Seattle has lost three of four and is 3-3, and our model has dropped the Seahawks from ten wins to seven with the quarterback the reason
Updated
Seattle lost 32-29 at home to San Francisco and 17-14 in Denver, is 3-3, and has lost three of its last four after a 2-0 start. When we ran the rest of the season the Seahawks came out at 7-10 with a 9.5 percent playoff chance, down from 10-7 and 57.5 percent two weeks ago and from 11.4 wins and a 90 percent playoff chance in July. That is nearly a three-win drop, second only to the Rams in this run, and the reason is the position our model has been hedging on since September.
Sam Darnold has started five of the six games and is 90 of 149 for 1,033 yards with six touchdowns and seven interceptions, a passer rating of 75, and has been sacked 14 times. Our model projects him for 3,559 yards, 17 touchdowns and 17 interceptions behind an 82.5 rating, down from 22 and twelve and a 95 rating two weeks ago. That is the model watching six weeks of a quarterback and believing what it sees, because it now leans harder on this season's snaps than it did a month ago. Drew Lock projects for 478 yards and a 123 rating in relief, which is the model saying it expects him to play again. Jaxon Smith-Njigba has 49 catches for 717 yards and seven touchdowns, the most receiving yards in the league, and projects for 125 catches, 1,732 yards and eleven scores, which is a career year by any measure and nearly the only thing on the offense that has gone right. Emanuel Wilson projects for 742 rushing yards. The unit projects for 342 points, 23rd in the league, down from 375 and 14th.
The defense has been the better half all season and its grade has slipped with the rest. Seattle projects to allow 354 points, 15th in the league, down from sixth two weeks ago, and the unit gave up 32 to San Francisco at home. The takeaways are still the strength. The Seahawks project for over 15 interceptions, fourth in football, with Julian Love at three and a half and Ernest Jones at two and a half plus 115 tackles and over three sacks. Byron Murphy II projects for nine sacks from the interior. Uchenna Nwosu's projection has dropped from eight and a half sacks to six and DeMarcus Lawrence's from seven and a half to under six as the edge has gone quiet. The team projects for 39 sacks, 22nd.
The division has settled around the 49ers and the Seahawks are a distant second. San Francisco is 6-0 and projects for twelve wins with the head-to-head over Seattle banked. The Rams are 2-4 and project for ten, which is still three more than Seattle. Arizona is 2-4 and projects for six. When we ran the rest of the season Seattle finished second in the NFC West more often than third, which is the record talking rather than the roster, and won the division almost never.
The next month is four home-and-away games the model mostly likes. Kansas City at home at 20 percent, Chicago at home at 60, Arizona at home at 58, and a trip to Las Vegas at 60. When we ran that stretch Seattle went 2-2 more often than anything else, with the Chiefs game treated as a loss and the other three as narrow wins. A 5-5 record entering the second week of November would put seven wins on schedule and the playoffs out of reach barring a run.
The fortnight in detail: a 32-29 home loss to San Francisco in which the defense gave up 32 to a 49ers offense that had not scored more than 36 all year, and a 17-14 loss in Denver in which the offense scored 14 against a Broncos defense the model has twelfth. Seven interceptions in five games is the number that did this. The model had Darnold at twelve for the season two weeks ago; it has him at 17 now, with the rating projection down 12 points, and that single line explains most of a three-win drop. The receivers have not changed. Smith-Njigba is on pace for the best season of any receiver in football. The running game has been adequate. The defense slipped from sixth to 15th, which is a real move but not a three-win one. What moves the number from here is the quarterback. Lock has a 123 rating in relief and the model has him for 55 more attempts, which is its way of saying it expects a change at some point. If that happens and Lock plays the way he has, the projection moves back toward nine. If Darnold starts all eleven remaining games and throws at his current rate, seven is generous.
The matchup record is a team the model has been too fond of all season. We gave Seattle 66 percent against New England, 72 in Arizona, 90 in Washington, 60 against the Chargers, 64 against San Francisco and 70 in Denver. That is more than four and a quarter expected wins from six games, and the Seahawks have delivered three. The three losses were all games the model had Seattle winning comfortably, and one of them was to a team it had at under five wins. The market opened Seattle at ten and a half wins. We sat a win over it in July, on it after Week 1, and three and a half under it now. The gap is the quarterback. If Darnold plays the rest of the season like the first six weeks, seven wins is the number. If Lock takes the job and plays like he has in relief, the model would move again.
