The afternoon window is closed, the couch has a dent in it, and NFL best bets today narrows down to two night games plus a pile of information I did not have this morning. That is the honest shape of Week 1. Every model in the country just got its first real data, and no model should trust it yet.
I also got one wrong already. That is further down. Nothing deleted, nothing cherry-picked.
| Matchup | Final | What I carry forward |
|---|---|---|
| Jets vs Titans | 23-10 NYJ | Breece Hall 102 rushing yards, Geno Smith 215 passing |
| Steelers vs Falcons | 20-13 PIT | Watt's 35-yard pick-6 broke a one-score game |
| Ravens vs Colts | 41-23 BAL | Lamar 324 and a rushing score, Henry three on the ground |
What I'm eyeing for NFL best bets today
- Cowboys at Giants, 8:20 ET tonight. Two teams with zero 2026 tape and a national audience, which is usually where my number and the market's number sit closest together. Turns into a play only if the inactives come out and the posted total disagrees with mine by more than a field goal's worth of process. Otherwise I am a viewer.
- Broncos at Chiefs, Monday 8:15 ET. Division opener, Kansas City at home, and the market has spent a decade paying a tax for that address. Turns into a play if the Monday designations come out clean for Denver's front and the number still treats this as a comfortable Chiefs night.
- Baltimore's receiver room. They hung 41 on Indianapolis and then walked off with the rookie Ja'Kobi Lane ruled out with a wrist and Zay Flowers questionable with a hamstring. If Flowers misses real time, Baltimore's next number gets built on Henry and Lamar's legs. I want to own my opinion on that before it is common knowledge.
- Tennessee, one week from now. Robert Saleh's debut against his old team ended 23-10 the wrong way, and 23-10 losses in openers get turned into identity by Wednesday. Turns into a play the moment a market prices that afternoon as a permanent condition.
- The Rams without Myles Garrett. Arthroscopic knee surgery, off to IR. That is a whole pass rush's worth of gravity leaving a defense, and the first posted number that has not fully discounted for it is the softest line on my board. Reading the injury report first is not a strategy. It is just homework.
The one I lean on most
Denver, plus whatever the points end up being, in Kansas City. That is a lean and only a lean, because I have not seen a posted number to measure my own against, and a lean that pretends otherwise is a tout move. If the Monday injury report treats both sides fairly, I expect my number and the board to have an argument. I like arguments.
Right side, wrong night in Pittsburgh
I had Atlanta's moneyline at +250 against a fair price of 234, call it 4.65 percent of edge, and I liked it as strong. The Falcons showed up short-handed, kept it to one score into the fourth, and then Watt jumped a Cooper Rush throw and took it 35 yards the other way. Right side, wrong night. The process that finds 4.65 percent twenty more times this season is the same process that just ate a pick-6, and I am not changing it because one bounce went the other way.
from the almanac: marcus allen sat third on the all-time rushing touchdown list for the better part of three decades, and derrick henry walked past him on a sunday afternoon in indianapolis. one beat of wonder, then back to work.
What I am watching next: the Ravens' Friday designation on Flowers, and whether Denver's number Monday looks anything like mine. Keep it 21+, keep it entertainment, and keep the variance in mind. You know where to find me.