San Francisco is 6-0 after winning in Seattle, and our model now has the 49ers at twelve wins and the NFC West favorite for the first time
Updated
Six games, six wins, and the one the model least expected was the one that mattered most. San Francisco went to Seattle as a 36 percent underdog and won 32-29, then held off Washington 23-17 at home, and the 49ers are the last unbeaten team in the NFC. When we ran the rest of the season this week they came out at 12-5, averaging twelve wins, with a 94 percent playoff chance and a three and a half percent title chance. Two weeks ago those numbers were 10-7, 68 percent and half a percent. In July they were 6-8 and eight percent. Only Buffalo has climbed as far since the summer, and this is the first run all season in which the model has San Francisco winning the NFC West more often than Los Angeles.
Part of the move is two wins. Part of it is that our model now leans harder on this season's snaps than it did a month ago, and six weeks of Brock Purdy at this level is a lot of evidence to lean on. He is 119 of 171 for 1,460 yards, 14 touchdowns and two interceptions, a passer rating of 118, and has been sacked five times after going four games without one. Our model projects him for 4,038 yards and 30 touchdowns behind a rating near 110, up from 106 two weeks ago. Christian McCaffrey projects for 996 rushing yards and twelve touchdowns, up from nine, plus 66 catches. Mike Evans has climbed to 79 catches and 1,206 yards. George Kittle projects for nine touchdowns. The offense projects for 397 points, tenth in the league, and the 49ers have scored 177 in six games, so the model still has the pace slowing.
The defense is the half the summer sim got most wrong, and it has allowed 110 points in six games. Our model projects the unit giving up 312 points, fifth in the league, which is a long way from the 424 it had in July and the 331 it had two weeks ago. Fred Warner projects for 127 tackles. Renardo Green's interception projection has nearly doubled to five, the biggest single move on the roster. Keion White projects for nine and a half sacks and Nick Bosa for under nine, with Bosa's line still carrying the games the calf cost him. The team projects for under 39 sacks, 23rd, which is the one soft number on the sheet, and Seattle was the first game in which a good quarterback made it matter. The 49ers gave up 29 that day and still won.
The division has been turned on its head and the model has stopped resisting. Los Angeles is 2-4 after losing at home to Buffalo and Arizona, and projects for ten wins with its title odds down to 8.5 percent from 44 in July. Seattle is 3-3 and projects for seven. Arizona is 2-4 at six. San Francisco is four games clear of the Rams with the head-to-head banked and three games clear of anyone else, and when we ran the rest of the season the 49ers won the NFC West in a clear majority of simulated years. That sentence has not been true in any run before this one.
The next month is three games and a bye. A trip to Atlanta at 52 percent against a Falcons team that has lost two straight at home, a bye in Week 8, Las Vegas at home at 56, and a trip to Dallas at 50. When we ran that stretch San Francisco went 2-1 more often than anything else, with the model treating all three as close to even. An 8-1 record entering the second week of November would push the projection toward 13 and the title odds into the conversation with Detroit and Kansas City.
The fortnight in detail is worth a paragraph, because the two games were different kinds of win. In Seattle the 49ers trailed a 3-1 team on the road, gave up 29 points to a Seahawks offense that had been struggling, and won anyway. That is the kind of game the model expected San Francisco to lose, and it is the kind of game the model had been waiting to see before it would move the division call. Against Washington the 49ers won 23-17 at home against a Commanders team playing its first game with Jayden Daniels back, and it was the first time all season the offense scored under 24 points. The model noticed both. The yardage projection for Purdy dropped 30 yards and the rating projection rose three points, which is the model expecting fewer throws and better ones, and the points-for projection climbed 16. What moves the number from here is simple. The 49ers play two of their next three on the road, and the model has them as a slight favorite in all three. Win two and twelve wins holds. Win all three and the title odds, which sit at three and a half percent because Detroit and Kansas City have the two best rosters in the model's eyes, start to climb toward the teams the record says San Francisco belongs with.
The matchup record is the clearest case in the league of a model being dragged along by a team. We gave the 49ers ten percent in Melbourne, 66 against Miami, 72 against Arizona, 64 against Denver, 36 in Seattle and 66 against Washington. Six wins from a little over three expected is the second-largest overperformance in football behind Buffalo, and the Seattle game, on the road against a team the model had at ten wins, was the one that finally moved the division call. The market opened San Francisco at ten wins. In July we sat more than three under it. We now sit two over it, and the gap has flipped direction entirely. The model bought the names in October after watching them play for six weeks, and Atlanta on Sunday is the first road game since it did.
