John Harbaugh and the Baltimore Ravens made a rare mathematical error in the fourth quarter on Sunday that might have cost them in a 31-30 loss to the Green Bay Packers.
The error, though, wasn't the one that will be endlessly debated. It wasn't Harbaugh's decision to go for 2 when down 1 with 42 seconds remaining.
Harbaugh's mistake was not giving his team a second chance if they missed.
Tyler Huntley's 2-point pass to Mark Andrews was broken up at the goal line, and all but ended Baltimore's hopes of a comeback victory. It wouldn't have, however, if it had come roughly four minutes earlier, after the Ravens scored to cut a 31-17 deficit to 31-23.
After that touchdown, Harbaugh sent Justin Tucker out to kick an extra point. And that decision, more than any other Sunday, lost the Ravens the game.
The vast majority of analytical models would have suggested a 2-point conversion right then and there, and a simple breakdown can explain the math. If, in this hypothetical scenario, the Ravens convert the 2-pointer and cut the Packers lead to 6, they merely have to kick an extra point to take the lead with less than a minute left. If they miss the hypothetical early 2-pointer, and remain down 8, they still have another chance to go for 2 and send the game to OT.
By kicking down 8, they essentially gave up an opportunity to know how many points they'd need — one or two — at the end of the game.
Harbaugh later justified the do-or-die 2-point attempt by saying that he was "try[ing] to get the win right there."
He should've tried to get the win four minutes earlier.
Was the late 2-point conversion the right call?
Harbaugh, speaking about the decision that superficially decided the game, said he thought "our chances of winning right there were a little higher than in overtime."
Analytics wouldn't strongly dispute that statement. The Ravens, with a backup quarterback against Aaron Rodgers, would've been underdogs in OT. They were still underdogs when they lined up for 2 because, even if the Ravens had converted, Rodgers and the Packers would have had 42 seconds and a timeout to drive into field-goal range for the win. Down 1 instead of tied, they would've had added incentive to push the ball downfield; to take risks; and to go for it on fourth downs. The Packers' chances of winning in regulation would've increased, too.
Both a 2-point conversion and an extra point, therefore, gave the Packers roughly the same chance to win — somewhere in the region of 35%. The choice, according ESPN's model, was a 51-49 toss-up.
According to EdjSports' model, it was a bit more tilted toward the extra point because, as EdjSports analyst Ian O'Connor wrote in an email, Tucker, one of the best kickers ever, "provides such a huge advantage to [the Ravens] in overtime. This is part of why we like the PAT better at the end of the game."
That preference, though, is marginal — and secondary.
The 'no-brainer' was the earlier 2-point decision that Harbaugh didn't make
The decision to go for 2 was a head-scratcher because, if Harbaugh felt confident enough in his offense to make it on the final drive, he shouldn't have hesitated to go for 2 when Huntley scampered into the end zone with 4:47 remaining.
If the decision inside a minute remaining was a toss-up, this one, according to ESPN's model, was a no-brainer.
At that stage of the game, with under five minutes remaining, the Ravens' only chance to win, in all likelihood, was to stop the Packers and then score again. Harbaugh should have foreseen this and made the win-now decision early.
"If Harbaugh knew he was going to go for the win at the end of the game, there’s no reason whatsoever he shouldn’t have gone for the [2-pointer] when down 8," EdjSports' O'Connor explained. "That way if they’re successful, they know they can eventually win it with a TD + PAT, but if they fail, they can still tie the game with a TD + 2PAT at the end."
Harbaugh's decision will likely become another referendum on analytics, "gut" and their respective places in coaching. He was asked after the game how much data factors into his fourth-down and 2-point decisions.
"It's mostly gut," he said. "The numbers are the numbers, but the numbers aren't perfect. I can tell ya this, I've shot a lot of holes in the numbers, with the numbers guys. And the numbers are never gonna be perfect. They don't take everything into account. So you just make a decision. The numbers are part of it. But the numbers aren't the main decision."
This pair of decisions, though, was less about numbers, more about logic. And Harbaugh, one of the most progressive coaches in the league, failed a surprisingly simple logic test.