- Mocks included
- 24
- Analysts
- 24
- Date range
- Apr 27–Sep 1, 2026
- Calculation
- live-2.3-source-independence
What counts as an eligible mock
A projection must target the correct draft year, contain a recognizable pick sequence, carry a verified writer and publication date, and come from an approved or manually reviewed source. Aggregator and roundup pages are retained only as private discovery provenance: Ultimate Big Board follows them to the original publication, canonicalizes that URL, and attributes the public record to the originating writer. Syndicated copies, simulations, withdrawn drafts, and low-confidence parses are excluded. Updated versions remain in history, but only the newest version of one analyst’s opinion votes by default. Big boards contribute to player ranking signals only and never to team-pick consensus.
Partial mocks are evidence only where they speak
A top-10 mock contributes to picks one through ten. A one-round mock contributes through pick 32. Absence after the author’s final published pick is never treated as evidence that a player went undrafted.
Restrained, configurable weighting
Each eligible selection receives a weight based on recency, independence, completeness, and parsing confidence. Repeated mocks from one analyst or publication are capped so a high publishing frequency cannot overwhelm the broader signal.
Consensus teams, unique players
Every team selection is tracked as a distinct projected pick—the club’s first, second, or later selection—and its weighted average position is calculated across sources. A maximum-weight assignment combines that average with exact pick support to build one coherent order without duplicating a single team pick. A club can still appear more than once when analysts genuinely project it to own multiple selections through trades. Players are assigned separately so each prospect also appears only once. Support, alternatives, sources, and the completed order are frozen with the snapshot.
Confidence is a label, not a verdict
Team-order confidence shows both the weighted support rate and how many eligible mocks cover the slot. Player confidence combines pick share, team connection rate, projection spread, independent source count, recency, source diversity, and order stability. The interface translates these signals into plain-language labels and displays the underlying counts.
Original summaries, cited themes
Each player page includes a source-attributed overview sized to the evidence available. It compares pick placement, team assignment, range, publication timing, and recurring themes from stored selection-level reasoning. Missing commentary is labeled instead of reconstructed. No source language is reproduced at article length, no summary is presented as broad consensus unless at least three independent publications contribute, and every eligible mock links to the original source.