How Draftanomics works

Draftanomics ingests public big boards from named analysts and outlets and aggregates them into the Public Draft Board: where the sources place each prospect, how much they disagree, and how placements move week to week. Every source counts once with equal weight, and a player needs at least three sources to receive a consensus rank. Aggregation follows the Consensus Big Board methodology popularized by Arif Hasan. Every ingested source is credited on /sources.

Two surfaces that sit on top of the aggregation are being rebuilt and are not published as finished products yet:

• Prospect dossiers
  Being rebuilt — generated prose embeds withdrawn model figures.
• Analyst Accountability Index
  Being rebuilt — grades pooled accuracy across years including the year being graded.

The re-grade archive takes a completed draft class and scores it against what the players actually became. The ruler is Career Score, our open-data career-value index (snap participation, durability, second-contract market value, and first-six-seasons honors). It is computed only from realized careers, so a re-grade is a record of what happened, not a forecast.

A class is scored on the seasons its players have completed. Classes fewer than four seasons old are marked provisional because their careers are still early.

The comparison engine matches a prospect's pre-draft profile — measurables, production shape, age, program context — against historical prospects with similar pre-draft profiles. The match is computed on pre-draft information only: how the comparable player's career turned out is never an input to the match.

The comparables' careers are then shown alongside the match, because similar profiles have had a range of outcomes, and the range is the useful answer.

Before the draft, boards are sealed into immutable weekly snapshots with a content hash. After the draft, nothing in a sealed snapshot can be edited: what a board said in February is verifiable by anyone. The Public Draft Board serves from the latest sealed snapshot, and week-over-week movement is the difference between the two most recent ones.

• No predictive player-value outputs
  Draftanomics does not publish tier probabilities, bust probabilities, contract projections, or projected picks.
• No medical or character flags
  Injury history and off-field information are not in the data. A serious medical can override anything on this site.
• No live tape grading
  Draftanomics doesn't ingest film. Film observations in reports are cited to named evaluators — see /how-reports-are-written.

Every product above stands on named upstream sources, credited here and in full on /sources:

• College Football Data (CFBD)
  College production, recruiting composites, and program context.
• nflverse
  Draft picks, rosters, snap participation, and the career records behind Career Score.
• Public analyst big boards
  15+ named analysts and outlets (Kiper, Jeremiah, Brugler, and others), aggregated with credit per board.
• Relative Athletic Score (RAS)
  Athletic-testing composites by Kent Lee Platte.
• SP+
  Team-strength ratings by Bill Connelly, used for program and conference context.
• Pro Football Focus (PFF)
  Season-by-season college charting grades, used under license.
• Consensus Big Board methodology
  The aggregation approach popularized by Arif Hasan.

The written scouting reports follow their own sourcing rules — film language must trace to a named evaluator, numbers must trace to the data. How these reports are written →

What is the Public Draft Board?

An equal-weight aggregation of public big boards and published analyst rankings for the current draft class. Every source counts once; a player needs at least three sources to receive a consensus rank. It shows where public sources place each prospect, not a Draftanomics ranking.

What is Career Score?

Career Score is Draftanomics's index of what actually happened in a player's career: snap participation, durability, second-contract market value, and first-six-seasons honors. It describes completed careers and is computed only from realized outcomes.

Does Draftanomics predict which prospects will succeed?

No. Draftanomics publishes no predictive player-value claims. The site shows where public sources rank the class, what the collected scouting says, how past classes turned out, and which historical prospects had similar pre-draft profiles.

Where does the data come from?

The /sources page names every upstream dataset and its lineage status. College Football Data, nflverse draft and roster records, public analyst big boards, and licensed charting data are the principal inputs.