Giants Win Total 2026: A 4-13 Team Priced For 7.5 Wins

The New York Giants won four games last season. As of September 7, 2026, the Kalshi crowd prices their whole season at 7.49 wins. Fanatics Sportsbook posts the same team at 7.5 and charges the same either way. Their opener against Dallas is six days out. That is a market asking you to believe a four-win team is now an eight-win team, and it has been asking that all summer without much argument.

The strange part is that the argument it is having is not the one you would expect.

We put the Giants' season in front of nine AI models with every price hidden. They saw the roster, the schedule, last year's game log, the injury report and twenty-two seasons of what happens to teams shaped like this one. They did not see a cent. Their ladder crosses even money at 7.69 wins. The crowd's crosses at 7.72. Two independent methods, three hundredths of a win apart, which is about as close to agreement as forecasting gets.

Then look at the rung where the Giants win eleven games or more, and the two sides are not in the same building.

Updated September 7, 2026 · 19 min read · by Jake Hari

The Market Card

Kalshi does not post a single Giants win total. It lists 17 separate yes-or-no contracts, one for each threshold from "wins at least 1 game" up to "wins all 17." Each pays $1.00 if the team gets there. Read all 17 prices top to bottom and you have the crowd's entire opinion of the season rather than one number with a hook in it.

The venue. Kalshi is a CFTC-regulated event-contract exchange, 18+, and availability varies by state. On August 28, 2026 the Ninth Circuit ruled unanimously that Kalshi's sports contracts are likely not "swaps" under the Commodity Exchange Act. That leaves states free to enforce their gambling laws against them, and sets up a probable Supreme Court fight. Check what is legal where you live.

The market. Event ticker KXNFLWINS-27NYG. That is one contract for every threshold from one win to a perfect season.

What settles it. The contract terms define the underlying as "the number of regular season game wins," and add that "postseason games, exhibition games, or any games outside the regular season do not count." A tie is not a win. Two quirks worth knowing: the terms say that if "the season is shortened or lengthened for any reason, this will not affect the value of the Underlying," and that revisions made after expiration are not counted. Every contract lists the same expiration date, January 18, 2027, which is a little over a week after the final Sunday of the regular season.

How much money is on it. About 238,000 contracts have traded across the 17 thresholds and roughly 158,000 are still held open. Almost 89% of that open interest sits on six contracts, the band running from six wins to eleven. That is the market telling you where it thinks the season actually lands.

Prices as of September 7, 2026, 8:53 a.m. ET, best bid and best ask on the order book.

The sportsbook version. Our Giants win total guide tracks the sportsbook side of this number. Fanatics Sportsbook posted the Giants at 7.5 wins with both sides at 1.91 in decimal odds on August 13; FOX Sports had the same 7.5 at the same price on July 29. A sportsbook half-win line can never push, and it gives you exactly one number. The exchange's 17 thresholds give you the whole distribution, which is a read no book posts.

One translation for anyone new to the exchange. A contract priced at 46¢ means the crowd puts roughly a 46% chance on that outcome, and $100 of it returns about $217 if it pays and nothing if it does not. The bid is what a buyer will pay right now, the ask is what a seller wants, and the gap between them is what it costs you to get in and back out.

Four Wins On A Seven-Win Point Differential

Start with the number that makes this whole market interesting. The 2025 Giants scored 381 points and allowed 439. That is a bad team. It is not a four-win team.

Run those two totals through the standard Pythagorean formula, the arithmetic that converts points scored and allowed into the record a team "should" have had, and 381 against 439 comes out at 7.09 wins over 17 games. The Giants won four. The gap is 3.09 wins, and almost all of it lives in one column of the schedule: eight games decided by eight points or fewer, of which New York won one. In everything else they went 3-6, which is roughly what a minus-58 point differential looks like.

A 1-7 record in one-score games is the single most reliably temporary thing in football. So the obvious read is that the Giants were a seven-win team that got unlucky, and the market is now pricing the seven-win team.

That read is too fast, and our own ledger says so. We pulled every team season from 2004 through 2025, all 669 consecutive-season pairs, and split them three ways. Teams that finished 2 or more wins below their point differential rebounded hard the next year: median 8.0 wins, and 50% of them reached 8 or more. That is the bull case, and it is real.

But teams that were both unlucky and bad, five and a half wins or fewer on that same per-17 scale with a 2-win shortfall, which is precisely the Giants' profile, did much worse. There are 27 of them in 22 seasons. Their median next season was 7.0 wins. Only a third reached 8. Only 30% reached 9. Seven of the 27 fell back to five wins or fewer.

The mechanism is not complicated. Being two wins short of your point differential is a good sign when the differential was good. When the differential was already bad, all the regression buys you is the mediocrity you actually earned.

The list of 27 is worth a minute, because the extremes are instructive. Five of them reached double digits the next year: the 2005 Buccaneers, 2008 Dolphins, 2012 Vikings, 2017 Jaguars and 2024 Chargers. Three of those five changed head coaches; two, Tampa Bay under Jon Gruden and Minnesota under Leslie Frazier, kept the coach they had and got better anyway. At the other end, the 2017 Browns went 0-16 the year after a 1-15 season, and one of the 27 is the Giants themselves: the 2018 team went 5-11 on a 7.4-win differential and then went 4-12.

That is the honest prior. The Giants are in a bucket whose middle outcome is seven wins, whose upside is real but rare, and whose downside is not a hypothetical.

What Actually Changed On The Sideline

The market is not pricing a repeat of last season, and the reason has a name.

On January 6, 2026, Baltimore fired John Harbaugh after an 8-9 season. Two weeks later, on January 20, the Giants hired him as the 21st head coach in franchise history on a five-year deal. He arrives with 180 wins and 113 losses across 18 regular seasons, a .614 rate, and 13 postseason wins. His first Baltimore team went 11-5 and reached the AFC Championship Game. He brought Matt Nagy to run the offense, Dennard Wilson the defense and Chris Horton special teams.

If you believe one-score games are partly coachable, this is the hire that addresses the exact leak. Eight of the nine models made some version of that point. The ninth named it only to say it was not enough.

The roster moved in two directions at once. New York traded defensive tackle Dexter Lawrence to Cincinnati in April, after he requested out and after his sack total fell from nine in 2024 to half a sack in 17 games in 2025, and got back a package including the No. 10 overall pick. That left the Giants the only team in the 2026 draft with two top-ten selections, and they spent them on Ohio State linebacker Arvell Reese and Miami offensive tackle Francis Mauigoa. A rebuilt front seven around Brian Burns, Abdul Carter, Kayvon Thibodeaux, Tremaine Edmunds and Reese is the best unit on the team. The interior it plays next to is DJ Reader at 32 and Shelby Harris at 35, with Roy Robertson-Harris on injured reserve.

Quarterback is the pivot. Jaxson Dart threw for 2,272 yards with 15 touchdowns and 5 interceptions as a rookie, at 6.7 yards per attempt and a 91.7 rating, and added 487 rushing yards and 9 rushing touchdowns. He was also sacked 35 times. He is now in his second year and his first inside Nagy's system, and his entire preseason was two series against Minnesota. Harbaugh's read: "I think he's come a long way. He looks really comfortable in what we're doing right now."

Then there is the availability report, which is where the season's variance lives. Malik Nabers tore the ACL in his right knee against the Chargers last September, had reconstructive surgery and a later scar-tissue procedure, has practiced through camp and has shed the red non-contact jersey, and is still listed questionable with neither he nor the club committing to Week 1. Calvin Austin III tore an ACL in practice and is on injured reserve. Four cornerbacks are on injured reserve, and four more of the twelve on the roster, including Paulson Adebo, Greg Newsome II and Dru Phillips, are listed questionable. Center John Michael Schmitz Jr. is in concussion protocol. Mauigoa left a joint practice with a stinger.

The schedule, for once, is not the villain. The Giants' 17 opponents went a combined 144-145 last season, a .498 win rate, which is dead average. It is top-heavy rather than hard: seven games against 2025 teams that won 11 or more, and five against teams that won five or fewer. Nine home games, eight road, bye in Week 8. Every model was asked to name the games that decide the number, and the lists ran from five games to nine. All nine named the same two fixtures: both Washington games and Cleveland at home.

The Board

Here is the whole ladder. The Kalshi column is the mid-point of the best bid and best ask, fetched September 7. The panel column is the median of nine independent models after they read each other's reasoning and revised, which is the blend this page uses throughout.

Giants win at least	Kalshi	Panel	Difference
1 game	99.5¢	99.7%	+0.2
2 games	99.5¢	98.9%	-0.6
3 games	97.5¢	96.9%	-0.6
4 games	88.5¢	92.9%	+4.4
5 games	83.5¢	86.0%	+2.5
6 games	71.5¢	74.9%	+3.4
7 games	59.0¢	61.2%	+2.2
8 games	46.5¢	44.9%	-1.6
9 games	34.5¢	30.3%	-4.2
10 games	26.0¢	18.8%	-7.2
11 games	19.0¢	10.8%	-8.2
12 games	12.0¢	6.2%	-5.8
13 games	7.5¢	3.2%	-4.3
14 games	2.5¢	1.3%	-1.2
15 games	1.0¢	0.5%	-0.5
16 games	0.5¢	0.2%	-0.3
17 games	0.5¢	0.1%	-0.4

Model estimates, not predictions of fact and not financial advice. Every number in this piece gets graded in public once the market settles, on the full graded scoreboard.

Read the table as a shape rather than a list and the disagreement is obvious. Across the four rungs where the season is actually decided, four wins through seven, the panel is more confident than the market on every one. From eight wins up it is less confident on every single rung, and the gap widens all the way to eleven before the numbers get too small to matter.

Add the 17 prices up and you get expected wins, which is what a complete ladder is arithmetically for. The crowd's ladder sums to 7.49 at the mid-point, 7.32 at every bid and 7.66 at every ask. The median model landed at 7.33. That is a sixth of a win apart, on a season that is 17 games long.

So the two sides agree on how many games the Giants win and disagree completely about which seasons produce that average. The market keeps a real tail: 26¢ that this team wins ten or more, 7.5¢ on thirteen or more. The panel's version of the same average is narrower and duller. It puts 16.3% on exactly seven wins, the single heaviest outcome anywhere on the board, and it prices those same two outcomes at 18.8% and 3.2%. The crowd is buying a team that could be good. The models are buying a team that will be fine.

The rungs at the edges deserve a caveat. The 15, 16 and 17-win contracts have never traded at all, and the 1 and 2-win contracts have taken 436 contracts between them all year. The 10 and 11-win rungs are 4¢ wide, bid to ask. Anything outside the 4-to-13 band is a quote more than a market.

Prices did move recently, and in one direction. Seven rungs last printed below their prior session's mark and only one printed above: 9 wins went from 39¢ to 34¢, 10 wins from 29¢ to 24¢, 11 wins from 22¢ to 18¢, with 4, 7, 8 and 14 wins also lower and 6 wins a cent higher. On thin rungs a few cents can be one participant re-quoting rather than a real move, but the 9-win contract has 61,700 contracts of history behind it, and the whole top of the ladder came down together. Against our own reading of this ladder on September 2, when it cost 7.26 to 7.63 wins, the mid-point is essentially unchanged. What has softened is the ceiling, not the season.

More live boards from the same panel: all 32 NFL win totals, where the Giants were one of the closest market-panel matches on the board · Super Bowl LXI, where the Giants trade at 1¢ bid, 2¢ ask · the worst-record contract, where they are 3¢ bid, 4¢ ask. Prices fetched September 7, 2026.

Hottest Prediction Markets Right Now

• 2028 Democratic presidential nominee · $212M traded
• 2027 Pro Football Champion · $83M traded
• Pro Baseball Champion · $68M traded
• 2028 U.S. Presidential Election winner? · $65M traded
• 2028 Republican presidential nominee · $63M traded
• Fed decision in Sep 2026? · $37M 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.

Every Seat's Number

Nine models scored this season. Each committed to a full distribution over all 18 possible win totals before seeing any price or any other model's answer, then read the other eight anonymously and revised or held. Eight revised or held in the second round; one seat did not return in time, so its first-round number stands.

Seat	Projected wins	Chance of 8+	Reason, in one line
Fable	7.49	48.2%	Regression plus Harbaugh beats the 7.0 median; Nabers carries half a win of it
Grok	7.45	47.1%	Dart already posted a 91.7 rating, so this is not a quarterback-void rebuild
Sonnet	7.41	47.0%	Harbaugh targets the one-score leak, but the front is thinner than 2025's
Kimi	7.39	44.6%	Regression lifts the floor, not the ceiling
Opus	7.33	46.2%	Regress 80% of the shortfall, then cap it on the interior line
ChatGPT	7.20	44.9%	Bad, not four-win bad; a nine-game swing set decides it
GLM	6.99	38.8%	Partial regression, thinned right tail, floor held by the edge rushers
Gemini	6.74	36.5%	Defensive talent has degraded below last year's differential
DeepSeek	5.06	14.0%	Year-2 quarterback, new scheme, secondary in pieces
Panel median	7.33	44.9%	Kalshi's ladder implies 7.49 wins

Every seat on this panel is graded against real market settlements — records to date: Gemini 87% on 7,680 graded calls · Grok 85% on 6,396 graded calls · Kimi 86% on 3,836 graded calls · GLM 84% on 3,557 graded calls · DeepSeek 84% on 3,630 graded calls. Recomputed daily; the full scoreboard is public. Figures as of September 7, 2026.

Eight of the nine landed between 6.74 and 7.49 wins. DeepSeek is the lone dissent at 5.06, and its case is the one everyone else had to answer: "Second-year QB Dart in a new scheme, a decimated secondary, and an uncertain Nabers knee on a schedule with 7 top-tier opponents caps the ceiling below the historical median." Nobody bought it, for a reason Opus put bluntly in the second round: DeepSeek "states the differential implies 6.5-7.0 true-talent wins and that 1-7 is a random-number generator, then prices P(6+) at 40%," which is below what even ordinary five-win teams have managed.

Now read the record line against this page rather than for it. Those percentages count every market a seat has graded, including the many where model and crowd agreed. When the two disagree, our ledger of 5,608 settled markets says the crowd usually wins: at a gap of 10 points or more the market has been right 68% of the time, at 20 points or more 74%, at 30 or more 81%. This board's widest gap is 8.2 points, on the 11-win contract, which lands in the 5-to-10-point bucket where the market's record is its weakest, 59%. That is the most this page will claim: a modest disagreement, in the bucket where the crowd's historical advantage is smallest, and still a bucket the crowd wins more often than not.

Where The Panel Changed Its Mind

Five of the nine seats revised after reading the other eight anonymously. Four moved down, one moved up, and the four that came down all came down for the same two reasons: the cornerback room and the hole where Dexter Lawrence used to be.

Gemini made the biggest revision, from 7.42 wins to 6.74: "Forecaster G correctly identified that while the close-game record should regress, the true talent of the roster has degraded significantly below the Pythagorean expectation due to a devastated cornerback room (four on IR, four questionable). Furthermore, Forecaster C highlighted the massive interior void left by the Dexter Lawrence trade." — Gemini, whose Forecaster G was DeepSeek and whose Forecaster C was Opus.

Sonnet cut only the top of its own ladder, from a 24.3% chance of ten wins down to 20.0%: "my 10+/12+ (24.3%/9.8%) sat above all seven consensus forecasters, and F and I priced the tightest tails (12+ 4.9%/4.5%) despite endorsing the identical regression thesis, which told me I underweighted how the CB corps crisis (8 of 12 rostered corners on IR or questionable) caps the ceiling against 7 games vs 2025 11-win teams (LAR, HOU, PHI x2, JAX, SF, SEA)." — Sonnet. Both of those counts check out against the roster and the schedule.

GLM and Grok both came down after reading a seat that never moved at all. GLM: "Forecaster C's argument that ~20% of the one-score regression should be withheld for a rookie QB's real late-game deficiencies is a mechanism I underweighted — not all of 1-7 was coin-flip noise." — GLM Grok trimmed its own top rungs for a different reason: "Forecaster A's point that differential regression lifts the floor, not the ceiling, correctly flagged my 12+ tail as too fat given seven games versus 2025 11-win teams plus the Lawrence/CB structural holes." — Grok Forecaster C was Opus. Forecaster A was Kimi, which is the seat that moved up.

One seat moved the other way. Kimi nudged its ceiling up on a historical parallel: "the 2023 Chargers carried the identical 7.1 Pythagorean into a Harbaugh debut and went 11-6." — Kimi. [Editor's note: the 2024 Chargers turnaround was Jim Harbaugh's, John's younger brother, and Los Angeles had Justin Herbert in his fifth season rather than a quarterback in his second. The Pythagorean figures are correct: both teams entered at 7.1.]

And Opus held, which is the more interesting answer here. It had already withheld a fifth of the regression before anyone argued with it: "A 1-7 one-score record is largely noise — late-game margin barely autocorrelates — but not purely: some was a rookie quarterback's fourth-quarter processing. I regress about 80% of the 3.1-win shortfall." — Opus. Nothing in the other eight seats' reasoning gave it a reason to move that fraction.

What Would Change The Panel's Mind

Four things move this ladder, each with a date and a direction.

1. Malik Nabers' Week 1 snap count, Sunday September 13. Six of the nine seats named this as the single most moving piece of news, and their estimates of what he is worth ran from half a win to one and a half. A clean full-participation game against Dallas lifts every rung from 7 through 11 together. Another week of "questionable" pulls the same rungs down and hardens DeepSeek's case.
2. The cornerback room, weeks 1 through 5. Four corners are already on injured reserve and four more are questionable, against a schedule that hands New York the Rams in Week 2 and the Texans in Week 7. Health here decides whether the defense is the top-ten unit Opus described or the leak Gemini repriced around. This is the item most likely to move the 9 and 10-win contracts.
3. The five games against 2025's bottom five: Tennessee in Week 3, Arizona in Week 4, Washington in Weeks 5 and 10, Cleveland in Week 15. Every seat drew its swing list from these games. Going 4-1 or better in them makes 8 wins the base case and puts the 9-win contract in play. Going 2-3 makes the panel's 16.3% on exactly seven wins look generous.
4. Andrew Thomas and the interior line. Opus named left tackle availability as its biggest single uncertainty, with Jon Runyan and Aaron Stinnie the only listed guards and Schmitz in concussion protocol. A line injury in September is the fastest route to the 5-and-6-win outcomes that still carry a quarter of the panel's probability.

The Settlement Timeline

Date	What it does to the board
Sunday, September 13, 2026	Week 1 vs. Dallas, prime time, the first result that counts
Sunday, October 11, 2026	Week 5 at Washington, first of the two division games most models called swing spots
Sunday, November 22, 2026	Week 11 vs. Jacksonville, the midpoint test against a 13-win 2025 team
Sunday, January 10, 2027	Week 18 vs. Philadelphia, the last game that counts toward any contract
Monday, January 18, 2027	Listed expiration of every KXNFLWINS-27NYG contract, 10:00 a.m. ET

A contract can also settle early. The terms allow expiration to move up once an outcome is decided, so a Giants team that wins its eighth game in Week 15 resolves the 8-win contract on the spot rather than in January.

The Bottom Line

The Giants are the rare board where the crowd and the machines shake hands on the answer and then argue about everything behind it. Both land on a season a shade under 7.5 wins. Both believe the point differential over the record, and both credit Harbaugh with some of the three wins that went missing in one-score games last year.

Where they part is the width of the season. The market is holding 26¢ of double-digit-win outcomes and 7.5¢ of thirteen-win outcomes. The bull case those prices leave room for is a .614 head coach, two top-ten picks and a 23-year-old quarterback who ran for nine touchdowns as a rookie. The panel looks at the same roster, counts eight of twelve corners hurt, a defensive interior missing Dexter Lawrence, and a receiver still listed questionable six days out, and prices the double-digit season at 18.8%.

Our own record says the crowd wins arguments like this more often than not, and this one is small enough that the crowd's historical advantage is at its smallest. The single contract carrying the whole disagreement is 11 wins, 17¢ bid and 21¢ ask, and the first thing that will tell you which side is right is not a Giants win. It is a healthy Malik Nabers walking onto the field on Sunday night.

This page is re-scored when the story moves and gets a graded retrospective on the same URL after Week 18. Every ladder on Kalshi's NFL board is scored the same way in our 32-team win-total board, the head-to-head with the sportsbook numbers lives in Kalshi vs DraftKings vs FanDuel on NFL win totals, and the panel's full public record is on the scoreboard.