The Outlier Was Atlanta, Not Green Bay

Green Bay has allowed 3.42 yards a carry and 105.0 rushing yards a game across the three games of its season that were not against Atlanta: 107 yards to Minnesota at 3.1 a carry, 67 to the Jets at 2.5, and 141 to Tampa Bay at 4.5 on Sunday. Against Atlanta it allowed 242 yards on 41 team carries. That is 5.90.

Atlanta's number was 2.48 yards a carry higher than what that same defense has conceded to every other opponent on its schedule.

Yesterday I argued that the run game is the only Atlanta sample large enough to reason from, and the cheapest objection to that position was available immediately: Lambeau was a soft front having a bad night, and the yardage travels nowhere. Green Bay's own four-game season does not support the objection. That defense was not permissive in the two games before Atlanta, and it was nowhere near Atlanta's number in the game after; the Tampa Bay figure only existed as of Sunday afternoon, which is why this test could not be run last week.

What that rules out, and what it does not

It rules out one explanation. It does not install another.

The same single game is the outlier in both samples, viewed from opposite sides of the line of scrimmage, and one game does not become two because it can now be described twice. There is also a mechanical reason the yardage was there to take. Green Bay rushed nine times for 17 yards and threw 53, which handed Atlanta 37:05 of possession and the carry volume that comes with it. Volume is downstream of game state. A front on the field that long against that many carries is taking a different test than one defending a balanced script, and I am not going to pretend otherwise to keep a clean finding.

The efficiency was already there

What the control test does not explain is the part of Atlanta's own three games nobody has had much use for. Atlanta rushed 32 times for 157 yards at 4.9 a carry in the game it lost 3-34 at home to Carolina. That is a full yard above the Pittsburgh figure -- 31 carries, 120 yards, 3.9 -- and a single yard below Lambeau. The run game was already working in the worst performance of Atlanta's season, with five turnovers attached to it.

So what changed at Green Bay was not the arrival of efficiency. It was volume: 41 carries against 32 and 31. It was explosiveness: a 55-yard long against longs of 11 and 20. And it was finishing: 15 rushing first downs against seven and nine. The mechanism was Bijan Robinson at 29 carries for 194 rushing yards. One precision note, because both versions are in circulation: Atlanta's two running backs carried 39 times for 244 of those yards, and Michael Penix Jr. carried twice for minus two. The team line is 41 for 242. Both totals are true and they measure different things.

A bad average that is getting better

New Orleans's run-defense aggregate is already on the record here: 4.71 a carry and 128.7 a game across 82 carries, and Pro Football Reference's opponent row agrees with the three box scores to the unit. The number I did not print yesterday is the one that complicates it. Inside those three games the trend runs the other way -- 5.0 a carry, then 4.9, then 4.2 -- and rushing first downs allowed fell harder still: 14, then eight, then four.

The oldest of the three games, which is also the one doing the most work on the aggregate, went to overtime at Detroit. Extra possessions inflate the carry count and the yardage both, so the per-game figure is slightly flattered against New Orleans by its own schedule.

Then the most recent game. New Orleans held Ashton Jeanty to 56 yards on 19 carries, 2.9 a carry, and allowed 113 rushing yards anyway, because 54 of them came on five carries by a backup -- Mike Washington Jr., including a 36-yard touchdown -- in a 27-35 home loss to Las Vegas. Strip the one run and that reads as a competent afternoon against the run. Excluding quarterback carries changes nothing: running backs have 73 carries for 347 yards against this front, 4.75 apiece.

That is not a front being displaced on first and second down. It is a front that has given up a long one. The distinction is the whole argument, because the two failures have different cures, and Atlanta's Week 3 profile -- a 55-yard long and 15 rushing first downs on the same afternoon -- tested both.

The two 29-carry games

The most load-bearing comparison available tonight is not an Atlanta number at all. In Week 1, Jahmyr Gibbs carried 29 times for 156 yards and two touchdowns against this New Orleans front. In Week 3, Bijan Robinson carried 29 times for 194 against Green Bay. Identical carry counts, the same one-back workload shape, two different fronts, and one caveat on the first: Gibbs did it in an overtime road game Detroit won by a point.

On the only measure available for both defenses, New Orleans at 4.71 across three games has been 1.29 yards a carry more permissive than Green Bay was against everybody except Atlanta. That is a comparison of what two fronts have allowed. It is not a projection, and three and four games is not enough sample to turn it into one.

What I cannot tell you

ESPN's depth charts label both clubs "Base 3-4 D," the identical string for each, which tells you nothing about how either front will line up tonight. No play-action rate, no personnel-grouping share, no pressure rate and no run-defense success rate was obtainable for this piece, so every claim above is built from what the New Orleans front has allowed rather than from how it aligns. The alignment argument waits for a source that publishes one.

Both fronts are also short, and a piece that counts only one side's absences is choosing its framing. Six New Orleans front-seven players are listed out, questionable or on injured reserve, and at one linebacker slot the first two names on the depth chart are both listed out. Atlanta's own front seven has James Pearce Jr. suspended, with no reason given in the league's data; Divine Deablo, Samson Ebukam and Yasir Abdullah listed questionable; Da'Shawn Hand and JD Bertrand on injured reserve; and DeAngelo Malone and Anterio Thompson listed out. Pete Werner was still listed questionable for New Orleans this afternoon, and nobody should write him in or out before kickoff. The guardrail I set yesterday holds today without amendment: beating a depleted front is a weak test, and if Atlanta runs for 200 yards tonight the yardage will have proved less than the total implies.

What tonight actually tests

Atlanta is the away team at Caesars Superdome, 8:15 PM ET, New Orleans favored by a point and a half as of Monday morning with a total of 47.5. It is Michael Penix Jr.'s second start of 2026 and his first career meeting with New Orleans. Carolina beat Detroit 32-26 on Sunday night and sits alone on top of the NFC South at 2-2; Atlanta and New Orleans are both 1-2 and Tampa Bay is 0-4. Tonight's winner reaches 2-2 and the loser falls to 1-3, and reaching 2-2 draws Atlanta level with Carolina on record without drawing it level in the division -- Carolina holds the head-to-head from the 34-3 win at Mercedes-Benz Stadium in Week 2, and keeps the lead either way.

What the control test bought is narrower than it looks and better than nothing. A month from now the Lambeau game will read as either the first entry in a pattern or the only one of its kind, and what decides which is not whether the yardage repeats. It is whether it repeats against a front that has spent three weeks getting harder to run through.