Washington Bet Big on Defense While the Offense Still Flatlines

Start with the scoreboard and you get a 5-12 Washington team that looks like a rebuild in progress. Start with the point differential and you get almost the same picture. The Commanders were outscored by 95 points last season, which our model converts to 5.8 structural wins — the accounting read on how good they actually were once you strip out the random noise of one-score finishes. Actual wins: five. Luck delta: -0.8. That's about as close to zero as it gets, which means the record wasn't a mirage and it wasn't bad luck. Washington was a bad football team, and the math agrees.

The one-score record — 2-5 in games decided by eight points or fewer — is worth noting mainly because it didn't distort anything. When a team goes 2-5 in close games on top of a brutal point differential, there's no bounce-back story to tell. The margin wasn't stolen by coin flips; it was earned the hard way. That's the honest framing for everything that follows: this is an organization building from a genuinely weak floor, not one unlucky bounce away from respectability.

Very little on offense. Washington's passing attack produced -0.01 EPA per attempt — fractionally below break-even, which is another way of saying it generated no value above what a replacement-level passing game would have produced. The run game was equally uninspiring at +0.01 EPA per carry, essentially neutral. The one number that flickers positive is the situational package: a 47.5% conversion rate on third and fourth downs and +0.09 EPA per play in the red zone. Those numbers suggest the offense could execute when the game demanded precision, even if it couldn't generate much momentum in between.

The defense was the real problem. Opponents gained 0.23 EPA per passing attempt against Washington — a number that sits in genuinely bad territory — and 0.05 EPA per carry on the ground. There was no phase of the defense generating stops at a rate that could compensate for the offensive stagnation. When both sides of the ball are losing the per-play battle, the ledger tends to be unforgiving, and Washington's -95 point differential is exactly what that profile produces.

Dan Quinn enters his second year in Washington having agreed with our decision model on 71.1% of his fourth-down and situational calls, which is middle-of-the-pack. The model estimates he left 0.423 wins on the table through suboptimal in-game decisions — his go-for-it rate of 22.9% trailed the model's recommendation of 43.4% by a wide margin. That's the quiet coaching inefficiency embedded in Washington's season: a conservative fourth-down philosophy on a team that needed to manufacture any edge available.

Washington's offseason tells a clear story: they spent heavily on the defensive front and pass rush, signed a veteran left tackle to protect the quarterback, and let the backfield sort itself out. The headline arrival is Laremy Tunsil at left tackle — a two-year, $30.1 million per year deal that the market prices at 2.51 wins of value. Tunsil is 32, which is the detail that earns this contract a WATCH verdict. The 98th-percentile price tag for a player entering his age-32 season carries real risk, and the exit isn't clean: $26.7 million in dead money through 2028 if things go sideways. The falsifier here is straightforward — if the left tackle market continues inflating so fast that this deal reads like a bargain by 2027, the concern evaporates.

On the edge, Washington added Odafe Oweh on a four-year deal at $24 million annually (WATCH, 96th-percentile price) and K'Lavon Chaisson at a more manageable $11 million per year. Oweh is 27, which is a better age profile than Tunsil, but the production receipts have to arrive quickly to justify that price. At tight end, Chigoziem Okonkwo comes over from Tennessee at $9 million per year, with a production baseline of 0.01 wins — not a lot to project from, but the price is reasonable. The supporting cast of Leo Chenal at linebacker, Tim Settle at interior defensive line, Amik Robertson at corner, and Nick Cross at safety collectively represents about 2.3 market wins of value in the $6.5-to-$8.25 million annual range.

The departures are modest. Tyler Biadasz left at center (0.77 market wins), Jacob Martin at edge (0.68), and Chris Rodriguez Jr. at running back (0.33, with a production value of -0.01 — the rare departure that theoretically helps). Nothing on this list is a significant loss.

The draft class is anchored by Sonny Styles at seventh overall, a linebacker out of Ohio State worth 2.05 chart wins — the largest single-slot investment in the class. After that, Antonio Williams (WR, Clemson, 0.71 chart wins at pick 71) is the only selection with meaningful slot value. The rest of the class — Joshua Josephs, Kaytron Allen, Matt Gulbin, and Athan Kaliakmanis — accounts for 0.83 combined chart wins across rounds five through seven. Total draft class chart value: 3.59 wins.

The production-basis net for the offseason is +0.02 wins. That's not a typo — after all the signings, the departures, and the draft picks, the ledger barely moves the needle on projected output. The market-basis number, $6.45 million net in annual value added, is more flattering, but it reflects what Washington paid rather than what the model expects them to get. Those are different things, and the gap between them is where most of the risk lives.

The cap posture adds context. Washington has $149.5 million in future proration already committed — a meaningful shadow on future flexibility. The QB cap hit sits at just 5.6%, which is either a sign of financial discipline or a reminder that the quarterback situation remains unresolved; either reading has consequences. The biggest cap commitments belong to Daron Payne (8.5% of cap at interior defensive line), Samuel Cosmi (6.2% at right guard), Terry McLaurin (5.5% at wide receiver), and Javon Kinlaw (4.9% at interior defensive line). Washington is spending like a team that believes the defensive line is the path forward — interior defensive line sits at a +1.61 z-score relative to league spending, edge at +1.36.

The other side of that ledger: running back is the most underfunded position on the roster at -1.62 z-score. There are no bargain contracts flagged in the brief and no albatrosses — the roster is neither carrying dead weight nor finding hidden value. The Tunsil and Oweh deals are the live wires, priced at the 98th and 96th percentiles of the market respectively, on a team that posted -95 points in 2025. High-end infrastructure spending on a low-end product is a bet that the infrastructure is the problem. Washington is making that bet explicitly.

Here's the arithmetic as plainly as it runs: Washington's 2025 point differential projects to a structural base of 5.8 wins. Add the net offseason production value (+0.02 wins) and the coaching decision delta (unchanged, no coordinator change that moves the needle in the model), and you arrive at 5.9 structural wins for 2026. That is an accounting projection from the ledger — not a pick, not a betting position, not a forecast with confidence intervals attached. It is what the numbers on hand produce when you turn the crank.

The falsifier is uncomplicated: if Washington wins eight or more games in 2026, this read is wrong. That would require either a dramatic improvement in per-play efficiency — particularly on defense, where the EPA profile needs to move from well below average toward neutral — or the kind of sustained close-game luck that the 2025 team didn't get and that the model doesn't expect to repeat. Tunsil and Oweh would both need to perform near the top of their position groups. Styles would need to contribute as a rookie. Quinn would need to close the gap between his fourth-down decision rate and what the model recommends. All of that is possible. None of it is assumed in the number.

The most notable vacated opportunity in Washington's backfield belongs to Chris Rodriguez Jr., who carried the ball 112 times and caught four passes for 92 PPR points in 2025 before departing for Jacksonville. The carries are the real asset here — 112 touches distributed across a season is a meaningful workload for whoever inherits the lead role. The receiving production was minimal (four targets), so this is primarily a rushing opportunity, not a pass-catching one.

The arrival worth watching in that context is Kaytron Allen, the sixth-round pick out of Penn State. The brief doesn't provide per-opportunity efficiency numbers for Allen, so the sleeper case can't be built on that foundation — what we can say is that the vacated carries are real, the new contract on the position is below market (-1.62 z-score at the group level), and sixth-round running backs occasionally outperform their draft slot when the depth chart is thin. That's the opportunity side of the equation. The price side depends on where the roster settles and how training camp shakes out. Worth monitoring, not worth projecting.

At tight end, Okonkwo arrives with a 0.01 production baseline from Tennessee — a number so small it's essentially a blank slate. The target data from his time with the Titans doesn't appear in the brief, so there's no efficiency figure to anchor a sleeper argument. He's a reasonable watch for anyone tracking Washington's offensive usage patterns in the preseason, but the brief doesn't provide the foundation to call him a value play. On the defense-side fade front, the EPA numbers are clear enough on their own: Washington's pass defense allowed 0.23 EPA per attempt in 2025, and the offseason additions are largely unproven at their price points. Opponents who stack against this defense have had good mathematical footing, and nothing in the ledger yet suggests that changes.

Everything above traces to a live instrument, and the receipts stay public: the GM boards carry Dan Quinn's decision card, the team's cap posture ($149.5M already charged to future caps), and every contract verdict with its falsifier and review date — including Laremy Tunsil (WATCH), Odafe Oweh (WATCH). When the season grades these reads, the grades post whether they flatter us or not.