From Kirk Cousins engineering a late Falcons touchdown to Matthew Stafford’s turnovers flipping the Rams-Panthers game, NFL Week 13 delivered some of the wildest swings of the season. Numbers tell one story, but the momentum swings tell another — and this is where advanced analytics like win probability capture the drama.

Atlanta Falcons vs New York Jets  – ESPN Analytics Win Probability 

With 10 minutes left, Kirk Cousins moved the Atlanta Falcons from negative territory to inside the 10-yard line in two plays. Tied 17-17 with the New York Jets at the time, the Falcons needed a play like that, a play to seize some momentum — before the sequence, ESPN Analytics gave them a 60 percent win probability, but it seemed like they could lean on running backs Bijan Robinson and Tyler Allgeier, who both had scored earlier in the game, to get the win.

Instead, it was a pass — a touchdown pass from Cousins to David Sills V on third down — that gave them the lead. After the score, their win probability rose to 83 percent. Against the 2-9 Jets, that was a guarantee, especially with the dynamic run game the Falcons don. As long as they played decent defense, logic said, they’d get the win. 

That was the problem, however. They couldn’t get off the field on third down; not once, not twice, not three times, but four times they had the Jets facing third down and they couldn’t keep them from crossing the first down marker or getting close enough to go for it on fourth. The Jets even had a false start to turn a third-and-4 into a third-and-9, but on play after, Tyrod Taylor found Alan Lazard for the first down. That’s unacceptable, and it paved the way for Taylor to run it in for the 10-yard game-tying touchdown with 1:53 left. Nonetheless, and despite their inability to get off the field on that drive, Atlanta still had a 68 percent win probability.

That proved meaningless. The Falcons’ ensuing possession was stagnant — after a bleak, one-yard run from Robinson on first down, Cousins threw two incompletions and the Jets forced a punt. Slowly, it was becoming a complete disaster for the Falcons. Their win probability had dropped to 52 percent. Defensive lapses had become offensive miscues. A drive that was supposed to get them back in control had lasted 27 seconds.

On the following possession, the Jets looked like themselves. A sack on the first play set them behind the chains and derailed the drive, and three plays later, they punted. The Falcons exhausted their timeouts on that drive, which lasted 29 seconds. The game was becoming a back-and-forth disaster, a contest of who could do a better job of finishing a lesson on how to lose a game. The win probability model seemed lost — neither team had momentum, so it gave the Falcons, again the better team on paper, a 62 percent chance of winning.

Guess what? Laughably, Cousins and the Falcons went three-and-out — Cousins threw three straight incompletions. On the ensuing drive, the Jets mustered up something that resembled pro football. They got a long punt return and found a few positive plays. Indeed, there were incompletions and offensive penalties mixed in, but they did just enough to get into Nick Folk’s range, and he kicked the 56-yard field goal as time expired to get the win.

Final: Jets win 27-24

Los Angeles Rams vs Carolina Panthers  – ESPN Analytics Win Probability Chart

Looking back, we should have expected Matthew Stafford to cool off at some point. What he was doing was unsustainable — going into Sunday’s game against the Carolina Panthers, he hadn’t thrown a pick in eight games. We probably shouldn’t have been surprised when he threw one in the middle of the first quarter. Now, when he threw a second interception, this time a pick-six, on the very next possession, we had every right to be shocked.

But even as the Panthers climbed ahead 14-7 with the pick-six, ESPN Analytics still favored the Rams’ track record, suggesting they had a 53 percent win probability. After all, it was just the first quarter. It was safe to assume Stafford and the Rams would get their act together.

They seemed to in the second quarter, putting together two touchdown drives. The problem for the Panthers had become extending the lead. After the Rams tied it at 14 early in the second quarter, the Panthers kicked a field goal — great for putting points on the board, but no team puts together a 12-play, 60-yard, seven-plus-minute drive just to get three points. That doom of a drive put Carolina up 17-14. And after, the Rams had a win probability over 60 percent.

The Rams responded with a seven-play, 80-yard drive of their own. The main difference? Unlike the Panthers, they cashed in, taking a 21-17 lead after a Blake Corum touchdown. That drove their win probability over 75 percent, and after the Panthers punted on the following possession to essentially end the half, it seemed like it could be a turning point. 

But the defense faltered. On the opening drive of the second half, Bryce Young and the Panthers put together 10 plays, went 71 yards, and scored a touchdown. That gave them the 24-21 lead, but the Rams still had 56 percent win probability. And in the same way their second quarter drives were elating, the following Rams’ drive was deflating — a three-and-out. Even after, their win probability was 51 percent. 

Even the analytics couldn’t buy into the Panthers. 

But that changed for a moment. At the end of the third quarter, they neared the red zone with nine- and 10-yard runs from Chuba Hubbard — and suddenly, they had a 64 percent win probability. Then they fell apart. First, they committed a penalty on second-and-6, and then Young took a sack. That, apparently, took them out of field goal range — it would’ve been a 56-yarder and worth a shot, I would argue, but they punted.

It was a mind-boggling decision, and I understand the field position game, but to beat an elite team like the Rams, teams like Carolina have to take risks. That’s just part of it. Looking back it’s moot because they ended up winning, but to me, they left points on the field. 

But the reality is that they won because the Rams finished by digging a hole  for themselves. At the beginning of the fourth quarter, they put together another drive you’d expect from a 9-2 team — a six-play, 89-yard, heels-dug-in drive that resulted in a Kyren Williams touchdown. Following the score, they led 28-24 and had a 76 percent win probability. 

Their win probability jumped to 80 percent as the Panthers encountered a fourth-and-2 from the 43 yard-line. But it flipped as Young’s pass met Tetairoa McMillan and McMillan ran 20 more yards for the touchdown. That put the Panthers’ win probability at 60 percent as they led 31-24. 

Did Stafford and the Rams have one more gasp? It looked like it. They were moving down the field. Stafford was spreading the ball around. As they moved inside the 25 yard-line with a catch from Puka Nacua, their win probability jumped to 57 percent. But then the villain from earlier in the night reappeared. This time, Stafford fumbled. And this time, it essentially ended the game.

Final: Panthers win 31-28