PLAYERS
2026 PLAYER PROJECTIONS
2026 PLAYER STAT LINES
Projected 2026 season stat lines and PPR fantasy points for every skill player, a pure model (multi-year usage × efficiency, age curves, and a games-played durability model), no betting or expert-ranking input. Floor–ceiling ranges shown.
60
RB STARTERS SHOWN
303
TOP RB PPR
| PLAYER | GP | RUSH YD | PPR |
|---|---|---|---|
 Jahmyr Gibbs DET | 16 | 1,084 | 303 |
 Bijan Robinson ATL | 16 | 1,213 | 297 |
ᴿ = rookie (draft-capital prior). PPR = standard PPR fantasy scoring. Ranges are the model’s floor–ceiling band (wider for rookies + thin histories).
NOTE: this board shows projected starters only (QB1, and RB1/RB2, WR1/WR2, TE1/TE2 for the committee positions), by each team’s real depth chart. Every rostered player still gets a projection; the rest just don’t clutter a starters’ board.
HOW THESE ARE BUILT
Each projection is team volume × player share × efficiency: a player’s multi-year per-game usage and per-opportunity efficiency, empirical-Bayes shrunk toward the position baseline, adjusted by a frozen age curve, and scaled by an availability model that projects games played from durability history, the single biggest accuracy lever. Pure model, zero betting or expert-ranking input. On backtests it beats a naive last-year projection at every position and agrees ~68% with expert rankings. See the methodology →
QUESTIONS
On a 2016–2024 walk-forward backtest the model beats both a position-average and a last-year-carry-forward baseline at every skill position on fantasy-PPR error, and correlates ~68% with the current expert rankings, while using zero market or ranking input.
Every stat line carries a floor–ceiling PPR band. Rookies and players with thin histories get visibly wider bands, because there is genuinely more uncertainty.
How steadily a player produced last season, not how much. It is the week-to-week spread of his 2025 PPR points (standard deviation over the average) ranked against every player at his position with at least eight games and a five-point average. 100 is the steadiest at the position, 0 the most boom-or-bust. STEADY is the top third, BOOM/BUST the bottom third. Rookies and thin samples get no score rather than a guess. Two players with the same projection can get there very differently; this is the column that says which.
Rookies have no NFL history, so they're seeded from their draft capital scaled to a position baseline, with the widest bands. Their projections firm up once they play.
Before kickoff
Get the model's picks every week, free
Every Wednesday the model picks every game and emails the slate before kickoff. Every pick is graded in public, wins and losses.
