New England won in Buffalo with Maye throwing for 29 points, and our model has moved the Patriots to nine wins and the division favorite
Updated
New England went to Buffalo as the model's slight favorite, which surprised people, and won 29-26, which surprised fewer of them by the end. The Patriots are 2-2, have beaten the Steelers and the Bills, and have lost to the Seahawks by three and the Jaguars by 29. When we ran the rest of the season New England came out at 9-8 with a 58 percent playoff chance, up from 8.4 wins and 40 percent after Week 3, and the Patriots are now the model's most likely AFC East champion. A 2-2 team projected to win its division is a statement about the division as much as the team, but it is also a statement about a defense that has been better than its reputation all year.
The offense has been the problem and it was not the problem on Sunday. Drake Maye is 73 of 117 for 854 yards with four touchdowns and seven interceptions, a passer rating of 71 through four games, and our model projects him for 3,869 yards, 21 touchdowns and 19 interceptions across the season. Nineteen is the highest interception projection of any starting quarterback in the league, and it is the line that keeps the Patriots at nine wins rather than ten. Romeo Doubs has 203 yards and two scores and projects for 73 catches and 1,115 yards as the number one. Mack Hollins has 209 yards but left Buffalo with a calf injury, which is the first thing to check. A.J. Brown remains on injured reserve with a high ankle sprain and is eligible to return against the Jets in Week 6. Rhamondre Stevenson projects for 838 rushing yards. The unit projects for 341 points, 20th in the league.
The defense is the reason the model has moved. New England projects to allow 329 points, eighth in the league, and the tenth-fewest yards. Dre'Mont Jones projects for ten sacks, Elijah Ponder for nearly nine, and Cory Durden for seven and a half from the interior. The team projects for nearly 46 sacks, eighth in football, and over 13 interceptions, sixth. Craig Woodson projects for 108 tackles and three picks. Robert Spillane for 112 stops. This is a unit that held Pittsburgh to three and Seattle to 13 and gave up 35 to Jacksonville in the one game that went wrong. The model has had the Patriots' defense graded as a top-ten unit since September. It took the Bills game for the record to catch up.
The division is the opportunity. Buffalo is 3-1 and projects for eight wins after the loss. The Jets are 1-3 and project for nine. Miami is 0-4 and projects for two. New England holds the head-to-head over Buffalo and has not played the Jets yet. When we ran the rest of the season the Patriots won the AFC East more often than any other team, which would have been an absurd sentence in July and a surprising one a week ago. The margin is thin. Nine wins versus eight and a half is not a gap, and the Jets are right there.
The next month is friendly. Las Vegas at home at 48 percent, the Jets at home at 48, a trip to Chicago at 52, a trip to Miami at 94, and Green Bay at home at 60. When we ran that stretch New England went 3-2 more often than anything else, and the Miami game is as close to a certainty as the model offers. A 5-4 record entering November with two division wins banked would make the nine-win projection look conservative.
The player lines are about the defense catching up to the grade. Maye's projection barely moved, from 3,836 yards and 19 touchdowns to 3,869 and 21, with the interception line stuck at 19. Doubs' touchdown projection climbed from five to seven. Hollins' yardage projection rose from 685 to 732 before the calf injury, and Stevenson's rushing line dropped from 916 yards to 838. On defense, Elijah Ponder's sack projection fell from 12.5 to 8.7, the largest downward move on the roster, and Dre'Mont Jones' from 11.2 to 10.3, while Spillane's interception projection more than doubled to nearly two and his tackle line rose from 98 to 112. Woodson's tackle projection dropped from 117 to 108. The number to watch is the points allowed, which fell from 331 to 329 and holds at eighth in the league. The model has had this defense graded in the top ten since September. The Buffalo win is the first time the record has agreed with it for two weeks running.
The matchup record is a team that keeps surprising the model in both directions. We gave New England 34 percent in Seattle and the Patriots lost by three. We gave them 50 against Pittsburgh and they won 20-3. We gave them 38 in Jacksonville and they lost by 29. We gave them 54 in Buffalo, which was the surprise, and they won. Two wins from about one and three-quarters expected is a mild overperformance, and the two wins came against the two teams in the group with the better rosters. The market opened New England at eight and a half wins. Our model sat a half-win over it in July, a bit under it after Week 3, and a half-win over it now. The lean is about a defense that has been real for a month and a division that has opened up. Maye's interceptions are the thing that could close it again.
