Bills–Lions, Week 2: Detroit Won the Rush and Lost the Down
Detroit won the pass rush and lost the down. Buffalo never let a snap last long enough for pressure to mean anything.
Buffalo 41, Detroit 31, in the first regular-season game played at the new Highmark Stadium.
Bills fans were ready and rowdy to welcome their team to their new home, and it paid off with a mostly dominant win over the Lions.
Going in
Question 1: Can Detroit get home with four?
Not just “can Detroit pressure Allen,” Buffalo’s tackle play last year is a huge reason why this question deserves to be asked: Spencer Brown and Dion Dawkins both finished top ten among tackles in pass block win rate (Josh Allen’s escapability and improvisation was a huge help here, but still), and Buffalo’s team pass block win rate was first in the league in week 1 at 71%. Against that, a defense that needs to blitz to generate pressure is in trouble, because blitzing Josh Allen means vacated zones behind a rusher who doesn’t arrive, and Josh Allen against a blitzing defense is a sight for sore eyes.
Baseline — Detroit, 2025:
| Stat | Value | Rank or note |
|---|---|---|
| Pressure rate | 24.1% | 10th |
| Blitz rate | 21.9% (21st) | Roughly ordinary |
| Sacks / sack rate | 49 / 8.1% (7th) | Finishes what it starts |
| Hurries / knockdowns | 54 (8.5%) / 50 (9.0%) | Completely average |
| Missed tackles | 76 | Among the league’s best |
| aDOT allowed | 10.0 | Highest in the league* |
| Passer rating allowed | 92.5 (19th) | 20th in passing yards |
| Passing defense EXP | −76.28 | Below average |
Every figure above from PFR 2025 opponent stats — Team Advanced Defense for pressure, blitz, hurries, knockdowns, aDOT and missed tackles; Passing Defense for sacks, rating allowed and EXP.
Answered if: Detroit sustains pressure while blitzing at or below their 2025 rate. Answered the other way if: their blitz rate climbs as the game goes — the tell that four isn’t working.
An interesting sub-question. That 10.0 aDOT allowed is a bit ambiguous due to game script and Detroit’s offensive firepower. Was Detroit’s coverage inviting deep shots, or were opponents simply chasing a Lions offense that scored 28.9 points per game in 2025, third in the league, while its defense faced 187 drives, the fourth-most in the league?
Scoring figure from Pride of Detroit; drives faced from PFR 2025 Drive Averages Against.
What happened
Buffalo’s method. Almost no shots downfield. The offense chipped — throws to the tight ends, intermediate corner routes, outs and comebacks for chunks, an effective run game with Cook, and Allen scrambling well when the pocket broke. Detroit generated pressure at a decent clip all night, and Buffalo had an answer ready every time: an outlet, or at minimum somebody to throw it at the feet of to avoid anything more negative than an incompletion.
The one deep completion was a bust. The touchdown to Josh Palmer came on broken coverage — Lions safety Avonte Maddox bit down on a crosser and had no help behind him on Palmer’s streak. That was the exception, not the plan.
Third down decided it. Buffalo reached 8-for-9 before Khalil Shakir dropped a critical conversion that went straight through his hands, and Keon Coleman converted again later to finish 9-for-11.
Buffalo’s answer to Gibbs. Jahmyr Gibbs started running well as the game went on. Buffalo responded by committing an extra defender: on Detroit’s drive down 34–24 with 8:32 left, Cole Bishop came downhill from the safety spot to meet the back at the line for no gain, then a batted ball at the line on 3rd and 10 — the tell that Buffalo’s interior rushers knew it was a pass and were playing hands-up rather than rushing. Detroit punted from inside their own 25.
The close. Buffalo’s final drive was surgical. Rock Ya-Sin went down on the field for Detroit (after Avonte Maddox went down earlier in the game). James Cook broke a 20-plus-yard carry down to the 1 and scored on the next play. Detroit could hardly ever make a stand defensively.
The answers
Detroit got pressure, and it really didn’t matter. The pressure was there at a reasonable rate and it produced nothing, because Buffalo’s answer was never about protection holding up — it was about the down ending before pressure could matter. That’s not an upset result; it’s exactly what the 2025 baseline above predicted. A defense that generates pressure and doesn’t convert it needs a way to surprise the play-caller and the quarterback, most likely with more simulated pressure (showing blitzers who drop and droppers who blitz) rather than more volume. Jack Campbell is the proof of concept: for all his coverage value, he’s one of the most effective finishers in football on the snaps Detroit does send him.
The Campbell case, with numbers. He finished 2025 as PFF’s highest-graded linebacker in football and blitzed 65 times — eighth-most in the league across every position, not just linebackers. Those 65 blitzes produced 12 pressures and five sacks, the best sack-per-blitz rate among the ten most-blitzed players in football — well ahead of the next best (Derrick Barnes, Detroit’s other linebacker in that group, four sacks on 53 blitzes) and miles past names like Kaden Elliss (three on 98) or Nick Bolton (one on 91). Detroit’s linebackers blitz often and finish at a rate nobody else in the top ten matches, which points the finger at the rest of the pass rush — a top-ten pressure rate with two elite finishers at the second level still nets the 19th-ranked passer rating allowed. That’s a problem more blitzing can’t solve. (One caveat: 12 pressures is a small sample, and small samples manufacture extreme rates.)
The aDOT sub-question, answered: Buffalo led the whole way and threw short all night — a leading team doesn’t take shots, and the one deep completion was a coverage bust, not a scheme concession. That points at game script, not an inviting secondary, as the real explanation for Detroit’s league-high 10.0 aDOT allowed in 2025. One game isn’t proof, but it’s evidence in that direction.
Buffalo’s third-down rate is the most regression-prone number in this game. 9-for-11 (82%) against a league average near 38–40% will not repeat. But separate the rate from the mechanism: the quick game, the tight end usage, and Allen’s decision-making under pressure are real and repeatable even after the number itself cools off. Buffalo won’t convert at 82% again, but not because the reasons they did were luck.
Where Detroit found something, with a caveat: Gibbs running well late was partly real, partly Buffalo softening with a lead. The run game needs to be more consistent to take pressure off Goff and St. Brown — the line’s pass-blocking was disappointing here too (more on that below).
Detroit’s line, and the other four-man story
Detroit’s pass protection looks excellent until you read the second column.
| Stat | 2025 | 2026, through Week 1 |
|---|---|---|
| Pressure rate allowed | 24.6% | 22.0% — 2nd-best in the league |
| No-blitz pressure rate allowed | not published | 23.5% |
| Pocket time / time to throw | 2.3 | 2.6 |
| Yards before contact per RB rush | 2.2 (15th) | 2.6 |
| Rush stuff rate | not published | 12.1% |
| Adjusted line yards | 4.78 (20th) | — |
| Clean pocket rate | 70.0% (17th) | — |
2025 pressure rate and pocket time from PFR offense advanced; yards before contact, adjusted line yards and clean pocket rate from FTN; the 2026 column from Sharp Football.
The no-blitz number should be lower than the overall number, and it isn’t. For almost everyone it is — Cincinnati allows 21.6% overall and 15.0% without the blitz, Las Vegas 22.6% and 13.0%, Arizona 25.6% and 20.7%. Detroit inverts it: their line holds up better against extra rushers than against a standard four. A blitz is a chess problem — a protection call, a hot route, a back kept in — that coaches solve. Losing reps against a plain four-man rush is personnel, and personnel doesn’t get fixed in a week. Buffalo’s 2025 defense blitzed on just 21.8% of dropbacks while pressuring on 23.9% — exactly the four-man profile Detroit’s line is least built to handle.
Around those protection numbers: 39 sacks allowed in 2025, 154 total pressures on roughly 626 dropbacks, 155 blitzes faced, 1.8 yards after contact per rush, 23 broken tackles. FTN pins last year’s decline on interior turnover — two first-time starters at guard — and had the unit 14th entering 2026, down from its usual top-five standing. Pocket time and pressure allowed normally move together, since more time in the pocket gives a rusher more chances to win; this year they moved apart, pocket time up three-tenths of a second while pressure allowed fell five points. That reads as a line that actually got better, not an offense throwing deeper. (PFR’s 2025 yards-before-contact figure of 2.8 includes QB runs; FTN and Sharp’s 2.2 is RB carries only — the number that lines up with this year’s 2.6.)
The run game is where Detroit needs to make up the difference, and the sources disagree. Sharp likes it — 2.6 yards before contact, 12.1% stuff rate. ESPN has the run block win rate as mediocre, 21st in the league. Adjusted line yards* has the unit collapsing, 1st in 2024 to 20th in 2025. Trust Sharp’s number least: a single breakaway run inflates yards before contact, and Detroit led the league in 30-plus-yard rushes last season while averaging 3.6 a carry without them. Three sources, one game, real disagreement — leave this open.
*Adjusted line yards separates the line’s work from the back’s: it weights each carry on a sliding scale (losses count for more than their full value, yards 1–4 count fully, 5–10 count half, 11+ don’t count at all), then adjusts for down, distance, opponent and situation.
So: the line is genuinely better than last year, but it carries one specific weakness — a standard four-man rush, the one kind of pressure a coordinator can’t scheme away — and Buffalo is exactly that kind of team. Thursday wasn’t evidence the improvement is fake; it was the one matchup built to hide it. The real test is a defense that has to send extra to get home. Watch for it.
What I expected
Going in, I thought this game would be decided by the pocket holding up for four seconds against Detroit’s rush, with Allen’s escapability as the tiebreaker. Right battlefield, wrong mechanism: Buffalo never fought that fight. The plan was making sure the ball was gone before 2.5 seconds, not surviving to four — outlets, tight ends, quick game, and a standing willingness to throw it at somebody’s feet rather than take a negative play. That’s play-calling and structure, not protection. A rush can win its rep and still lose the down if the down was built to end before the rep resolved.
What carries
Detroit’s pressure not converting is a season-long question, not a Thursday one. The check: by Week 6, is Detroit still top-half of the league in pressure rate and bottom-half in passer rating allowed? If yes, that’s the defining fact about this defense — everything behind the rush is giving the quarterback somewhere to go, and the secondary injuries won’t help. If it breaks, the coverage improved and Thursday was just a bad matchup.
Second thing to track: whether Buffalo’s quick-game answer to pressure is a Josh Allen skill or a plan built specifically for Detroit’s personnel. Next Gen Stats time-to-throw over the coming weeks should answer it.
Numbers still to pull
This game’s stats won’t be available for another couple days, and I will try to update this, but in case I don’t, here are some stats and their sources for you to see for yourself:
| Stat | Source | Available |
|---|---|---|
| Allen’s time to throw | Next Gen Stats | Within a day |
| EPA per play, both sides | rbsdm | Same night |
| Win probability chart | ESPN Gamecast | Now |
| DET pressure rate and sacks, this game | PFR 2026 opponent stats | Tuesday \~6pm ET |
| DET aDOT allowed this game vs. the 10.0 baseline | PFR 2026 opponent stats | Tuesday |
| BUF third down rate, league context | PFR 2026 offense | Tuesday |
| BUF pressure rate faced, this game | PFR 2026 offense advanced → Pressure | Tuesday |
Time to throw is the one to chase hardest. It converts “Buffalo got the ball out before the rush arrived” from something observed into something measured — the empirical spine of this whole piece. (Allen’s accuracy while pressured is a paid PFF stat; the free substitutes — pressure rate faced, sack rate, scramble production — constrain the picture well enough without it.)
What’s next
Both teams are home on September 27 — Buffalo hosts the Chargers, Detroit hosts the Jets — and only one of them gets a real test.
Buffalo’s comes immediately. The Chargers pressured on 23.6% of dropbacks in 2025 while blitzing on just 19.0% — the same get-home-with-four profile Detroit brought — and opened this season second in the league in pass rush win rate at 78%. If the quick game is an Allen skill rather than a plan built for one opponent, we’ll know in ten days.
Detroit’s is the wrong opponent to learn from. The Jets managed 26 sacks and a 4.8% sack rate last season while surrendering a 110.9 passer rating — a line with a four-man-rush problem won’t be asked a hard question, and a defense that can’t convert pressure will be handed an easy one. Both of this piece’s open questions survive to Week 4.
Whatever those two games bring, the lesson from this one holds. Pressure rate quietly assumes the pressure gets a chance to matter. Buffalo spent the whole night making sure it never did.