Draft Capital (DPOS): The Strongest Single Signal
Data Dictionary · what the NFL's own behavior tells you about a prospect
DPOS is the draft-capital component of DMX: a prospect's pick number transformed into a z-score within his position cohort, with undrafted players assigned pick 300. It compresses the NFL's entire scouting apparatus — thirty-two front offices spending real money — into one number. And it is the strongest single-variable predictor of career outcomes at every position.
The numbers
Standalone, draft capital explains roughly 5.6–13.4% of career-outcome variance depending on position (published on the Model Transparency page) — more than any other single input, athletic testing and college production included. In hit-rate terms, splitting the 2001–2020 labeled cohort into draft-capital quartiles within each position:
| Draft capital | N | Star or Starter % | Bust % |
|---|---|---|---|
| Top quartile | 465 | 49.0% | 20.6% |
| 2nd quartile | 463 | 32.2% | 41.0% |
| 3rd quartile | 463 | 11.2% | 68.9% |
| Bottom quartile | 462 | 12.8% | 73.2% |
Offensive prospects, 2001–2020 classes, DPOS quartiles within position. Live query of the public database.
The top quartile of draft capital hits at nearly four times the rate of the bottom half. This is why “fade the combine, trust the draft” is half-right: the NFL is, collectively, a very good scout.
Why within-position z-scores
Raw pick numbers don't compare across positions, because the league prices positions differently — quarterbacks get drafted ahead of their talent on scarcity, running backs get pushed down on replaceability. Normalizing within position cohort asks the right question: how much capital did this player command relative to others at his job? A pick-40 RB and a pick-40 QB carry very different signals, and DPOS encodes that.
Why not only draft capital
If DPOS were sufficient, DMX would be redundant — but the full composite improves on capital alone at every position (the side-by-side R² comparison is on the Model Transparency page). The reason is that draft position is a market price, and markets misprice: teams reach for need, overpay for tools, and let productive-but-unspectacular profiles slide. Athletic testing and age-adjusted production are the independent witnesses that catch those mispricings — in both directions. The valuable signal isn't capital or the model; it's the disagreement between them.
Where capital decays
Draft capital is a draft-day prior, and its protective halo fades as real performance accumulates — a first-round pick gets two or three years of benefit-of-the-doubt, not ten. That decay is modeled explicitly on the in-career side: DPX applies a draft-capital decay term in its dynasty modifier, so a former first-rounder's pedigree counts for less with each season his production fails to back it.
Frequently asked questions
Is draft capital all that matters for rookies?
It's the strongest single signal, but not sufficient: the full DMX composite outpredicts draft capital alone at every position. The market misprices prospects - reaches, tool-based overdrafts, productive sliders - and testing plus age-adjusted production catch those errors.
How big is the hit-rate gap by draft slot?
Large. In the 2001-2020 labeled cohort, top-quartile draft capital within a position hit (Star or Starter) at 49.0%, versus about 11-13% for the bottom two quartiles, where nearly three in four prospects busted.
How are undrafted free agents handled?
UDFAs are assigned pick 300 before the within-position z-score transform, keeping them in the dataset at the extreme low end of the capital scale rather than excluded.
Why is DPOS computed within position?
Because the league prices positions differently - QBs are drafted ahead of talent on scarcity, RBs pushed down on replaceability. A within-position z-score isolates how much capital a player commanded relative to his own job market.