NFL Week 10 reminded us momentum doesn’t always follow logic — and neither does win probability, What looked inevitable on paper unraveled in real time, as small moments became catalysts for chaos. The Colts looked safe until a small special teams play tilted the field. The Giants watched a 95 percent win probability vanish in minutes. The Jaguars managed to turn a 98 percent chance into a stunning collapse.

Here’s how each of those swings unfolded:

Indianapolis Colts vs Atlanta Falcons – ESPN Analytics Win Probability Chart

When Jonathan Taylor ran 83 yards for his second touchdown of the day with six minutes left, it looked like the Indianapolis Colts would run away with it. They were up against the sub-par Atlanta Falcons and they had a 22-17 lead. Even after failing on the two-point conversion attempt, they were set up to leave Berlin with the win — a defensive stop and a field goal would have done it. ESPN Analytics gave them a 84 percent chance of closing it out.

But then on the ensuing kickoff, Falcons returner Jamal Agnew took it from Atlanta’s 1 yard-line to the 29. It didn’t seem like much in the moment, but it gave the Falcons a head start, if for no other reason than not having to start the drive in their own end zone. And then much because of that return, the Falcons were in Colts territory after just two plays from scrimmage. In other words, even though ESPN Analytics still gave the Colts a 72 percent win probability, the Falcons seemed to be in a groove. Sure enough, they kept getting 7-, 8-, 9-, 10-yard plays until Tyler Allgeier found painted turf and put the Falcons ahead 25-22 with 1:44 left. That flipped the numbers — after the score, the Falcons had a 71 percent win probability.

The Colts struggled on the following possession, backing themselves up to third-and-21 just behind midfield after Daniel Jones was sacked for a significant loss. Before the play, the Falcons’ win probability was 89 percent, but then Jones scrambled for 19 yards on third down and converted on fourth down. That conversion brought some momentum to the Colts; it brought the Falcons’ chances of winning down to 56 percent. And then it happened again — the Colts had an incompletion on first down and sack on second to bring up third-and-long. It was costly in the moment; rather than having a chance to take the lead late, they needed to pick up a few yards and try to get the game-tying field goal. They did with 25 seconds left, essentially forcing overtime.

The Colts owned overtime. When it began, ESPN Analytics gave the Falcons a slight edge win probability-wise, but the Colts defense shut them down. They had two incompletions, a negative run, a penalty, and a sack — and then they punted. That opened the door for Taylor to win it with his third touchdown of the game, a fitting end.

Final: Colts win 31-25

New York Giants vs Chicago Bears – ESPN Analytics Win Probability Chart

With 5:06 left, the New York Giants had all the momentum. Up 20-10, they seemed to have the Chicago Bears on the ropes — Caleb Williams had just thrown a deep incompletion to Rome Odunze on first down, and they just needed to play clean, and they’d be able to close it out. ESPN Analytics said they had a 95 percent chance of doing so. 

But then Williams and his poise shined through. On the next two plays, he connected with Odunze for an 18-yard play and then scrambled for 29 yards to get the Bears on the 2 yard-line. And then on the following play, he and Odunze linked up for a touchdown to cut the deficit to 20-17 with under four minutes left. It was just a hiccup for the Giants. Right? The analytics thought so; the G-Men still had a 78 percent win probability.

Wrong. When they got the ball, with Russell Wilson in to replace Jaxson Dart, who was being evaluated for a concussion, the Giants did nothing on offense. After a 5-yard run and two sacks, they punted, and their momentum was dwindling bit by bit — as the Bears received the punt, their win probability dropped to 65 percent.

That’s when Williams put on his cape one more time. Just before the two-minute warning, he delivered a pass to Luther Burden III, who ran nine yards into the red zone. After that play, the Bears had a 51 percent win probability. Two plays later, Williams did it with his legs, rushing 17 yards into the end zone to put the Bears ahead 24-20 with 1:47 left. The score gave the Bears an 84 percent win probability in the moment — and it catapulted them to the win and the Giants to their fourth loss in a road game where they led by 10 or more points.

Final: Bears win 24-20

Houston Texans vs Jacksonville Jaguars  – ESPN Analytics Win Probability Chart

At the beginning of the fourth quarter, the Jacksonville Jaguars had the Houston Texans in desperation mode. Leading 29-10, they had Davis Mills — who replaced CJ Stroud after he sustained a possible concussion earlier in the game — in a must-have fourth down situation on their own 17 yard-line. It seemed like an impossible situation to escape,especially for a backup quarterback. No room for error. According to ESPN Analytics, Jacksonville’s win probability was above 98 percent. It was a sure-thing. 

But with a short pass to Jaylin Noel, Mills converted the first down. It set up Mills to connect with Jayden Higgins for a 12-yard touchdown a few plays later to cut the deficit to 29-18. But it didn’t do much for win probability — the Jags still had a 96 percent chance of winning. That’s because win probability calculates what momentum does in real-time and predicting what followed would have been impossible for even the computer models.

The Jags’ disaster ramped up on the ensuing kickoff in the form of an unnecessary roughness penalty that backed them up inside their 15. After a 5-yard run by Travis Etienne Jr on first down — following the play, their win probability was at 93 percent — they stalled on second and third down. They lost four yards on a run, conceded a sack, and had to punt on fourth-and-14, unacceptable drive for a team with momentum that was already vanishing.

Mills opened the next possession with a 22-yard pass to get to midfield. And five plays later, on a third-and-3 play from the 7 yard-line, he connected with tight end Dalton Schultz for another touchdown — and after the failed two-point try, it was 29-24, a one-possession game with 7:29 left. Houston officially had earned a fighter’s chance and Jacksonville’s win probability shrunk to 78 percent. 

One would think the onslaught of points from the Texans would’ve led the Jaguars to get their act together, but they again went three-and-out. Again, though, they still had the momentum, a 81 percent win probability. The Jacksonville defense, had a chance to make up for the past quarter or so of bad football — they had the Texans at third-and-10 on their own 7 yard-line. But then they gave up another big play to Schultz, this time a 20-yard catch. A few plays later, as the  Texans continued to move down the field, the Jaguars did the unthinkable. A pass interference penalty on the 2 yard-line. Momentum completely flipped. What was a 63 percent win probability for the Jaguars became exactly that — a 63 percent win probability — for the Texans.

The Texans got through eventually, but not without struggle. On the first play from the 2, tackle Trent Brown moved early, a five-yard penalty. Two plays later, center Jarrett Patterson did the same thing. Another five-yard penalty to back the Texans up to the 14 yard-line with 41 seconds left. Inexcusable sequence. In fact, ESPN Analytics said momentum had swung back to the Jaguars — they had a 75 percent win probability ahead of second-and-goal on the 14. But on third down, Mills did it. He ran 14 yards for the touchdown to give the Texans the 30-29 lead.

Momentum flipped again. After the score, ESPN Analytics gave the Texans a 93 percent win probability. Good defense would have closed it out — and when it mattered most, they got it. They pinned the Jaguars behind midfield, and on the last play of the day, they forced a fumble and returned it 32 yards for the touchdown to cap off the 

Final: Texans win 36-29