specific pattern as a division-wide law. That is where I most refuse to move. — Opus
Underneath the pattern is the mechanism. Philadelphia's plus-54 differential last season was built on defense, not offense: 325 points allowed, fifth-fewest in the league at 19.1 per game, carrying an offense that ranked 19th in scoring. Defense-driven seasons carry forward better than offense-driven ones, and this defense is largely intact. Vic Fangio is still the coordinator, Zack Baun wears the green dot per Fangio's July 30 announcement, and Jalen Carter signed a four-year, $152 million extension through 2031, the largest ever for a defensive tackle, per ESPN. Philadelphia also draws the softest 2026 schedule in the division, its opponents having gone .481 last season against Washington's .502.
The counterweight is the receiving corps, and it is real. Philadelphia traded A.J. Brown to New England after June 1 for a 2028 first-round pick and a 2027 fifth, per ESPN, splitting a $43.45 million dead-cap charge across two seasons. The replacements are rookie Makai Lemon, traded-for Dontayvion Wicks, and free agents Marquise Brown and Elijah Moore. As of this week the ESPN ledger has Lemon limited with a hamstring, Moore missing practice with a neck injury from Wednesday's joint session with New England, and DeVonta Smith limited with a hamstring, per Dave Zangaro of NBC Sports Philadelphia. Edge rusher Jonathan Greenard's pectoral injury has the club worried he misses the Week 1 division opener against Washington, per Mike Garafolo of NFL Network.
An 11-6 champion that got worse at the one position group it could least afford, and better nowhere obvious, is why the panel lands at 37 rather than 45. It still leads the board. It just does not lead it by the margin the standings imply.
Washington Commanders: The Team Priced Like It Lost 13 Games
Here is the gap, and it is the cleanest one on the board.
Washington finished 5-12 with a minus-95 point differential, the ugliest line in the division. It is also the least informative, because Jayden Daniels missed ten of those games with a sprained knee, a strained hamstring and a dislocated elbow, per NFL.com's division camp preview. He is healthy entering the season. A record built largely without a team's starting quarterback is a record about that quarterback's absence, and the market appears to be reading it as a record about the team.
The panel's agreement here is the striking part. Eight seats, 19 to 22, with a median of 20 against a 14.5-cent price. Nothing else on this board drew that kind of consensus.
Washington gets 20 because Daniels missing 10 games makes the -95 differential too harsh, and the defensive spending creates upside. But hardest schedule, Tunsil out, Newton out after pectoral surgery, Armstrong suspended, and both coordinators changing keep me below the more bullish Washington takes. — ChatGPT
The roster underneath supports it. Washington spent the offseason on defense: edge rusher Odafe Oweh on a reported four-year, $100 million deal, linebacker Leo Chenal, safety Nick Cross, nickel Amik Robertson and defensive tackle Tim Settle, plus linebacker Sonny Styles at No. 7 in the draft. On offense the club added tight end Chigoziem Okonkwo and running back Rachaad White. Dan Quinn is credited with eight seasons of NFL head-coaching experience, second-most in the division behind John Harbaugh.
Every seat also named the same three brakes, and they are worth stating plainly because they are what keeps 20 from being 26. Washington replaced both coordinators: Kliff Kingsbury is out, assistant quarterbacks coach David Blough was promoted to run the offense, and Daronte Jones took over the defense, per ESPN's camp preview. The ESPN injury ledger this week carries left tackle Laremy Tunsil out, edge rusher Dorance Armstrong suspended, and defensive tackle Jer'Zhan Newton out after pectoral surgery with no timeline, which head coach Dan Quinn confirmed on August 20. And the schedule is the division's hardest, its opponents having gone .502 last season.
Kimi opened highest on Washington at 26 and came down to 20 after reading the other seats, and its first-pass framing is still the cleanest statement of the bet: "Washington is my biggest disagreement with a naive 2025-record ranking." Opus, which held the Commanders third rather than fourth, named the same thing from the other side: "Washington's healthy Daniels is the biggest upside lever on the board." A 20 percent estimate against a 14.5-cent contract is what that reads like once the coordinator turnover and the injury ledger are taxed back out.
New York Giants: A Great Coach Walking Into An Injury Ledger
The Giants got the biggest name of the NFC East offseason and the panel's lowest number, and both facts are defensible.
New York hired John Harbaugh after 18 seasons in Baltimore, where he won 180 regular-season games, made 12 playoff appearances, took six AFC North titles and won a Super Bowl. He is, by a distance, the most accomplished head coach in this division: 18 seasons of NFL head-coaching experience against Nick Sirianni's five, Dan Quinn's eight and Brian Schottenheimer's one, per Yahoo Sports' offseason review of the hire. Jaxson Dart enters the season as the starter after a rookie year in which he played 12 games, missing two in concussion protocol, completed 63.7 percent of his throws for 2,272 yards at 6.7 yards per attempt with 15 touchdowns and five interceptions, and added 487 rushing yards and nine rushing touchdowns, per Pro Football Rumors.
What the panel could not get past is the cost of the rebuild running alongside the hire. New York traded Dexter Lawrence II, the best interior defensive lineman it had, to Cincinnati for the No. 10 pick in the 2026 draft, per NFL Network. Then the injuries arrived in clusters: cornerback Thaddeus Dixon, returner Gunner Olszewski and defensive lineman Roy Robertson-Harris each tore an Achilles in offseason work, and Malik Nabers has been held out with a knee issue, including this week's joint practices with Miami, per Jordan Raanan of ESPN. The Giants added Najee Harris to a reshuffling running back room this month, per Mike Garafolo of NFL Network.
The models still gave New York twelve percent rather than a rounding error, which is the right instinct on a four-team board where every team plays the other three twice. Sonnet was the seat that moved up on New York in the revision round, from nine to thirteen, and its first-pass reasoning is the case for the price: "Harbaugh is a real long-term upgrade, but this year pairs a sophomore QB with a traded-away Dexter Lawrence and an alarming camp injury cluster." Long-term upgrade, this year taxed. Twelve percent against a 14.5-cent price still makes New York the second team on this board the panel would rather sell than buy.
Where The Panel Changed Its Mind
The models scored the division once, then read each other's anonymized numbers and reasoning and scored it again. This board moved more between rounds than most, and the movement is the most useful thing on the page.
The clearest way to see it: in round one, Gemini and DeepSeek were the two seats with Dallas ahead of Philadelphia. In round two, both had moved off it, and two different seats, ChatGPT and GLM, had taken their place. The argument for Dallas changed hands rather than getting settled.
- ChatGPT Moved The Most Of Any Seat, 24 points of total revision across the four teams, taking Philadelphia down from 43 to 34 and Dallas up from 25 to 36 after reading the Cowboys case its peers had built.
- Gemini Moved Almost As Far In The Opposite Direction, cutting Dallas from 43 to 32 and lifting Philadelphia from 32 to 37, after other seats pushed back on treating the 13 percent repeat rate as a flat discount on the Eagles. Those two seats made the single largest team-level move in the round, 11 points each, on the same team, in opposite directions.
- DeepSeek Made The Biggest Move On Washington, from 14 to 22, while cutting Dallas from 40 to 30. Its second pass kept the Cowboys' defensive-rebuild case and reweighted what it was worth relative to the row below: "Washington's -95 is misleading because Jayden Daniels missed 10 games; he is healthy, and the Commanders spent heavily on defense."
- Opus Made The Biggest Lift On The Champion, from 35 to 41, on the franchise decomposition of the repeat rate above. Three other seats nudged Philadelphia up as well, none of them by more than five points. Its note on Dallas barely moved at all, from 28 to 27: minus-40 is still minus-40, and the Cowboys shed Micah Parsons a year ago.
What Would Change The Panel's Mind
Four checkable events, each with the direction it pushes the board.
- A Jayden Daniels Injury Of Any Kind. The entire Washington case is one player's availability. Any absence for Daniels drops the Commanders back toward their 14.5-cent price and pushes the difference into Philadelphia and Dallas. A healthy Daniels through October does the opposite and makes 20 look conservative.
- Dallas's Points Allowed Through The First Quarter Of The Season. The Cowboys must show that a 3-4 conversion under a first-time coordinator has produced something like an average defense. Give up 27 or more per game through Week 5 and the panel's 31 is generous. Hold opponents in the low twenties and the market's 36 cents was cheap.
- The Week 1 Division Doubleheader On September 13. Washington at Philadelphia in the afternoon and Dallas at the New York Giants at night. Head-to-head is the first tiebreaker the NFL applies, and all eight seats weighed the division-record column in their answers. Jonathan Greenard's availability for the Eagles is the specific injury to watch in that opener.
- Philadelphia's Receiving Corps In September. Makai Lemon, Elijah Moore and DeVonta Smith all carry dated injury notes this week. If the rebuilt group looks functional through the first month, the Eagles' defense-plus-easiest-schedule profile makes 37 look low. If Jalen Hurts is throwing to a patchwork group in October, the panel's discount was not deep enough.
Settlement Timeline
| Decided By | Regular-season record, then the NFL's published tiebreaker sequence |
| Season Opens For This Division | Sunday, September 13, 2026: Washington at Philadelphia (4:25 PM ET), Dallas at the New York Giants (8:20 PM ET) |
| Next Catalysts | Washington at Dallas, September 20 · Dallas at Philadelphia, Monday, October 26 · Philadelphia at Dallas on Thanksgiving, November 26 |
| Final Weekend | Sunday, January 10, 2027: Dallas at Washington and Philadelphia at the New York Giants, both division games in Week 18 |
| Kalshi Expected Expiration | January 11, 2027, 10:00 AM ET; early close permitted once a title holder is declared |
| Re-Scored | when the story moves; prices on this page are stamped as of August 21, 2026 |
Hottest Prediction Markets Right Now
- 2028 Democratic presidential nominee · $187M traded
- 2027 Pro Football Champion · $68M 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.
- NFC West Odds 2026: The Champion Is Priced Second
- The 72-Year Streak Traders Are Betting Against
- Trump's NASA Speech Is Priced For A Rally
- AFC West Odds 2026: The 14-3 Champion Trades Third
- 2026 Governor Races: Every Kalshi Board We Track
The Bottom Line
Kalshi and our panel agree on the top of this division and disagree about everything underneath it. Both make Philadelphia the favorite, and the strongest evidence for the Eagles sits underneath the 11-6 record: all three defending-champion repeats in this division since 2002 belong to Philadelphia, and none belong to the other three franchises combined. Where they part company is the middle. The board asks 36 cents for a Dallas team whose case rests on a historically bad defense becoming ordinary, and 14.5 cents for a Washington team whose case rests on a quarterback simply being available. The panel prefers the second bet to the first, by five and a half cents on one and five on the other, and it agreed with itself more tightly on Washington than on anything else here. Kickoff is three weeks out and both theses get their first real test in the same Sunday of division games. We will re-score this board in public as they do.
Prediction markets are for adults 18 and older and carry risk. These are model estimates, not predictions of fact and not financial advice.
