Atlanta Falcons 2026: The Coverage Report — Fewest Tracked Takes in the League, and the Worst Skill Score on Any Club

Twenty-nine of the 32 clubs we track drew more attention than the Atlanta Falcons. Atlanta is last: 323 takes in the database, against a 32-club average of 604 and a median of about 600. Of those, 137 have been graded — 30th of 32, ahead of only Arizona (130) and Carolina (120), and roughly two-thirds of the 212-per-club average.

One caveat, stated once and meant plainly: this measures the outlets and shows MissedTakes tracks, not all football media. Atlanta's local coverage is not in this file. What is in it is the national conversation, and in that conversation the Falcons are an afterthought.

That is not obviously unfair. The site's season records have Atlanta at 8-9 in 2025 and 8-9 in 2024, with 7-10 in each of the three years before that and no playoff appearance since 2017. A club that finishes a game under .500 two years running does not force itself onto anybody's show.

The Falcons were the league's most expensive team to talk about

Volume is one thing. What the graded 137 say about the people who made them is worse. Atlanta's average skill delta — the site's measure of how far a call beat a naive baseline, where zero means no better than naive — is -19.76, the lowest figure of any of the 32 clubs. Average accuracy is 44.5.

The damage is concentrated in exactly one place: preseason optimism about the Falcons as a contender.

• Playoff Picks: 8 graded, 0.0 average accuracy, -200 average skill. Every graded call putting Atlanta in the field was wrong, and each one was worse than simply assuming last year's answer.
• Division Winners: 9 graded, 0.0 average accuracy, -13 average skill. Same story. Chris Brockman in December 2024: "It's not just hot or overreacting to say they're gonna win the division. They're in first place right now." Graded 0, skill -117. Jonas Knox took Atlanta in the NFC South a year later. Also 0.
• Sleeper Calls: 14 graded, 28.6 accuracy, -79 skill. The Falcons were a favorite source of breakout picks that did not break out.

The contrast with the plainest question — how many games will they win — is stark. On season win totals, the field graded out at 86.7 accuracy across six calls, and the consensus file is close to unanimous: five 2025 win-total calls on Atlanta, four of them exactly 8 (Jason McIntyre, Danny Parkins and Nick Wright twice), one at 10 from Brady Quinn. Atlanta won 8. The skill score on those calls is still -13.33, because the naive number was 8.25 and being right about a mediocre team is not hard.

So the pundits knew Atlanta would win about eight games and picked them for the playoffs anyway.

The cutdown: 18 calls, 2 grades, and neither is from this year

This is the series' news hook, and for Atlanta it is close to empty. There are 18 roster-spot calls on the Falcons in the database. Two are graded. They are the same claim, made on the same March 2025 show: that Kirk Cousins would still be a Falcon in Week 1. Brockman said it ("Do you feel like he's on the Falcons week one? I kinda do"), Rich Eisen agreed, both graded 100 for accuracy, both worth a skill delta of exactly 0 — the naive basis was that Cousins had been on somebody's 53 the previous season, and he ended up on Atlanta's for 17 of 18 weeks.

The other 16 are pending. Thirteen of them were published in August 2026, and none has a grade yet, so there are no surprises to report from this cutdown — no consensus lock who got cut, no long shot who stuck. What we can say is what was claimed, and by whom:

• Scott Kennedy, August 30: Atlanta keeps three quarterbacks — Tua Tagovailoa, Michael Penix Jr. and undrafted rookie Jack Strand.
• Tyler Sullivan, same day, on Strand: "Most likely, Atlanta will try to sneak him onto its practice squad unless more significant moves occur at the QB position that free up a spot."
• Carter Bahns, August 27, flagging Cooper Rush as a cut candidate: "Complicating things for Rush is the emergence of undrafted free agent Jack Strand."
• Kennedy, August 11, on the locks: Kyle Pitts, Austin Hooper and Charlie Woerner at tight end; Drake London, Jahan Dotson, Olamide Zaccheaus and Zachariah Branch at receiver.
• Kennedy, August 29, on released defensive tackle Eric Johnson: "He should catch on with someone, maybe even the Falcons' practice squad next week."

Check back when those settle. Two graded roster-spot calls, both restating that a quarterback under contract would remain under contract, is not a record anyone should be judged on.

What the club's graded file actually covers

Note the depth-chart line: 80 calls, 11 graded. Most of what gets said about this roster has not resolved yet.

The player split explains the rest of the deficit. Drake London is the most-graded Falcon at 19, and calls about him average 17.5 accuracy and -46.84 skill — Andrew Erickson ("Drake London finishes as the wide receiver one overall, that is my mile take") and Derek Brown both had him as a possible overall WR1; he finished WR17. Cousins, by contrast, carries 15 graded calls at 92.8 accuracy, because "the veteran stays where he is" is easy. Penix (7 graded, 84.5 accuracy, +60.38 skill) is the one Falcon the field has read well.

Who has a named record here

Eight pundits clear five graded Falcons calls. That is a floor, not a sample worth ranking on, so treat these as records rather than a leaderboard.

Eisen's 89.2 is the highest accuracy on the club, and it is almost entirely Cousins: three graded depth-chart calls placing him correctly in the quarterback room, plus the Week 1 roster call, all worth little or nothing in skill terms because they matched the naive answer. The largest single Falcons file belongs to *The Herd with Colin Cowherd*, with 40 graded calls at 25.4 accuracy and -56.62 skill.

The summary for Atlanta is short. The tracked field talked about this team less than any other, got the win total right, got the postseason spectacularly wrong every time it tried, and has not yet been graded on a single call from the roster that just got cut to 53. See the scoring methodology for how the skill numbers are built.