DFS Projection Accuracy Report — Week 1 (2026)
Through Week 1 of the 2026 NFL season, here's how our four projection sources (CBS, FantasyPros, Sleeper, ESPN) performed against actual DraftKings scores — and how those findings are shaping next week's lineup weights.
Key Takeaways
- Cbs is the most accurate source this season (RMSE 6.85 pts) and receives the highest blend weight (1.0).
- Espn is the least accurate (RMSE 7.29 pts) and gets down-weighted to 0.94.
- Cbs over-projects Pass Yds by 22.4 on average — treat Cbs pass yds numbers with caution heading into Week 2.
- Cbs over-projects Pass TD by 0.4 on average — treat Cbs pass td numbers with caution heading into Week 2.
- Espn over-projects Pass TD by 0.6 on average — treat Espn pass td numbers with caution heading into Week 2.
- Fantasypros over-projects Pass TD by 0.5 on average — treat Fantasypros pass td numbers with caution heading into Week 2.
- Sleeper under-projects Pass Yds by 15.2 on average — treat Sleeper pass yds numbers with caution heading into Week 2.
- Sleeper over-projects Pass TD by 0.4 on average — treat Sleeper pass td numbers with caution heading into Week 2.
Overall Source Accuracy
| Source | RMSE | MAE | N players | Blend Weight |
|---|---|---|---|---|
| Cbs | 6.85 | 4.89 | 249 | 1.0 |
| Sleeper | 7.01 | 5.13 | 298 | 0.978 |
| Fantasypros | 7.13 | 5.22 | 258 | 0.96 |
| Espn | 7.29 | 5.28 | 249 | 0.94 |
RMSE = root mean squared error; MAE = mean absolute error; lower is better.
Accuracy by Position
| Position | Cbs RMSE | Sleeper RMSE | Fantasypros RMSE | Espn RMSE |
|---|---|---|---|---|
| QB | 7.99 (n=34) | 9.16 (n=32) | 9.35 (n=34) | 9.27 (n=35) |
| RB | 7.76 (n=68) | 7.73 (n=71) | 7.38 (n=65) | 7.62 (n=59) |
| WR | 6.88 (n=85) | 7.05 (n=106) | 7.22 (n=97) | 7.42 (n=96) |
| TE | 4.76 (n=62) | 5.26 (n=62) | 5.02 (n=62) | 5.06 (n=59) |
| DST | — | 5.16 (n=27) | — | — |
Stat Category Bias (Mean Error)
Mean error shows the direction each source is wrong. Positive = source over-projected; negative = source under-projected. A consistent directional bias is a lineup signal — e.g. if CBS reliably over-projects passing yards, CBS-heavy QB projections are inflated and should be treated with skepticism.
| Stat | Cbs | Sleeper | Fantasypros | Espn |
|---|---|---|---|---|
| Pass Att | — | -2.6 | — | — |
| Pass Cmp | — | -1.5 | — | — |
| Pass TD | +0.4 | +0.4 | +0.5 | +0.6 |
| Pass Yds | +22.4 | -15.2 | -0.8 | -3.5 |
| REC | +0.3 | +0.4 | +0.4 | +0.3 |
| Rec Yds | +2.4 | +4.0 | +4.0 | +4.0 |
| Rush Att | — | +0.7 | — | — |
| Rush Yds | +3.0 | +3.2 | +2.0 | +1.1 |
| Tgt | — | +0.3 | — | — |
Stat Category Magnitude (MAE)
MAE shows how far off projections were on average, regardless of direction.
| Stat | Cbs | Sleeper | Fantasypros | Espn |
|---|---|---|---|---|
| Pass Att | — | 6.5 | — | — |
| Pass Cmp | — | 4.5 | — | — |
| Pass TD | 0.7 | 0.8 | 0.9 | 0.9 |
| Pass Yds | 64.5 | 68.8 | 73.4 | 71.4 |
| REC | 1.3 | 1.3 | 1.3 | 1.2 |
| Rec Yds | 17.7 | 18.8 | 18.3 | 18.7 |
| Rush Att | — | 3.2 | — | — |
| Rush Yds | 15.1 | 17.7 | 15.8 | 15.9 |
| Tgt | — | 1.6 | — | — |
Data through Week 1 of the 2026 NFL season. Accuracy computed across 2 weeks.
