NFL Week 4 was unbelievable. Late chaos. Lead changes. Fanbases getting their hopes up in the second half just to get disappointed in the end. Win probabilities flipping in seconds. Drama, drama and more drama.
Before I get into the analytics chaos, I want to highlight two bonus games that shocked me, that in the moment, felt like epitomes of win probability chaos but actually weren’t. In the Dublin game, the Minnesota Vikings scored 15 unanswered points in the fourth quarter to cut the Pittsburgh Steelers’ lead to 24-21, but according to the game’s win probability chart on ESPN Analytics, they never shifted the game win probability-wise — the closest thing to it was when they scored their last touchdown to get within a field goal with two minutes left, but even then ESPN Analytics gave them a 91 percent chance of winning.
Similar thing in the Seattle Seahawks-Arizona Cardinals game; the Seahawks had a 97 percent chance to win based on ESPN Analytics’ game win probability model, but then the Cardinals scored two touchdowns — with the Seahawks missing a field goal in between — to tie the game with 28 seconds left. But because a game-winning field goal loomed, the Seahawks still had a 72 percent chance of winning. Sure enough, Jason Myers hit a 52-yarder as time expired.
Without further ado, here are the games that sent win probability charts wild:
Indianapolis Colts vs. Los Angeles Rams – ESPN Analytics Win Probability Chart
Up 20-13, the Indianapolis Colts had an 84 percent chance to win with 4:06 left, according to ESPN Analytics. They had controlled much of the fourth quarter up until that point — the only play the Los Angeles Rams had run from scrimmage in the fourth ended in a Kyren Williams fumble. Matthew Stafford had just completed a pass to Puka Nacua on third down to create a fourth-and-2 situation inside the Colts’ 10 yard-line. The Colts just needed one more stop.
And then the Rams went for it all; Stafford went back to Nacua on a short pass and Nacua ran four yards into the end zone. Tie game. 20-20. The Colts still had the edge but barely — ESPN Analytics gave them a 56 percent chance of winning. It seemed like momentum favored the Rams even more, though. The Colts had a chance to negate a second straight Rams possession — the one before ended with the Williams fumble — but they didn’t. And rather than icing the game, they were in a battle.
On the following possession, the Colts ran six plays for 20 yards. It was a drive haunted by a penalty and a fumble by quarterback Daniel Jones — although they were able to recover the fumble, they lost 9 yards on the play. So at the two-minute warning, it was a tie game and the Rams had a 62 percent chance to win.
And then The Play happened. The Play that completely flipped the game, the momentum. The Play to not only open but close the drive. The Play where Stafford bounced in the pocket behind his own 12 yard-line and connected with Tutu Atwell, who then ran 62 yards for the score and the 27-20 lead. It was The Play that didn’t just break the tie but gave the Rams a 93 percent chance of winning with 93 seconds left, The Play that stole the game and ended the Colts’ perfect season
Final: Rams win, 27-20
Chicago Bears vs Las Vegas Raiders – ESPN Analytics Win Probability Chart
From the end of the first quarter until the closing seconds, the Las Vegas Raiders seemed to havethe edgeover theChicagoBears, according to ESPN Analytics. It felt like they set the tone in the first half as they scored touchdowns while the Bears settled for field goals — in fact, that was the difference in their 14-9 halftime lead. It felt like they could cruise to the win; heading into halftime, they had a 65 percent chance of winning.
And then the second half happened. On the first play from scrimmage, Geno Smith threw an interception, giving the Bears the chance to take the early third quarter lead. After that play, likely because the Chicago offense had been struggling, the Raiders still had a 57 percent chance of winning. But although it had two offensive penalties, the ensuing drive ended in a Bears touchdown. That momentarily flipped the script, giving the Bears the 16-14 lead and a 56 percent chance to win.
Slowly but surely, it was becoming a back-and-forth affair. On the following possession, Smith connected with running back Ashton Jeanty for a 9-yard touchdown to give the Raiders a 21-16 lead and 66 percent win probability. And then to begin the fourth quarter, the teams traded field goals and the Raiders led 24-19 and had a 74 percent chance to win.
It seemed like Las Vegas looked to have the game under control — they even had a 81 percent chance to win after running back Jeanty ran for 13 yards on third-and-1 to break into the 10 yard-line with nine minutes left — but hope continued to loom for the Bears. The Raiders could’ve gone up two scores, but instead, they ran three plays — two where they lost yards and an incompletion. That forced them to go for a field goal, which stretched the lead to 24-19. After that, they had a 74 percent chance to win, but it felt more even than that.
That’s when Caleb Williams and the Bears turned it on. Overcoming a penalty, Williams led an impressive five-minute, 11-play, 69-yard drive to take the 25-24 lead. After the D’Andre Swift score, they tried to get the two-point conversion but failed. So they led by a point with 94 seconds left — and had a 72 percent chance of winning. Live, I didn’t think the lead meant too much, assuming the Raiders would simply drive down the field to set up the game-winning field goal. I assumed wrong. They drove down the field, trotted Daniel Carlson out to put them ahead with 38 seconds remaining. But then the Bears’ special teams stole the show, blocking Carlson’s 54-yarder and essentially sealing the victory.
Final: Bears win, 25-24
Green Bay Packers vs. Dallas Cowboys – ESPN Analytics Win Probability Chart
The Green Bay Packers led the Dallas Cowboys 34-30 in Micah Parsons’ return to AT&T Stadium when chaos unleashed, or continued, depending on how you see it. The past five drives had ended with touchdowns, and even though the star on defense had been the story coming in, the offenses had stolen the show. ESPN Analytics gave the Packers a 86 percent chance of winning after they took that 34-30 lead with 1:54 left.
But then Dak Prescott and the Cowboys started driving. After Prescott threw an incomplete pass on first down and a short completion on second to bring up third-and-3, the Packers needed a stop to take complete control. Their win probability stood at 72 percent. Two plays later, Prescott threw a 28-yard touchdown to George Pickens to go up 37-34, and the win probability flipped. The Cowboys had an 87 percent chance of winning with 43 seconds left.
87 percent seemed high, considering the Packers just needed a field goal to send the game to overtime. And that’s what they did: they forced the extra period with a 53-yarder as time expired.
The Cowboys effectively moved down the field with short plays as overtime started. And then Prescott threw a deep ball to Tolbert to pick up 34 yards and get within five yards of scoring. After that play, ESPN Analytics gave the Cowboys a 93 percent chance of winning. It felt like a guaranteed touchdown, but yet again, this is the NFL. Sure enough, the next three plays were a run for one yard, a sack, and an incompletion. They settle for the field goal. Unacceptable. Rather than needing a touchdown to escape with a mere tie, because the Cowboys couldn’t finish the job, the Packers needed a field goal and had a chance to win with a touchdown. I have never played pro football, but I imagine those two situations have immensely different impacts on the psychology of competing and approaches to calling plays.
Nonetheless, the Cowboys had a 76 percent chance of winning after the field goal. It seemed high to me, but that’s beside the point. And luckily, the Packers couldn’t find the play they needed. In fact, they struggled to put plays together; they got down to fourth-and-6 on their own 24 yard-line. On that play, the Cowboys had an 87 percent chance of winning. They converted on that play, but incompletions and negative plays haunted them. They ended up settling for the field goal as time expired, escaping Dallas with a tie.
Final: Tie, 40-40