Week 8 Model Picks

Week 7, if you had bet the posted card: 1-0 on spreads, 0-3 on parlays, -1.15 units. The 2-unit Giants bet at +10 did not need the points. New York won outright, 27-21 in Houston, so “ten is too many” turned out to be an understatement. The parlays gave it all back. Carolina killed the first one by losing 43-13 to Tampa Bay. Baltimore (35-13 to Cincinnati) and the Jets (28-25 to Miami) both killed the second. San Francisco killed the third, losing 29-20 in Atlanta. Season to date: about -9.8 units. Here is Week 8, on the same rules. A spread has to clear a raw edge of 4.0. Totals stay on the page and off the card. Three 1-unit moneyline parlays. The budget is eight units. This week spends three.

Our Bets

Spreads (0 units)

None. The first run of The Model had one: Eagles +6 at Washington, a 4.4-point edge. It was built on the wrong quarterback. Our data source still listed Jalen Hurts as Philadelphia’s starter, with a 55.2 QBR. Hurts is out for the season with blood clots in his lungs, and Tanner McKee is starting his third game, with a 35.4 QBR over his first two. We put McKee in, refit, and the number moved to Washington by 2.2 against a 6-point line. That is a 3.8-point edge, which misses the bar. The honest card has no spread bet, so we are not making one.

Parlays (1 unit each)

  • Chiefs ML (-118) / Steelers ML (-192) / Rams ML (-278). Pays +282.
  • Bills ML (-170) / Lions ML (-170) / Cowboys ML (-325). Pays +230.
  • Seahawks ML (-166) / Packers ML (-192) / Patriots ML (-310). Pays +222.

The flag on the parlays is Buffalo. The Model’s Baltimore row is Tyler Huntley at 61.2, and Lamar Jackson is expected back from his high ankle sprain. If Jackson plays, the Ravens are better than The Model thinks, and the Bills leg is the one we trust least. Jacksonville (-485) and Cincinnati (-625) are too short to be parlay legs. Seattle’s row is Sam Darnold this week, so the Lock problem from earlier in the season is fixed.

No Thursday spread. Carolina at Green Bay is Packers -3.5, and The Model calls it Green Bay by 3.3. Bryce Young and Jordan Love have nearly identical numbers in the file (57.9 and 58.3), and The Model is a quarter of a point off the market. No edge, no bet.

The Model

Same ensemble: Ridge, LightGBM, XGBoost, and a small neural net, refit on games through Week 7, using team stats, EPA, and starter QBR. Week 7 went 9-5 straight up and 6-7-1 against the spread. On the season it is 50-55-2 against the spread, and it is still behind the market on margin error, 9.9 to 9.6.

The Model picks for Week 8, 2026: per-model margin of victory, pick, DraftKings spread, edge, win probability, and moneyline for all 14 games

The table is the full 14-game card, with McKee in for Hurts: per-model margin, pick, DraftKings spread, raw edge, straight-up win probability, and moneyline. Four teams are on bye: the Giants, Houston, New Orleans, and San Francisco. For the second week in a row, The Model and the market pick the same winner in all 14 games. Ten of the fourteen edges are under 1.6 points.

Over/Under Picks

Totals went 7-7 in Week 7 and are 45-62 on the season. We are showing the numbers and not betting them. Two Unders stand out. Arizona at Dallas is 54.5 and The Model says 48.2. Baltimore at Buffalo is 54.5 and The Model says 49.0, though a Jackson return could change that one. Still not bets.

The Model totals for Week 8, 2026: predicted team scores, model total, DraftKings over/under, and difference for all 14 games

Where we disagree with Vegas

  • PHI at WAS: WAS -6; model MOV +2.2 Commanders. Off the card at 3.8. The Eagles are 2-5 and have lost five straight, and McKee has a 35.4 QBR in his two starts. The Model still thinks six is a lot for a Washington team it rates only a little better. Jayden Daniels practiced in full with his elbow.
  • LAC at LA: LA -6.5; model MOV +9.5 Rams. Off the card at 3.0. Justin Herbert’s 31.3 QBR is second-lowest on the board, ahead of only McKee, and Matthew Stafford is at 65.0. That gap is most of the story.
  • CLE at PIT: PIT -3.5; model MOV +5.6 Steelers. Off the card at 2.1. Deshaun Watson (45.1) against Aaron Rodgers (50.1) is not much of a gap, so The Model likes Pittsburgh for the rest of the roster.

Remember, the best way to use this is as one input into your own thinking. The Model does not know who is holding out, and it only knows the quarterback our data source lists. This week that bit us hard enough to erase the only spread bet on the card. Baltimore is the open question: Huntley is in the file and Jackson is probably on the field.

These are our personal bets and opinions, not betting advice. Wager at your own risk and within your means.

Good luck everyone.

Leave a Reply

Your email address will not be published. Required fields are marked *