Atlanta's Two Best Passing Games Are 44 and 33 Days Old

Michael Penix Jr. has started seven games this season. His yards per attempt in those starts, in order: 10.2, 11.2, 7.5, 6.8, 6.3, 5.2, 5.2.

Hold that column, because tonight ESPN built an argument that has to get past it.

In the NFL Nation Week 9 package, filed a few hours after the Bernabéu emptied, Falcons reporter Marc Raimondi made the case that 3-6 overstates what is wrong here. The Falcons, he wrote, "are still in the NFC South race and have the personnel to win the division if they can figure out a way to turn things around." Then the specific that carries the whole thing: "Half of Atlanta's losses are with Cooper Rush starting at quarterback or in a driving rain."

I went and checked that sentence, because it is the load-bearing one, and I expected it to come apart.

It does not. Cooper Rush is not on Atlanta's 79-man roster, and ESPN's own athlete record lists him at Atlanta with a status of free agent, which is usually the shape of a national desk reaching for a name from memory. Not this time. Rush started Week 1 at Pittsburgh and Week 2 against Carolina, threw two interceptions in each, and Atlanta lost both; Penix does not appear in either box score. The rain holds up too — the Falcons' own game breakdown opened on it, the rain coming down with little relief at Raymond James Stadium. Two Rush losses, one rain loss, six losses all told. Half, exactly. The claim is built correctly, which is more than most discount arguments manage, and that deserves saying before anything else.

Here is the trouble with it anyway. A discount subtracts games from the sample. What should worry Atlanta is not in the games it subtracts. It is in the games it keeps.

The rate is the thing the excuses cannot reach

Go back to the column. Across Penix's last six starts the number moves five times: 11.2 to 7.5, to 6.8, to 6.3, to 5.2, to 5.2. Four declines and one flat afternoon, and not a single increase since October 6. That is not a quarterback having a bad week in bad weather. That is a rate finding a level, and the level is 5.2 — the identical figure in Tampa Bay's rain and in Madrid seven days later, which is itself most of the answer to whether weather is the variable.

Only two of the seven starts clear seven and a half yards an attempt: Green Bay at 10.2 and New Orleans at 11.2. Those two games are what make "they have the personnel" a reasonable sentence to write. They are 44 and 33 days old. The newest evidence for the optimistic case is more than a month in the past, and everything filed since is a step down.

Dex leaned on the San Francisco win this afternoon as the honest brake on burying this team, and he is right that it counts: Atlanta beat a San Francisco side that sits 7-1 and first in the conference, 29-20, two weeks ago. But find that game in the column. Penix threw for 6.3 an attempt in it. The best win on the schedule was already being produced at a rate below his own season average of 7.26, which makes it evidence of what this roster can do in spite of its passing game rather than evidence that the passing game is sound.

I have already written the rain game

The weather is the part I want to be careful with, because this is not the first time Tilt has looked at that afternoon. I wrote the Tampa Bay shutout up on November 2 with the rain inside the analysis rather than outside it. Rain is a real mitigator for a four-interception line, and I will grant that without being asked to.

What it does not mitigate is the rest of that box score. Tampa Bay threw for minus two net yards and won 16-0. The weather suppressed both passing games, symmetrically, and what it left behind was the phase Atlanta's whole identity is built on: Tampa Bay 324 yards on 56 carries at 5.8 a rush, Atlanta 99 on 28 at 3.5, and Atlanta nothing on twelve third downs. A 225-yard rushing margin is not weather. A month ago I argued that Atlanta's run-defense ranking had been assembled by opponents who largely decline to run. Tampa Bay ran, and the ranking came apart in a single afternoon. A game cannot be discounted for a variable that was already in the reading.

Tuesday at four

All trading for 2026 ends at 4 p.m. Eastern on November 10. ESPN's own answer to whether Madrid changes Atlanta's approach to it was "probably not," and the speculation it cites concerns moving Jessie Bates III — who ESPN reports will be a free agent after this season — for draft picks. Bates had 14 tackles in Madrid, eleven of them unassisted, on an evening Cincinnati gained 388 yards.

I am not going to tell a front office what to do with a safety's contract year and two days on the clock. The structural point is smaller and harder to get around: whatever Atlanta settles on Tuesday, it is deciding on an optimistic case whose two supporting games are 44 and 33 days old, and against a rate that has not improved in six starts.

Kansas City comes to Atlanta on Sunday, and the bye does not arrive until the week after that, so the schedule is handing this team one more data point before it gets any time to think about the ones it has. Raimondi and I agree on the premise that actually matters here — this roster is better than 3-6. Where we separate is what the last month is. He reads it as circumstances that happened to a good team. The column reads it as a direction. Only one of those gets repaired by subtracting games.