Our futures model was frozen on 2026-05-16 and has not been re-run. The market has traded every day since, absorbing camp reports, injuries and trades the model never saw.
So when the two disagree, the honest starting assumption is not that the model found something the market missed. It is the reverse: the market usually knows something the model cannot.
This note takes the eight largest gaps on our 15 August board and asks, case by case, what the market learned that the model could not. We could answer for three of eight. For the other five we could not, and we are saying so rather than filling the space.
This piece has a short life. Every figure here is a snapshot dated 15 August 2026 measured against a model frozen in May. Once games are played these comparisons describe history rather than the board.
The clearest case is Green Bay
Micah Parsons joined the Packers in a 2025 trade, so the May model already knew he was on the roster. What it does not know is that he is recovering from a torn ACL and expected to miss at least the opening weeks.
The model holds a Packers team with a healthy Parsons. The market prices one without him for a month. That is the 12.3-point gap in a single fact, and it is the model being out of date rather than the market being wrong.
A gap between a frozen model and a live market is a question to investigate, not an answer.
The other two are narrower. The Rams' division number sits higher in the market than in the model, and part of that traces to a trade the freeze never saw: Los Angeles acquired Myles Garrett on 1 June, two weeks after. Cutting the other way, Matthew Stafford has been managing a back problem and missing August practice. Baltimore's coaching change, with Harbaugh out and Jesse Minter hired, predates the freeze and is already reflected structurally. What a static model cannot capture is three months of evidence about how a new staff actually works.
Where we came up empty
| Team | Market | Model | Market implied | Gap |
|---|---|---|---|---|
| Philadelphia Eagles | Division | 58.0% | 40.6% | +17.4pp |
| Seattle Seahawks | Playoffs | 79.7% | 65.1% | +14.5pp |
| Green Bay Packers | Playoffs | 67.1% | 54.8% | +12.3pp |
| Detroit Lions | Playoffs | 76.2% | 64.7% | +11.5pp |
| Baltimore Ravens | Playoffs | 84.6% | 73.3% | +11.4pp |
| LA Rams | Division | 40.2% | 47.3% | -7.1pp |
| Minnesota Vikings | Division | 9.1% | 15.7% | -6.6pp |
| Dallas Cowboys | Division | 23.7% | 30.2% | -6.4pp |
Philadelphia's is the largest gap on the board and the one we could least explain. We found some camp injuries in the secondary, and news that Washington left tackle Laremy Tunsil suffered a torn triceps expected to cost him significant time. Neither comes near explaining 17.4 points.
Minnesota runs the other way: the market is more bullish than the model. Kyler Murray signed in March, before the freeze, so that is already priced. Safety Jamal Adams suffered a season-ending preseason knee injury, which should push the number toward the model rather than away from it. We could not identify why the market is where it is. Seattle and Detroit have no specific post-freeze development behind their playoff gaps that we could find.
- Explained by a specific event. Green Bay, the Rams, and Baltimore in part.
- Not explained. Philadelphia, Seattle, Detroit, Minnesota, Dallas. We looked for a post-freeze story on each and found nothing that accounts for the size.
- The ratio is the finding. Three of eight. That says more than any single team's number.
None of this is a case for taking a position on either side of these gaps. It is a record of where a fixed reference and a live market have drifted apart, and an honest account of how much of that drift we can trace. What a frozen model is good for covers how to read one; the full board sits in the Forecaster.
Snapshot dated 15 August 2026. Division figures are drawn from the six books quoting that market, playoff figures from four. Once games are played these comparisons describe history rather than the board. Sources for each development are linked inline.