Hi there, argmin readers. Today I’m Monday Morning Quarterbacking and wrote one of my sporadic football analytics posts.
NFL Sunday was a wild and weird one. Rare events dominate the NFL and are what makes it so fun to watch. The Tampa Bay Bucs destroyed the Atlanta Falcons in a hurricane, rushing for 324 yards and passing for… zero yards. Zero! They didn’t even complete a pass. No team had won a game with zero completions in over fifty years. The Broncos, quarterbacked by Bo Nix, orchestrated his twelfth comeback win down double digits in the fourth quarter. He’s only started 42 games! And yet, everyone, including his coach, still thinks Bo Nix is a bad quarterback. And yesterday was a banner day for my favorite hobby horse, the stupidity of the go-for-2-down-8 (G42D8) gambit.
To the relief of most of my readership, I haven’t written about G42D8 in a while. So let me remind those who are fortunate enough not to share my dumb obsession. A team down 14 points in the fourth quarter starts a comeback by scoring a touchdown. They are now down by 8. They face the decision to kick an extra point or attempt a 2-point conversion. In 2015, analytic-brained weirdos convinced themselves that the laws of probability insisted your chances of winning were only 50% if you kicked the extra point, but were 62.5% if you attempted the 2-point conversion, even if you missed the first one. And for the next decade, they’d show you this stupid infographic and smugly yell about the probability differential to show you how smart they are.
Late in the third quarter, down 17-3 to the Cowboys, the Cardinals benched journeyman quarterback Jacoby Brissett for journeyman backup quarterback Gardner Minshew. The Cardinals immediately scored on a 63-yard run by Tyler Allgeier. Allgeier was apparently stoked about the QB change. Down 8, they kicked the extra point. Even though it was technically still 57 seconds until the fourth quarter, G42D8 zealot Seth Walder exclaimed they should have gone for two. Regardless, the Cardinals faced the down-eight test again after finding themselves again down 14 a few minutes into the fourth quarter. They scored a touchdown and kicked the extra point. The Cardinals failed the G42D8 test, twice in a single game. They weren’t having any of this shamalytics. Of course, they went on to win.
In 2024, I wrote about the full logic of the G42D8 gambit and why it was a terrible misapplication of decision theory.1 Since the start of the 2025 season, teams attempted the 2-point when down 8 13 times. Teams kicked extra points 11 times. The strategy is certainly mainstream enough now that teams apply it more than half the time. And announcers remain completely befuddled every time. The announcers are not wrong. The analytic explanation is not “simple math.” It’s tortured, mathematicized reasoning.
But since it’s becoming the dominant strategy for losing teams, it’s worth checking how those teams fare. The G42D8 gambit teams won 2 of the 13 games where they followed the analytics (and one win is a clear technicality, which I’ll address in a second). The stodgy, conservative, meathead teams won… twice as well. In fact, they additionally tied once. That is, team G42D8 is 2 and 11 (.154 pct). Team Meathead is 2, 8, and 1 (.227 pct).
Yes, these sample sizes are absurdly small, and it’s impossible to say one is better than the other. That is actually my point. We decide NFL seasons with 17 games, after all. There is no evidence that G42D8 leads to more wins than kicking extra points.
Since 2015, 45 teams have attempted the G42D8 gambit. In these games, 22 teams have missed the first two-point conversion. The G42D8 brain geniuses tell us this is okay, because when they score the second touchdown, they can go for two again and take the game to overtime. Except that in all 22 of these games, no team has ever taken the game to overtime. 19 of these teams never scored another touchdown! In one game, the Browns scored the second touchdown but missed the second 2-point conversion and lost. In another, the Jets made the 2nd 2-point conversion, but there was still 4:30 left in the game, and the Lions subsequently marched down the field to score a winning touchdown. The truly bizarre remaining game was a Thursday Night Football NFC South masterpiece between the aforementioned enemies, the Falcons and Bucs (IYKYK). The Falcons missed the 2nd 2-point conversion, but took advantage of miserable clock management by Bucs coach Todd Bowles and managed to get a third possession. They kicked a field goal to win by 1.2
A third possession in regulation isn’t even considered a possibility in the G42D8 decision tree, where you lose 100 percent of the time you miss both conversions. Except the Falcons miraculously won. I’ll allow this one to stay in the win column for team G42D8, but it’s certainly a technicality, and certainly one that shows football strategies are far more complex than what can be captured in an infographic. The models don’t describe the game of football as much as the statistical analysts think they do.
If you’re going to insist you’re smarter than everyone, at some point you need to show results. Otherwise, the people who don’t even understand what “percent” means are going to stuff you in a locker. Down 14 points in the fourth quarter means you’re probably going to lose. You can go for two. You can go for one. But trying to argue that one is obviously superior to the other is, and always will be, dumb. Just like fighting with people on sports shows and social media.
Committed to my own dumb bits, I had a few follow-up posts in 2025 digging into the actual 2-point completion rates, the actual success rates of the strategy, and the absurdity of games where it “works.”
Bucs-Falcons games are always terrible and weird, and adding the sloppy, tired dynamics of Thursday night only made it weirder. Four plays into the drive after Atlanta’s botched 2-point attempt, Bucs QB Baker Mayfield threw an awful interception, giving the Falcons the ball back at their 33-yard line. They drove down the field and scored with 3:34 left. Even though the Falcons missed the second two-point conversion, 3:34 is an eternity in football, especially because the Falcons had all of their timeouts. They managed to stop the Bucs before the two-minute warning (there’s that clock management malpractice by Todd Bowles). With two minutes left and no timeouts, they drove 45 yards and kicked an easy 43-yard field goal. They won by 1, 29-28.