Ten days out, four Week 1 numbers still don't agree with themselves

Ten days before kickoff, most of the Week 1 board has stopped moving. Eight of the twelve Sunday main-slate games have identical numbers across every current source checked. Two haven't — and one of those two isn't a two-way disagreement anymore, it's a three-way one. If you're building an environment read off "the total," it matters which kind of number you're looking at.

The ranking, as of today

Totals and spreads below come from same-day checks against CBS Sports/SportsLine's Week 1 model table and VegasInsider's per-game pages, both dated 2026-09-02 or 2026-09-03. Where sources disagree, both numbers are shown rather than picked.

TB@CIN's shootout shape — the tightest spread among the slate's high-total games — is the cleanest target environment on the board and has been since the market first repriced off its stale May number. ARI@LAC and CLE@JAX sit at the opposite end: double-digit and near-double-digit favorites with total scoring capped below the slate median, the two games most likely to turn into one team running out the clock.

Where the board actually disagrees, and why that's different from "it moved"

BAL@IND and ATL@PIT aren't cases of a number moving — they're cases of three current, same-week sources not agreeing with each other. That's a meaningfully different situation than a line ticking up or down, and it argues for checking again closer to lock rather than picking a side now.

BAL@IND has carried 48.5 consistently since the market first repriced in late August. On 2026-09-03, CBS Sports/SportsLine's table showed 47.5 for the first time in this total's entire history, while VegasInsider's own page still showed 48.5, and a Covers.com check the day before had shown 50.5. Three sources, three numbers, inside 48 hours. 48.5 remains the best-supported figure on weight of history, but three independent reads landing three different places is not the shape of a converged market.

ATL@PIT is worse: both its total and its spread are in dispute at the same time. CBS's table has it at PIT -3/41.5; VegasInsider's own page, checked the same day, has PIT -3.5/42.5. Neither figure has moved toward resolution — the spread dispute is new as of this week, layered on top of a total dispute that's been open since late August.

Both games are also the two least-covered on the slate — neither has drawn a single dedicated write-up from any DFS-industry outlet checked. There's no market correction available to lean on for either one; whatever read you build here should be rechecked, not trusted as final, right up to lock.

What the total doesn't tell you

A total and a spread describe expected scoring. They say nothing about who's actually on the field producing it, and two of this week's better-looking environments have real personnel questions underneath the number.

Carolina's offensive line is not the one that earned CHI@CAR's 47.5. Both of Carolina's starting tackles are out: Taylor Moton is on the Non-Football-Injury list with a blood clot in his lung, and Ikem Ekwonu is still recovering from a torn patellar tendon suffered in January's wild-card loss (ESPN/SI.com, 2026-07-21). Their Week 1 fill-ins are free-agent signee Rasheed Walker at left tackle and first-round rookie Monroe Freeling at right tackle (Panthers.com, 2026-07-21). A total moving up and a quarterback's protection getting worse are two separate facts, and the market's total says nothing about whether it has actually reconciled them.

Washington's line is worse than PHI -4.5 alone implies. Starting left tackle Laremy Tunsil tore his triceps in an August 8th practice and needs surgery, ruling him out for essentially the entire season (CBS Sports/Fox Sports/NBC Sports, multi-outlet, 2026-08-09). That argues for more pressure on Washington's quarterback specifically than a one-score spread alone would suggest.

Two more situations affect confidence in a read without moving a number at all. Atlanta's starting quarterback is still unresolved ten days out — Adam Schefter, on the Pat McAfee Show September 1st: "It's down to Tua and Penix and they don't have a decision yet." Every ATL@PIT read built around a default starter is a read built on an assumption the team itself hasn't confirmed. And Arizona's backfield remains a genuine coin flip: NFL Network's Mike Garafolo put Jeremiyah Love's Week 1 availability at "about 50-50" as of September 2nd, with head coach Mike LaFleur declining to commit either way — worth knowing before treating ARI@LAC's blowout shape as settled at the player level, even though the game's own line (LAC -10.5) hasn't moved in two weeks.

Elsewhere, personnel is more settled than it was a week ago. Miami traded presumptive 2026 starter Quinn Ewers to Jacksonville; Malik Willis is now confirmed as Miami's Week 1 quarterback across multiple outlets, with no dissenting report since. Green Bay's Josh Jacobs is on the Commissioner's Exempt List and cannot play Week 1 — confirmed fact, not risk, since August 30th. New Orleans is without Alvin Kamara for a Week 1 backfield now built around Travis Etienne. None of these change a line by themselves, but each changes who's actually producing the scoring the line implies.

Weather still doesn't exist for any of it

Six of these twelve games are outdoors: Cincinnati, Philadelphia, Carolina, Pittsburgh, Jacksonville, and Tennessee. None has a usable forecast yet. The National Weather Service's forecast horizon, as of a direct check this morning, runs through the night of September 10th — still three days short of the September 13th kickoff. Anything claiming to forecast conditions for these six games right now is a long-range climatological average, not a real prediction. That will change in the next several days as the horizon closes the gap; until it does, treat every outdoor game above as weather-unknown, not weather-neutral.

The honest summary

Eight of twelve games are settled numbers you can build a read around today. Two — BAL@IND and ATL@PIT — are genuinely still moving, not just noisy, and deserve a recheck rather than a decision. Two more — CHI@CAR and WAS@PHI — have settled totals sitting on top of real offensive-line questions the total itself can't see. And weather for a third of the slate remains, correctly, unknown. Ten days out, that's not a complaint about the data. It's what Week 1 actually looks like before the calendar hands it anything better.