Pittsburgh lost in Cleveland as an 84 percent favorite, and our model cut the Steelers to nine and a half wins with the title odds gone
Updated
Pittsburgh went to Cleveland as the model's 84 percent favorite and lost 27-24, and that one result cost the Steelers more in the projection than any game this season. When we ran the rest of the season Pittsburgh came out at 9-7 with a 55 percent playoff chance and half a percent title chance, down from 10.3 wins, 69 percent and three percent after Week 3. The Steelers are 2-2 with wins over Atlanta and Cincinnati and losses in New England and Cleveland, and the two losses came against teams the model had them beating. That pattern, more than the record, is what moved the number.
The offense has been the issue in both losses. Aaron Rodgers is 88 of 153 for 999 yards with seven touchdowns and four interceptions, an 82 rating through four games, and has been sacked 13 times, which is the most of any quarterback in the league. Our model projects him for 4,340 yards and 25 touchdowns against 14 interceptions, and the sack pace is the number that is dragging everything else. Jaylen Warren has 309 rushing yards and no touchdowns and projects for 1,115 yards. Rico Dowdle missed Cleveland with a toe injury and his status this week is uncertain. D.K. Metcalf has 16 catches for 213 yards and projects for 97 catches and 1,257. Roman Wilson has two scores and projects for 54 catches. Pat Freiermuth has two touchdowns on 14 catches. The unit projects for 349 points, 19th in the league.
The defense is still the reason the Steelers project for nine and a half rather than eight. Pittsburgh projects to allow 325 points, seventh in the league, and the seventh-fewest yards. T.J. Watt has five sacks and an interception in four games and projects for 14 and a second pick. Alex Highsmith has three and projects for twelve. Keeanu Benton projects for six and a half from the interior. The team projects for 46 sacks, seventh in football. Robert Spears-Jennings has gone from a four-tackle projection to 73 as the rookie safety has taken a starting role, and Jaquan Brisker projects for 84 stops and two picks. The unit gave up 27 in Cleveland to a Browns offense the model has 30th in the league. That is the game it has to answer for.
The division is tighter than the roster quality suggests. Baltimore is 3-1 and projects for eleven wins with Lamar Jackson's ankle the only variable. Cleveland is 3-1 and projects for seven, holding the head-to-head over the Steelers. Cincinnati is 2-2 and projects for ten. When we ran the rest of the season Pittsburgh won the AFC North in a small minority of simulated years, down from a real share a week ago, and finished third more often than second. The Browns rematch in Week 8 and the Ravens and Bengals games later in the year are where that changes if it changes.
The next month is as kind as the schedule gets, and it has to be used. Indianapolis at home at 70 percent, a trip to Tampa Bay at 70, a trip to New Orleans at 74, Cleveland at home at 78, and a bye in Week 9. When we ran that stretch the Steelers went 3-1 more often than anything else and 4-0 often enough to matter. A 6-2 record entering November would restore the ten-win projection and the title odds with it. A 4-4 record would confirm what the Cleveland game suggested.
The player lines are about a quarterback under pressure. Rodgers' projection climbed from 4,119 yards to 4,340 and from 22 touchdowns to 25, with his interception line up from twelve to 14 and his rushing projection from 26 yards to 63, which is the model noticing him scrambling. Metcalf's projection rose from 95 catches to 97 and from 1,105 yards to 1,257. Roman Wilson's touchdown line went from three to five. Pittman's dropped from 71 catches to 65. Warren's rushing projection fell from 1,169 to 1,115 and Dowdle's from 384 to 338 with the toe. On defense, Highsmith's sack projection dropped from 13.4 to 12.2 and Watt's held at 14.1 with his interception line at 1.2. Spears-Jennings' tackle projection went from four to 73. Holcomb's climbed from 75 to 88. The number to watch is the sacks allowed. Thirteen in four games is the most in the league, and the model's 349-point offensive projection assumes the pace slows. If it does not, nine and a half wins is generous.
The matchup record is two wins the model expected and two losses it did not. We gave Pittsburgh 46 percent against Atlanta and the Steelers won. We gave them 50 in New England and they lost 20-3. We gave them 54 against Cincinnati and they won by three. We gave them 84 in Cleveland and they lost. Two wins from about two and a third expected is close to even, but the shape matters: the Steelers have beaten the teams the model had as coin flips and lost the one it had as a lock. The market opened Pittsburgh at nine wins. Our model sat a half-win under it in July, a win and a half over it after Week 3, and a half-win over it now. The lean is the defense. The sack count is what keeps it from being larger.
