2026 NFL model performance study
A transparent, continuously updated study of completed EdgeLine NFL official selections. It reports every graded win, loss and push using the original locked number. Pending member forecasts are excluded.
12-18
30 completed official selections
-23.6%
-7.09 units at standardized grading
Results by market and model tier
These splits help readers evaluate where the model has performed well or poorly without selecting only favorable outcomes.
| Segment | Record | Win rate | Units | ROI | Avg. CLV |
|---|---|---|---|---|---|
| All official selections | 12-18 | 40.0% | -7.09 | -23.6% | +0.10 |
| Point spreads | 8-10 | 44.4% | -2.73 | -15.2% | +0.17 |
| Game totals | 3-8 | 27.3% | -5.27 | -47.9% | 0.00 |
| Other / legacy market labels | 1-0 | 100.0% | +0.91 | 90.9% | — |
| Strong tier | 2-6 | 25.0% | -4.18 | -52.3% | +0.06 |
| Play tier | 10-12 | 45.5% | -2.91 | -13.2% | +0.12 |
How this study is calculated
- Only completed official Play and Strong recommendations are included in the primary record.
- Each result is graded against the line locked before kickoff, not a better later price.
- Units use the standardized grading stored with each selection.
- Closing-line value compares the locked line with the recorded closing line when available.
- Non-official leans are tracked separately and are not blended into the official record.
What the numbers do—and do not—show
Win rate and ROI describe realized historical outcomes. Closing-line value evaluates whether the locked price beat the closing market. Neither metric guarantees future performance. Small samples, injuries, data revisions and random game outcomes can materially affect results.
For full row-level verification, open the completed NFL results. For model inputs and weighting, read the NFL model methodology. For public metric definitions, visit the Research Center.