Atlanta's Run Defense Was Built by Teams That Don't Run

On September 20 at Mercedes-Benz Stadium, Carolina scored 34 points, Atlanta scored 3, and Carolina ran the ball 22 times doing it. Twenty-two carries on 60 offensive plays, with a 17-point lead at the interval and a 31-point lead entering the fourth quarter. A team that far ahead is supposed to put the game in the ground and go home.

Carolina kept throwing.

Yesterday I printed that Atlanta allows 48.3 rushing yards a game, fewest in the league, and I raised the obvious objection against myself: some of that low volume was donated by the scoreboard. That objection was half right. The other half is that two of the four offenses Atlanta has faced had the lead and declined to run at it anyway.

Four offenses, and not one of them a running team

Rush rate says what an offense wants, and the honest version counts sacks, because a sack is a called pass play that failed. Carries divided by carries plus pass attempts plus sacks taken, across four games, from ESPN's by-team splits: Pittsburgh 33.9%, Carolina 35.4%, Green Bay 31.1%, New Orleans 34.0%.

Four of the seven lowest rush rates in the league belong to the four teams Atlanta has played. Baltimore's is 53.1%.

Underneath the rates it gets worse. Green Bay is last in the NFL in rushing yards, rushing yards per game and rushing attempts — 232, 58.0 and 73. Neither Pittsburgh nor Green Bay has scored a rushing touchdown this season. Carolina has 24 rushing first downs against 63 passing.

Here is the plain version, and it is less flattering than the number it explains. The most run-committed offense Atlanta has faced all season is Carolina at 35.4%, and "most run-committed" is doing a great deal of work in that sentence. Not one of these four is a downhill running team. The four average 22.1 carries a game. Baltimore averages 31.8.

Half of it was the scoreboard. Half of it was the opponent.

Since the game-state objection is mine, here is the audit. Green Bay carried 9 times for 17 yards, 14.3% of its offensive plays, trailing by three scores at Lambeau Field. New Orleans carried 19 for 50, 27.5% of its plays, trailing at Caesars Superdome. Game script explains those two.

Pittsburgh won its game and ran 21 times, 32.8% of its plays. Carolina won by 31 and ran 22 times, 36.7%. Game script explains neither.

A defense spared by the scoreboard in two games was spared by the opponent's own preference in the other two, and those are different kinds of untested. The second kind governs Sunday, because the input this number has never been given is volume from an offense that wants to hand it over. Atlanta has not been asked to defend more than 22 carries in a game this season.

What losing Lamar Jackson actually takes off the field

Baltimore head coach Jesse Minter told reporters on Friday that Jackson will not play, with a left ankle injured in the win over Tennessee; the Falcons' own site carries it. Jackson has 19 of Baltimore's 127 carries this season, 15.0% of them, and 144 of its 614 rushing yards, 23.5%, at 7.6 a carry — at once the most efficient part of that run game and the smallest share of its volume. Take his carries out of the ledger and Baltimore is still 108 attempts for 470 yards, 4.35 a crack, 117.5 a game. That is 2.4 times what Atlanta has allowed.

So what leaves the field Sunday is the efficiency of the quarterback run, not the quarterback run. Tyler Huntley, whom the Associated Press expects to start, has 169 career carries for 811 yards at 4.8 across 31 appearances, and he has run in every season he has played. He carried four times for 16 yards in relief at home against Tennessee. If what you plan to watch Sunday is which quarterback Baltimore starts, you are watching the wrong half of that offense.

Then the part I cannot resolve. The two observations of Baltimore without Jackson disagree.

In the three 2025 games Jackson did not appear and Huntley was the primary passer, Baltimore ran 125 times on 208 plays — a 60.1% rush rate and 41.7 carries a game, against a full-season 52.1% and 29.8. But that sample is three games long and belongs to a different staff: Minter is in his first year, and nothing in it is evidence about what he calls.

The only sample that is his is six days old. Jackson was hurt in Baltimore's 24-18 win over Tennessee, Huntley finished it, and Baltimore scored nothing after the interval, Tennessee outscoring them 11 to nothing in the second half. Baltimore ran 30 times for 112 yards, 3.7 a carry, its worst rushing rate of the season. Derrick Henry carried 23 times for 73 yards, his worst game of the year.

Pick whichever sample flatters your argument and you can have either Baltimore. I am not going to. The defensible claim is narrower and it is about Atlanta: this front has not been asked to hold up against that volume, and nobody has seen this staff build a plan without its quarterback.

One thing the season line hides about Henry. His yards per carry has fallen every week — 6.0, then 4.3, then 3.4, then 3.2 — and over his last two games he has carried 49 times for 162 yards. He is still tied with Buffalo quarterback Josh Allen for the league lead in rushing touchdowns at 7. A two-time rushing champion averaging 3.3 a carry is a different problem than the preview copy describes, not a smaller one.

The front he is running at

Atlanta posted a game status report on Friday, which it had not done when I wrote yesterday, and every player on it plays in the front seven. Yasir Abdullah is out with a hamstring. Divine Deablo, Samson Ebukam and Za'Darius Smith are listed questionable — two hamstrings and a concussion. No offensive player and no defensive back appears in that block.

That is not a footnote to a run-defense argument. Smith and Ebukam are the two outside linebackers in Atlanta's base 3-4, and between them they have produced 4 of the team's 7 sacks: 57% of the pass rush out of two men who have each played three games rather than four.

The chart behind them is the uncomfortable part. Behind Smith it lists Cameron Thomas, then Abdullah, who is out. Behind Ebukam, Jared Ivey, then James Pearce Jr., who carries a suspension designation and has not appeared in a game this season, then DeAngelo Malone, who is out. Behind Deablo, Christian Harris, then JD Bertrand, who is on injured reserve. Every questionable starter in this front is backed by a man unavailable one rung further down. The interior — LaCale London, Gervon Dexter Sr., Brandon Dorlus, Maason Smith — is the healthy part, and against 31.8 carries a game it is also what gets hit first.

Eight days ago I wrote that Atlanta's linebacker problem was a coverage problem rather than a run-fit problem, because the two compromised slots sat beside each other on one side of the front. The attrition has crossed the formation since. Sunday brings the one shape of offense that can falsify the run-fit half of that sentence, and I would rather say so now than after the fourth quarter grades it.

The one genuine snap figure available comes from the club, not a feed: Deablo played 81% of Atlanta's defensive snaps at New Orleans. One player, one game. The same report notes he and Ebukam both carried this designation before that game and both played — the only base rate on offer, and it covers two cases.

What Sunday is actually measuring

Atlanta is 2-2, 0-1 at Mercedes-Benz Stadium, and second in the NFC South behind a Carolina team it shares a record with and lost to by 31 points. The run defense is the one thing on this roster that genuinely changed, and I am still going to defend the number.

What I will not do is pretend it has answered a question nobody put to it. Four opponents told us what this front does against offenses that would rather throw, and the answer — 2.7 yards a carry, the rate I printed yesterday — is excellent. At 8:20 on Sunday night the other kind of offense arrives, without the quarterback who made its run game efficient and with every reason to lean on the back who makes it heavy. Whether 48.3 described one defense or four opponents is a question the first 10 carries will begin to answer.