Roster Advice
Turns a posted roster into real add / drop / hold decisions (pairs with league sync).
Intelligence engines
How we read it
Process grade: the composite
The within-position percentile of each predictive anchor, averaged. QBs are graded on EPA/play + DAKOTA + PACR; receivers on WOPR + target share + EPA; backs on volume + EPA. Anchors persist and forecast; production is the noisy output. The gap is the edge.
Buy-low: inputs say more is coming
When the process grade sits well above the actual production percentile, the opportunity says the points are coming. Buy-low before it corrects.
Sell-high: running hot
When production outruns the process grade, the player is converting above his usage. Sell-high before it regresses. In-line means the points are earned by the opportunity. Not a pick.
The process grade behind every player. is a Pro-tier analytics engine. This is the same data served by the premium /api/intelligence/player-model endpoint, gated for Pro and Elite members.
founding rate · locked for life · 3-day refund window
Data via nflverse (nflverse-data), licensed CC BY 4.0. How we source data
Turns a posted roster into real add / drop / hold decisions (pairs with league sync).
Combines the player grade, expected points, team environment, and passing signal into one ranked pool that feeds every fantasy tool.
Two public QB ratings side by side. When they disagree, we show it instead of averaging it away.
Yards over expected per carry, next to volume and stacked-box context: bell-cow, buy-low, or volume-dependent.
Air yards and target share combined into one opportunity score, with a clear buy / sell read.