Sign in
NFL
Write-ups
Write-upsPlayersBacktestHow this works

Week 3 preview

Sep 22 · predictions are published to show the model works. nothing here is betting advice.Last updated: predictions Oct 1, 11:02 AM ET · results graded Sep 28, 11:22 PM ET
Written by our system from the week's data, with the same language model that reads the news; no human edited it. Every number comes from the Record and Week 3 pages. If a sentence reads wrong, tell us.

The model grades itself honestly, which means it knows the market beats it about 56 to 44. This week we have 16 games to test that claim.

Where we differ from the market

Two games stand out where our read diverges sharply from the betting price.

In CAR @ CLE, we have Cleveland at 54% while the market favors Carolina by 2.5 points. Cleveland's home field is worth 8 points in our model, but Carolina's better recent form costs them 3, and Cleveland's more efficient offense adds 3 more.

In MIN @ TB, we have Tampa Bay at 51% while the market favors Minnesota by 1.5. Tampa is the better team overall by 11 points in our view, but Minnesota's home field is worth 8, and Tampa's better form adds 6. This one is nearly a coin flip.

History says the market wins more of these than we do, so these are the games that test the model, not the ones to be sure about.

The confident ones

Five games grade out well above 70 percent:

  • Green Bay Packers over Atlanta Falcons at 72%: GB has the QB edge, is the better team, and plays at home.
  • Buffalo Bills over Los Angeles Chargers at 86%: BUF is the better team, has the QB edge, and plays at home.
  • Detroit Lions over New York Jets at 83%: DET is the better team, has the QB edge, and plays at home.
  • Seattle Seahawks over Washington Commanders at 76%: SEA is the better team by a wide margin, is in better form, and this one is notable because Washington plays at home but that advantage is overwhelmed.
  • San Francisco 49ers over Arizona Cardinals at 78%: SF is the better team, has the QB edge, and is in better form.

The coin flips

Four games fall between 51 and 54 percent, where the model has almost no edge:

  • CLE at 54% over Carolina
  • HOU at 54% over Indianapolis
  • TB at 51% over Minnesota
  • LA at 51% over Denver

Injury adjustments

The news layer made meaningful moves in several games. Seattle's pick moved 5.7 points toward the Seahawks after Washington's QB was ruled out. Buffalo's pick moved 4.8 points toward the Bills after the Chargers' 0-2 start and reported morale issues. Detroit's pick moved 3.8 points toward the Lions after New York's injuries mounted. Tampa's pick moved 7.5 points toward the Buccaneers after Minnesota's running back went to injured reserve.

Every pick

Thu 8:15 PM · ATL @ GB: GB 72% · projected 21–26 Sun 1:00 PM · CAR @ CLE: CLE 54% · projected 21–22 Sun 1:00 PM · CIN @ PIT: CIN 60% · projected 25–21 Sun 1:00 PM · HOU @ IND: HOU 54% · projected 23–22 Sun 1:00 PM · KC @ MIA: KC 63% · projected 27–21 Sun 1:00 PM · LAC @ BUF: BUF 86% · projected 18–31 Sun 1:00 PM · NE @ JAX: JAX 59% · projected 21–24 Sun 1:00 PM · NYJ @ DET: DET 83% · projected 19–30 Sun 1:00 PM · SEA @ WAS: SEA 76% · projected 26–19 Sun 1:00 PM · TEN @ NYG: NYG 68% · projected 20–26 Sun 4:05 PM · ARI @ SF: SF 78% · projected 18–29 Sun 4:05 PM · MIN @ TB: TB 51% · projected 22–22 Sun 4:25 PM · BAL @ DAL: BAL 65% · projected 28–23 Sun 4:25 PM · LV @ NO: NO 64% · projected 20–25 Sun 8:20 PM · LA @ DEN: LA 51% · projected 24–24 Mon 8:15 PM · PHI @ CHI: PHI 61% · projected 24–22

Predictions are published to show the model works. Nothing here is betting advice.

← All write-ups