Mostly Pass
Most games aren't bets.
Betting research from the bookmaker's side of the counter: price first, prediction second
Most games aren't bets.
Daily research on NFL betting prices: what the market implies, where the vig lives, where a transparent model disagrees, and why most games are a pass. Every signal timestamped. Every result published.
No locks. No screenshots. No deleted losses. Every signal is PAPER until a model earns the right to call something a bet.
Get the daily research by email (free)
Mostly Pass: NFL Week 4 research, 2026-10-02
Most games aren't bets. Today's card: 0 BET · 4 LEAN · 11 PASS across 15 games. All signals are PAPER, timestamped before kickoff, and graded automatically against the Pinnacle close.
Bets
None. Our model has not beaten NFL closing lines out of sample, so it is not allowed to call anything a bet. What follows is research.
Where the model disagrees on the spread (paper signals)
- Buccaneers +3.5 -108 (draftkings): Pinnacle fair -104, model -152, would clear the BET bar at -134 or better
- Texans -3 -106 (lowvig): Pinnacle fair -101, model -134, would clear the BET bar at -119 or better
- Patriots +7 -112 (lowvig): Pinnacle fair -106, model -123, would clear the BET bar at -109 or better
- Raiders +4.5 -108 (draftkings): Pinnacle fair -107, model -118, would clear the BET bar at -104 or better
Every game: market vs model
| Game | Pinnacle spread (home) | Model spread (home) | Gap (pts) | Verdict |
|---|---|---|---|---|
| Patriots @ Bills | -7 | -5.2 | -1.8 | LEAN |
| Packers @ Buccaneers | +3.5 | PK | +3.5 | LEAN |
| Cowboys @ Texans | -3 | -5.5 | +2.5 | LEAN |
| Chiefs @ Raiders | +4.5 | +3.1 | +1.4 | LEAN |
| Colts @ Commanders | +4 | -1.9 | +5.9 | PASS |
| Jets @ Bears | -3 | -10.1 | +7.1 | PASS |
| Rams @ Eagles | +3.5 | -0.7 | +4.2 | PASS |
| Jaguars @ Bengals | -2.5 | +1.2 | -3.7 | PASS |
| Cardinals @ Giants | +2.5 | -3.1 | +5.6 | PASS |
| Titans @ Ravens | -11.5 | -12.3 | +0.8 | PASS |
| Dolphins @ Vikings | -11 | -10.9 | -0.1 | PASS |
| Chargers @ Seahawks | -7.5 | -11.8 | +4.3 | PASS |
| Broncos @ 49ers | -2.5 | -2.1 | -0.4 | PASS |
| Lions @ Panthers | +3.5 | +3.4 | +0.1 | PASS |
| Falcons @ Saints | -3 | -1.1 | -1.9 | PASS |
Why we're passing
Dolphins @ Vikings: Pinnacle has Vikings -11; our model makes it -10.9. The best model edge on either side of the spread is 0.3%, under the 1.5% bar for even a paper signal. The price is fair, so not betting is the default.
When the model is probably wrong
A big gap between our model and the sharp market usually means the model is missing information, such as a quarterback change or an injury, not that we found a mispricing. These games are automatic passes:
- Colts @ Commanders: market +4, model -1.9 (gap +5.9 pts)
- Jets @ Bears: market -3, model -10.1 (gap +7.1 pts)
- Rams @ Eagles: market +3.5, model -0.7 (gap +4.2 pts)
- Jaguars @ Bengals: market -2.5, model +1.2 (gap -3.7 pts)
- Cardinals @ Giants: market +2.5, model -3.1 (gap +5.6 pts)
- Chargers @ Seahawks: market -7.5, model -11.8 (gap +4.3 pts)
The record
BET: none (the model has not earned BETs) · LEAN paper signals published: 0, none graded yet
How this works: https://mostlypass.pages.dev/methodology/ · Every row: https://mostlypass.pages.dev/record/record.csv Research, not betting advice. Paper signals only. 21+. Gambling problem? Call 1-800-GAMBLER.