Start with the number that matters most heading into this offseason: Seattle went 14-3 last season with a point differential of +191. Run that through the fitted points-per-win model — 8.5 plus point differential divided by 35.8 — and you get a structural win total of 13.8. The luck delta is +0.2. That's it. That's the whole luck audit. What Seattle did in 2025 was almost exactly what a team with their scoring margin should have done.

That's rarer than it sounds. Most teams with elite point differentials either outrun their margin in close games or fall short of it. The Seahawks went 6-3 in one-score games, which reads as slightly fortunate — flip two of those and you're at 12-5 — but the sample is small enough that it barely moves the needle. The structural read of 13.8 wins is essentially the whole story: this was a legitimately great football team, not a smoke-and-mirrors team hiding behind a fluky record.

The practical implication for 2026 is that Seattle doesn't get a bounce-back narrative. They were as good as their record said. The question is whether they can hold that level against a harder schedule and after an offseason that, on balance, trimmed some talent from the edges of the roster.

The EPA profile is the cleanest picture of what Seattle actually was. On offense, they posted a passing EPA of +0.14 per attempt — above-average, efficient, not spectacular. The rush game was a mild drag at -0.07 per carry, which is normal for a team that leaned on the pass. Their 46.5% conversion rate on third and fourth downs was among the better marks in the league, and a red-zone EPA of +0.07 per play means they generally cashed their opportunities in scoring position. The offense didn't have one transcendent mode; it had very few bad modes.

The defense was the quieter story and arguably the better one. Seattle held opponents to -0.10 EPA per pass attempt and -0.15 per rush attempt. That rush figure is particularly striking — a defense that takes away the run at that clip forces opponents into obvious passing situations, which compounds the benefit across an entire game. A defense that is genuinely good in both phases doesn't show up often at that EPA level; Seattle's did.

Head coach Mike Macdonald's decision index warrants a close look. His model agreement rate was 78.9% — a respectable number — but his go rate on fourth down was 13.2% against a model-recommended rate of 28.9%. The gap cost Seattle an estimated 0.249 expected wins last season. That's not a crisis, but it's half a win over a full season sitting in a drawer. Whether Macdonald closes that gap in year two matters on the margins of a tight NFC West race.

The departures are the headline. Boye Mafe left for the Bengals on a deal worth $20 million per year — a market-wins value of 1.63. Tariq Woolen went to the Eagles at $12 million per year (0.94 market wins). Coby Bryant landed in Chicago at roughly $13.3 million annually (1.06 market wins). And Kenneth Walker III headed to Kansas City on a $14.35 million per year deal, with his production-wins figure from last season at essentially zero — a data point worth holding onto. On the market-wins ledger, Seattle walked out the door with 4.77 wins' worth of cap contracts.

The one notable arrival is Rashid Shaheed from New Orleans, signed at $17 million per year, carrying a market-wins value of 1.37. His production-wins number from 2025 was zero — meaning the model didn't register a meaningful win contribution from his output last season. Seattle bought market price for a player the numbers haven't confirmed yet.

The draft class carries 3.14 chart wins in aggregate. Jadarian Price (RB, Notre Dame, pick 32) brings 1.07 chart wins as the first-round answer to the Walker departure. Bud Clark (S, TCU, pick 64) adds 0.77 to address the Bryant hole at safety. Julian Neal (CB, Arkansas, pick 99) is a reasonable attempt to backfill some of the Woolen loss. Beyond that, the class thins quickly — picks five through seven contribute a combined 0.56 chart wins.

On extensions, four contracts were written this offseason, and the ledger has opinions on all of them. Jaxon Smith-Njigba's four-year, $42.15 million per year deal grades as FAIR PRICE: he sits at the 99th percentile in production and the 100th in pay, and at 25 with an aging multiplier of 0.82, the runway is long. The exit in 2028 carries $21 million dead — manageable if the production holds, uncomfortable if it doesn't. Devon Witherspoon (four years, $33 million per year) and Charles Cross (four years, $26.1 million per year) both grade WATCH — top-of-market prices without enough production data to confirm the value yet. The most eyebrow-raising entry is Leonard Williams at three years, $30 million per year: 99th percentile pay, no clean exit, and age 33 at signing. Three projected 2027 compensatory picks (rounds four, five, five) add a combined 0.82 chart wins to the longer view.

Net the ledger honestly and the offseason hurt — but barely. The production-wins balance is -0.01, which rounds to a push. The market-wins balance is -3.4, which reflects how much cap value walked out relative to what came in. Those two numbers tell different stories, and the truth is somewhere between them: Seattle let go of players whose market prices had outrun their on-field output (Walker's production wins were essentially nil) while paying top dollar to keep players whose value is still being demonstrated (Witherspoon, Cross) and one whose best years may be behind him (Williams).

The cap posture deserves its own paragraph. Sam Darnold sits at 12.1% of the cap at quarterback — a number that is either excellent or alarming depending on what Darnold is. Williams consumes another 9.4%. Uchenna Nwosu is at 6.4%, and Cooper Kupp at 5.6%. That Kupp number is the one that stings: a 47th-percentile producer eating 92nd-percentile pay is the definition of an albatross contract, and $164.7 million in future cap proration means Seattle has structural obligations that limit flexibility. The special teams group is spending at a z-score of +2.85 above the league mean — an unusual concentration — while the running back group sits at -1.32, which makes sense given that Walker is gone and Price is on a rookie deal.

The aging curves give back with one hand what the extensions offer with the other. Smith-Njigba's aging multiplier of 0.82 is the only one the data carries, and it's encouraging — a 25-year-old at the top of the receiver market is the right kind of bet. The Williams deal has no aging multiplier listed, which is the model's polite way of saying it didn't bother running one for a 33-year-old defensive tackle on a three-year commitment with no exit. That contract will be tested well before it expires. On balance, Seattle trimmed real edge-roster talent and replaced it with draft chart value and one unproven arrival, while locking in high-priced extensions that concentrate risk in the middle of the roster.

The arithmetic is straightforward. Seattle's structural base from the 2025 point differential is 13.8 wins. The offseason production-wins adjustment is -0.01 — a rounding error, not a meaningful change. The coach decision delta is zero; Macdonald is back, same model agreement rate, same gap on fourth down. Add it up: 13.8 + (-0.01) + 0 = 13.8 structural wins projected for 2026. This is an accounting projection from the ledger, not a betting model output and not a pick.

The falsifier is simple: if Seattle wins fewer than eleven games, this read is wrong, and the likeliest culprits would be a regression from Darnold at quarterback (the number most exposed to single-player variance), the defensive secondary losing more than expected from the departures, or the Williams contract producing nothing in year one. If they win fifteen or sixteen, this read was conservative, and the draft class or Smith-Njigba's continued ascent probably explains the gap. Eleven wins to the downside, fifteen to the upside — that's the range where the model's confidence lives.

The most concrete opportunity question in Seattle's backfield is who absorbs Kenneth Walker's workload. Last season Walker ran the ball 221 times and caught 36 targets for 191.9 PPR points. That is a substantial share of a productive offense, and it now belongs to whoever wins the job. The draft answer is Jadarian Price, a first-round pick from Notre Dame who arrives on a rookie deal. The opportunity is real; whether Price is ready to handle a full three-down role as a rookie is the variable the market will spend all summer debating.

On the receiver side, Rashid Shaheed saw 92 targets last season with the Saints and produced 156.6 PPR points on a per-target rate of 1.70. That efficiency number is the part worth watching. He arrives into an offense that already has Smith-Njigba as the clear top option, which creates a genuine question about how many looks Shaheed actually sees. The opportunity exists, but it competes with an elite target-absorber at the top of the depth chart.

The sleeper case here is straightforward if unexciting: Smith-Njigba himself. He posted 99th-percentile production last season and is signed through 2028 at a current cap number of 3.3% — 82nd percentile in pay against 99th-percentile output. The market has since corrected with the extension, but the production profile and his age suggest the output is more likely to grow than shrink. The aging flag at a multiplier of 0.82 is a mild caution, not a warning. In DFS terms, the efficiency the price once hadn't caught up with is what you'd want in a stack anchor; now you're paying for it, and the question is whether the production justifies it. Last year's numbers say yes.

Everything above traces to a live instrument, and the receipts stay public: the GM boards carry Mike Macdonald's decision card, the team's cap posture ($164.7M already charged to future caps), and every contract verdict with its falsifier and review date — including Jaxon Smith-Njigba (FAIR PRICE), Devon Witherspoon (WATCH), Leonard Williams (WATCH), Charles Cross (WATCH). When the season grades these reads, the grades post whether they flatter us or not.