2026 Sleepers, Values and Fades
Where our machine learning projections disagree with the draft market, by position. Model rank comes from our 2026 board, market rank from live ADP across real drafts, and every list below recomputes each morning as both move.
Draft-day values
Players going at least 4 positional spots later than the model ranks them, inside the first eight or so rounds. If the model is right, these picks return more than they cost.
| Player | Team | Model rank | Drafted at | ADP | Gap | Proj pts |
|---|---|---|---|---|---|---|
| Wan'Dale Robinson | TEN | WR26 | WR41 | 86.4 | +15 spots | 170.5 |
| Michael Wilson | ARI | WR27 | WR37 | 75.6 | +10 spots | 170.3 |
| Quentin Johnston | LAC | WR38 | WR46 | 97.6 | +8 spots | 155.7 |
| Jadarian Price | SEA | RB22 | RB29 | 73.9 | +7 spots | 169.2 |
| Jeremiyah Love | ARI | RB7 | RB13 | 27.1 | +6 spots | 240.4 |
| Brian Thomas | JAX | WR30 | WR35 | 73.0 | +5 spots | 169.1 |
| George Pickens | DAL | WR7 | WR11 | 20.8 | +4 spots | 234.2 |
| Tetairoa McMillan | CAR | WR13 | WR17 | 30.9 | +4 spots | 211.5 |
Sleepers (ADP 100 and later)
The same signal in the double-digit rounds: late picks and waiver-priced players the model ranks well inside startable range at their position.
| Player | Team | Model rank | Drafted at | ADP | Gap | Proj pts |
|---|---|---|---|---|---|---|
| Emmett Johnson | KAN | RB32 | RB57 | 157.7 | +25 spots | 123.0 |
| Jordyn Tyson | NOR | WR57 | WR75 | 157.4 | +18 spots | 115.8 |
| Jauan Jennings | MIN | WR50 | WR65 | 147.8 | +15 spots | 126.2 |
| Tre Tucker | LVR | WR53 | WR58 | 132.2 | +5 spots | 120.0 |
| Justin Herbert | LAC | QB10 | QB14 | 105.0 | +4 spots | 271.7 |
| Romeo Doubs | NWE | WR47 | WR51 | 113.0 | +4 spots | 136.8 |
| Keenan Allen | LAC | WR55 | WR59 | 132.9 | +4 spots | 117.2 |
| Malachi Fields | NYG | WR69 | WR73 | 155.3 | +4 spots | 97.0 |
Fades: where the market is higher than the model
Not "bad players", just prices the model would not pay. The market drafts these players at least 4 positional spots ahead of where our projections rank them, so someone else can pay the sticker price.
| Player | Team | Model rank | Drafted at | ADP | Gap | Proj pts |
|---|---|---|---|---|---|---|
| Mike Evans | SFO | WR40 | WR27 | 55.2 | -13 spots | 155.2 |
| Chris Godwin | TAM | WR48 | WR38 | 79.6 | -10 spots | 134.1 |
| Terry McLaurin | WAS | WR33 | WR24 | 47.7 | -9 spots | 166.4 |
| Luther Burden | CHI | WR36 | WR28 | 55.6 | -8 spots | 162.6 |
| Jayden Reed | GNB | WR52 | WR44 | 95.0 | -8 spots | 121.5 |
| Rashee Rice | KAN | WR14 | WR8 | 15.3 | -6 spots | 208.3 |
| Garrett Wilson | NYJ | WR18 | WR13 | 26.2 | -5 spots | 199.0 |
| Jakobi Meyers | JAX | WR45 | WR40 | 86.2 | -5 spots | 140.1 |
| Dak Prescott | DAL | QB9 | QB5 | 65.8 | -4 spots | 274.8 |
| Bhayshul Tuten | JAX | RB27 | RB23 | 52.4 | -4 spots | 143.8 |
How this list is built
Player projections come from position-specific models (ElasticNet and RidgeCV) trained on 16 seasons (2010-2025) of NFL data. We rank every player at his position by projected 2026 fantasy points, then line that order up against market ADP order at the same position. Disagreements of 4 or more spots land here.
One honest caveat: when the market is far ahead of the model on a player, the market sometimes knows something the stats do not yet show, like an injury timeline, a holdout, or a coaching promise. We show the gap and the numbers behind it, and you make the call.
Sleepers and values FAQ
What is a fantasy football sleeper?
A sleeper is a player being drafted much later than his likely production deserves, usually in the double-digit rounds. On this page a sleeper is any player with an ADP of 100 or later whose model rank at his position is at least 4 spots better than where drafters are taking him.
Who are the biggest 2026 fantasy football values right now?
The biggest gaps between our model and the draft market right now: Wan'Dale Robinson (model WR26, drafted WR41); Michael Wilson (model WR27, drafted WR37); Quentin Johnston (model WR38, drafted WR46). The full list on this page refreshes every morning.
How does DraftAI find sleepers and values?
We compare each player's rank in our machine learning projections against his market rank at the same position (ADP order among 2026 drafts). A gap of 4 or more positional spots in the model's favor makes him a value or, past ADP 100, a sleeper. A gap in the market's favor makes him a fade. Both lists recompute daily from fresh projections and ADP.