Philadelphia Eagles: The Defense That Changes Everything

The Eagles went 11-6 last year, and the honest version of that sentence is: they earned most of it. Our fitted model — 8.5 base wins plus point differential divided by 35.8 — puts their structural win total at 10. The actual result was one win better than that, a luck delta that's small enough to take seriously but not so small you can ignore it. They outscored opponents by 54 points across 17 games, which works out to a little over three points per game. That's a legitimate good team, not a mirage.

The one-score record is where it gets interesting. Philadelphia went 8-4 in games decided by one score. That's a strong number, and it's also the number that most reliably regresses toward the mean. Close games are the NFL's least predictable commodity — the league's best teams tend to drift toward 50/50 in one-score situations over time, and 8-4 is far enough above that line to notice. The structural read says ten wins was the real floor. The extra win came from winning close, and winning close is the hardest thing to budget for twice.

The defense is the story, and the pass defense is the story within the story. Opponents generated -0.13 EPA per passing attempt against Philadelphia last season. To put a ruler on that: consistently holding opposing passing attacks to negative expected value per snap is what separates defenses that look good from defenses that actually are. The run defense added another -0.04 EPA per carry, so this wasn't a unit with a soft underbelly. It was a complete, genuinely elite defense.

The offense was more complicated. A 0.06 EPA per passing attempt is positive — meaning the passing game was adding value — but it's modest. The rushing attack was slightly underwater at -0.02 EPA per carry, which is close enough to neutral that it's not a crisis, but it's also not a weapon. The situational numbers help explain how this group stayed productive: a 43% conversion rate on third and fourth down is above average, and 0.13 EPA per red-zone play is a real number — the kind that says the offense gets better, not worse, when the field compresses.

Nick Sirianni's decision index is the one quiet concern. His agreement rate with our model's fourth-down recommendations was 89.7%, which sounds high until you see that he went for it on fourth down at a 12.1% clip against a model-recommended rate of 19.6%. That seven-and-a-half-point gap cost Philadelphia an estimated 0.131 expected wins last season. It's not a firing offense, but it's a real number, and it's worth watching whether the offense's modest EPA profile and the missed fourth-down opportunities are connected.

The departures are the headline. Jaelan Phillips left for Carolina on a $30 million-per-year deal — a market value of 2.5 wins by our pricing model. His production wins aren't in the brief, which tells its own story about a player whose health and snap counts make him difficult to underwrite, but the market priced him like a franchise edge rusher and Philadelphia let him go. Nakobe Dean went to Las Vegas at $12 million per year (0.94 market wins), and Reed Blankenship signed with Houston at $8.25 million per year (0.61 market wins). Add it up and the Eagles lost 4.05 wins' worth of market-priced talent from their defense in the same offseason.

On the arrivals side, Tariq Woolen came over from Seattle on a one-year, $12 million deal (0.94 market wins), and Marquise Brown signed from Kansas City at $5 million per year (0.33 market wins). The net offseason move on a market-wins basis is -2.78. On a production-wins basis — adjusting for what players actually produced rather than what the market paid — the net is 0, reflecting the fact that several of the departing players have production records that don't match their market prices.

The contract verdicts on the retention side require a look. Jonathan Greenard signed a two-year extension at $30 million per year — 98th percentile pay for an edge defender. Jordan Davis got three years at $26 million per year, also 98th percentile for an interior defensive lineman. Both carry WATCH verdicts graded purely on price discipline, with exits in 2028 at $14.1 million and $11.6 million dead cap respectively. These are top-of-market bets on players whose production percentiles aren't available to grade against. That's the honest uncertainty here.

The draft class carries 3.52 combined chart wins by pick value. First-rounder Makai Lemon (WR, USC, pick 20) accounts for 1.2 of those. Eli Stowers (TE, Vanderbilt, pick 54) adds 0.8, and Markel Bell (OT, Miami, pick 68) brings 0.74. The rest of the class is developmental depth. A projected third-round compensatory pick in 2027 adds another 0.47 wins by chart value.

The honest net is a lateral move, and lateral is not what you want when you're replacing a defense that was genuinely elite. Philadelphia lost three contributors from that defense, replaced one with a cornerback whose market value roughly equals Dean's, and spent heavily to retain Greenard and Davis at the very top of their positional markets. The production-wins net of zero is the league's way of saying the incoming players cost what the outgoing players produced — but the outgoing players were part of a defense that held opponents to -0.13 EPA per pass attempt, and that unit is now being rebuilt at significant expense.

The cap posture adds texture. Jalen Hurts sits at 10.1% of the cap, which is low for a franchise quarterback in the current market — that's genuine flexibility. Lane Johnson at 6.4% and Jordan Mailata at 5.0% reflect the cost of protecting a quarterback well. But the biggest cap-allocation gaps by our z-score model are at safety (-1.61) and edge (-1.02), meaning Philadelphia is spending significantly below the league average at two positions on a defense that just lost three contributors. The $154.8 million in future proration tells you the extension commitments are real and long, which limits how much that underinvestment at safety and edge can be corrected without another hard trade-off.

The case for optimism is straightforward: the offensive line is intact, Hurts is cheap by quarterback standards, and the draft added a first-round receiver at a position of need. The case for concern is equally straightforward: the defense that made this team worth writing about has been substantially reconfigured, and the contracts handed to Greenard and Davis at the 98th percentile of their positions are the kind of commitments that look different if either player's production doesn't match the price.

Here's the arithmetic, laid out plainly. Philadelphia's 2025 point differential of +54 translates to a structural base of 10 wins. The offseason production-wins net is 0, adding nothing. The coaching decision delta is 0, with Sirianni returning and no change to the decision-index baseline. The structural wins projection for 2026 is 10.

That is an accounting projection from the ledger — not a betting model output, not a pick. It says that the pieces on paper, priced at production rates rather than market rates, add up to a ten-win team. The falsifier is straightforward: if Philadelphia wins eight games or fewer, this read was wrong — the defense declined more than the production-neutral offseason net suggested, or the 8-4 one-score record was masking a team whose underlying quality was closer to nine wins than ten. If they win twelve, the Greenard and Davis extensions paid off faster than the price discipline flags implied, and the rebuilt defense held its level.

The brief doesn't show vacated targets on the departure side, so there's no clean arithmetic to run on who absorbs Philadelphia's outgoing opportunity. What the brief does show is Marquise Brown's production record: 74 targets last season, 137.7 PPR points, and 1.86 PPR points per target. That per-target rate is the number worth holding onto. He's running routes behind DeVonta Smith and whoever emerges from the receiver depth, which caps his target ceiling, but 1.86 points per look on a $5 million contract is efficiency the market almost certainly hasn't fully priced.

The sleeper case for Brown is exactly what a sleeper should be: measurable per-opportunity production at a price that says the broader market hasn't caught up. He didn't produce fantasy-relevant volume last season — 74 targets isn't a featured role — but if his usage increases in Philadelphia's offense, the efficiency is already there. The risk is that the usage doesn't come. Makai Lemon arriving as a first-round pick signals the team's intentions about the receiver room's future, and that future may not leave Brown with the opportunity the efficiency warrants.

The brief flags no aging-curve concerns on the offensive skill positions, so there's no fade case to run on that side of the ball. On the whole, Philadelphia's fantasy footprint this season looks more stable than exciting — a complete offensive line protecting a quarterback on a favorable cap number, with the receiver room's pecking order still being settled.

Everything above traces to a live instrument, and the receipts stay public: the GM boards carry Nick Sirianni's decision card, the team's cap posture ($154.8M already charged to future caps), and every contract verdict with its falsifier and review date — including Jonathan Greenard (WATCH), Jordan Davis (WATCH), Tariq Woolen (WATCH). When the season grades these reads, the grades post whether they flatter us or not.