Chicago at Seattle

Seattle is favored by 3 over Chicago at home on Monday night, and the model has Seattle by 8.9, a 5.9-point disagreement in favor of laying the points. Where that projection lands matters as much as its size. Three is the most common final margin in football, 14.7% of the 2,068 settled games in the model's training data, and seven is next at 8.5%. The 8.9-point projection clears both.

Seattle is the top-rated team in football, 1st in opponent-adjusted net EPA at +0.369, and Chicago is 8th at +0.138. The difference is defense. Seattle's ranks 1st at -0.294, where lower is better, and Chicago's ranks 10th at -0.040. Seattle's special teams rank 2nd at +0.257 against Chicago's 16th at +0.125. Chicago is better in two places: its offense ranks 8th at +0.098 against Seattle's 11th at +0.075, with success rates of 47.6% and 46.8%, and its turnover margin of +0.333 ranks 6th, while Seattle's -0.176 ranks 23rd. Seattle's rating is built on its defense, not on turnovers.

Results have lagged the rating. Seattle is 4-3, and all three losses came by three points or fewer: 33-31 at Washington, 32-29 to San Francisco, rated 6th, and 17-14 at Denver, rated 9th. Its wins came 13-10 over New England, rated 3rd, 31-7 at Arizona, 30-23 over the Chargers and 20-3 over Kansas City last week. It is 3-1 at home and has outscored its opponents 168-125. Chicago is 5-2 and has won two straight, the latest 30-24 over New England. Its losses came 9-3 to Minnesota and 31-7 at Green Bay; its other wins came over Carolina, Philadelphia, the Jets and Atlanta, rated 26th, 16th, 30th and 20th. Chicago has outscored its opponents 180-143.

This is the only path from an injury to the projection, so it is worth showing rather than describing. The model reads the expected starter’s rating:

When a starter is ruled out, the depth chart promotes his backup, so the play probability above reads 100% and the post-model quarterback adjustment does not fire. The absence shows up in the rating instead, which falls to the backup’s — and that rating is one of the twenty features the model selects. A play probability below 100% with no designation filed is the default weighting applied to any unfiled starter, not a report that he is hurt.

Every input is perturbed by the standard error of a full 17-game season average and the game re-projected, four hundred times. The question each trial answers is whether the model still favors SEA -3 after the perturbation.

It does in 100.0% of trials at that realistic noise level, and 99.8% when the noise is doubled.

The other side of this game is priced too — CHI +3 — and the model puts it at 48.8% against 51.2% implied by its own price. It is the same projection seen from the other end, not independent evidence — preferring this side and declining that one are one decision, not two.

The side held in 100% of trials and 99.8% under doubled noise. Put plainly: every input to the model was nudged at random by the standard error of a season's average, 400 times over, and Seattle stayed ahead of 3 every time. At twice that noise, 1 trial in 400 flipped.

SEA opened at -4.5 and sits at -3: the market has moved 1.5 points toward CHI. 40% of tickets and 49% of the money are on this side.

The move came with most of the tickets on Chicago, so it is not reverse line movement, and the splits do not show what moved it. On Seattle, the money share runs nine points ahead of the ticket share, so the Seattle bets have been larger on average. The move also changes the bet itself. At -4.5, Seattle had to win by five or more. At -3, a four-point win covers and a three-point win, the most common margin in football, is a push. This is a Monday game, so the line has two more days to move.

Public betting splits from Action Network, captured 2026-10-30, across 9,829 tracked bets. They describe the market; they are not an input to the model.

Saturday's final report for the Monday game moved several names on both sides, and none of it reaches the projection: the model's only injury input is the quarterback adjustment, and it does not fire. Drew Lock is questionable with an elbow injury after a limited Friday, but Sam Darnold has taken nearly every Seattle dropback since week 3 and is the established starter; Lock's figure is the workload he carried in weeks 1 and 2. Caleb Williams is healthy. Seattle is the more depleted side at 1.836 against Chicago's 1.225, and the offensive line takes the largest hit: tackle Charles Cross is out with an ankle injury at 19.4% of the tackle snaps, and guard Anthony Bradford is questionable at 19.9% of the guard snaps after missing practice. Receiver Rashid Shaheed is out with a knee injury at 13.5% of the receiver snaps, linebacker Uchenna Nwosu is doubtful at 13.7% of the linebacker snaps, and running back Jadarian Price remains on reserve at 27.8% of the running-back snaps. The news runs the other way on defense: Leonard Williams, at 63.9% of the defensive tackle snaps, and linebacker Drake Thomas practiced in full on Friday and carry no designation. Chicago lost running back Kyle Monangai on Saturday, ruled out at 40.0% of the running-back snaps, and is without tight end Cole Kmet, on reserve at 15.4% of the tight end snaps. Linebacker D'Marco Jackson is questionable at 23.9% of the linebacker snaps and cornerback Malik Muhammad is questionable at 15.8% of the cornerback snaps; tight end Colston Loveland carries no designation. Chicago also traded for defensive tackle Jeffery Simmons on Thursday. Its ratings come from seven games he did not play for it, so he is not in the 8.9 either.

Final report: Wednesday for a Thursday game, Friday for the weekend. A player ruled out or listed as doubtful counts as not playing; a questionable non-quarterback plays about 62% of the time, a questionable quarterback about 35%. Reserve lists are folded in separately because players on IR never appear on a weekly report.

Depletion is the share of a normal week’s snap workload expected to be missing: each player’s share of his position group’s snaps in prior games, multiplied by his probability of not playing. 1.00 is roughly one full starter’s workload. It is context, not a model input.

The case against starts with the shape. Across 1,391 walk-forward games, laying points at a 3-7 point disagreement has covered 44.1% of the time, below the 52.38% break-even, so a 5.9-point gap is no evidence on its own. Seattle's record has not matched its rating: it is 4-3, with three losses by three points or fewer, two of them on the road. Chicago has won two straight, beat New England, rated 3rd, last week, and has the better offense and the better turnover margin. Seattle plays without Cross and Shaheed, and the market has moved 1.5 points toward Chicago since the open.

Play-by-play from every game since 2019 — roughly 345,000 plays — plus snap counts, depth charts and multi-book closing lines. An ensemble of gradient-boosted trees, a regularized linear model and a random forest predicts the point margin directly; that continuous prediction is converted into a probability against any line using the model’s own out-of-sample error distribution.

Twenty features, ranked fresh each week inside the training window and never using the games being predicted. Validation is strictly walk-forward: every season is predicted using only the seasons before it. Betting odds are never permitted as model inputs — the model is not allowed to see the market it is being asked to beat.

See every Week 8 game on the edge board →

Every game played this season, with the margin. These are results, not ratings: the ranking above is opponent-adjusted expected points added per play, which is why a team’s record and its rank can disagree.