Every mock draft is a series of choices. A fan takes one prospect while other players are still available. Fanspeak brings those decisions together to build the Consensus Big Board.
The ranking reflects how people draft in On The Clock. It can help you see which prospects the draft room values, build a shortlist and compare your own board with the crowd.
Where the rankings come from
The inputs are human drafting decisions in On The Clock. When a person makes a pick, that choice supplies head-to-head evidence against players who were still available within a 64-slot comparison window.
Each choice contributes to the larger picture. The model combines that evidence across picks to estimate a relative order for the prospects. The board is a measure of drafting preferences; it does not promise the order NFL teams will select or how successful a player will become.
How the ranking is built
Consensus uses a Bradley–Terry model, fitted over accumulated choices. That model estimates how likely one player is to be chosen ahead of another, then uses those estimates to rank the board.
The published implementation is called bt-elo-nodecay. It uses a minorization-maximization fit, and older comparisons retain their weight. One new pick joins the existing evidence; it does not automatically cause a visible jump in the rankings.
Consensus, Crowd and Balanced
Fanspeak's three boards answer related questions with different methods.
- Consensus fits a Bradley–Terry model over accumulated human drafting choices and publishes a deeper prospect list.
- Crowd uses sequential Elo updates as new head-to-head choices arrive and publishes the top 150.
- Balanced blends analyst rankings with drafting data. That blend is separate from the Consensus method.
Consensus and Crowd both draw on drafting decisions. Their different rankings reflect different models; comparing them is not a comparison between fans and an independent group of experts.
What the board can tell you
A rank is a position in the published list. Filtering by position narrows that list while preserving the original ranks. A player outside the published coverage has no recorded rank in that snapshot.
When comparing dates, use the same board, draft class and method. A change in coverage can bring a player into the list without proving that the player jumped from an assumed last-place rank.
The number of ranked players, the number of times a board has been used and the number of qualifying drafts behind a ranking are different measures. The publisher has not supplied an edition-specific completed draft count, exact input window, inclusion audit or method revision for the available saved rankings. A sitewide total cannot stand in for those missing details.
Downloading and citing the board
On the Consensus page, open Download & cite to download the spreadsheet or copy a citation. Both are available without an account.
The spreadsheet includes its source, draft class, update timestamp and citation. The copied credit refers to that same ranking. If it matches a saved record, the citation links to that fixed record. Otherwise it identifies the live board and its update time, so it does not pass a changing page off as a permanent snapshot.
When quoting a rank, credit Fanspeak and retain the date and source link. Explain any filtering or analysis you add. Player photographs, school marks and third-party material are separate from the ranking data; the site's Terms and Conditions continue to apply.
