Methodology
How the Draft Kit works
Two inputs, one subtraction. We take what the market charges for a player in 2026 drafts and set it against what he actually produced in 2025 — and the rows worth your attention are the ones where those two disagree. Everything below is arithmetic you could redo yourself.
3,899
real drafts behind our ADP
247
players ranked
6,095
2025 games logged
4
scoring formats tracked
Where ADP comes from
ADP — average draft position — is the typical pick a player goes at. Ours is read from real 12-team drafts, not a panel of rankers: 3,899 of them behind the top of the board, refreshed daily. It's displayed as round.pick, so 4.07 is round 4, pick 7.
We snapshot all 4 scoring formats — PPR, half-PPR, standard and 2QB/superflex — every day, which is how the risers and fallers board knows what moved. Daily snapshots started Tue Jul 21, so the movement window is still short and we say so rather than extrapolating.
Where production rank comes from
Every 2025 regular-season game is logged — 6,095 of them across 615 players. A player's production rank is simply where his PPR points per game placed him at his own position last season. No projections, no adjustments, no opinion.
The value number is the gap between those two ranks. Positive means the market drafts him later than last season's production justifies; negative means he's going earlier than production alone supports and you're paying for a story — a bigger role, a new offence, a leap. Sometimes the story is right. The number just tells you that's what you're buying.
What the 2026 Outlook is — and isn't
It is not a projection model, and we won't call it one. It's last season's per-game scoring pulled toward the positional average in proportion to how many games you actually saw — a full season barely moves, eight games moves a lot — with a floor and ceiling from that player's own week-to-week swing.
That's why an eight-game player with a big average comes back noticeably lower: the sample, not our opinion, is doing the work. Players who changed teams keep the same midpoint and get a wider range, because a new situation is uncertainty, not a prediction.
76 of 684 players get no Outlook at all — rookies and anyone who missed 2025. We could emit a number for them. We don't, because there's nothing behind it: their draft price is projection and draft capital, and that's what we say.
Who writes The Fantasy Desk
The Desk is written by a language model from a fact sheet built out of these same tables — not by a person, and we're not going to pretend otherwise with a byline.
What makes that trustworthy rather than a shortcut: the model is given the numbers and forbidden from adding any, and the exact fact sheet used is stored alongside every article. If a piece cites a number, that number came from a query, and it's on record. It can't invent an injury or a depth chart, because it's never told one.
What we don't have
No 2026 projections in the industry sense — no touchdown forecasts, no target-share modelling, no strength-of-schedule adjustment applied to player estimates. No player age, contract, or draft-capital data, which is why rookies get honesty instead of a ranking.
We track 3,165 news items to explain *why* a price is moving, but we don't grade our own fantasy calls — there's no fantasy accuracy record to show you yet, and we'd rather say that than publish one that isn't real.
Runs on the same data