NFL intelligence, built around the market.

1. Step 1Learn the ideasShort guides: the Intel Score, NO READ, prospective tracking, model vs market.
2. Step 2Read the researchLonger pieces on market timing, probability, projections and how models are tested.
3. Step 3See the productA frozen pregame example of the Desk and a Player Props demonstration.
4. Step 4MembershipWhat membership includes and what it costs.

• 01See a sampleA frozen pregame example of the member Desk, plus a fictional Player Props demonstration.
• 02Learn how FDI worksShort guides to the ideas the Desk is built on.
• 03Intel ScoreWhat the number measures — and what it is not.
• 04NO READWhy a game can have no read, and what that means.
• 05Prospective trackingFrozen before kickoff, checked after the game.
• 06Model vs marketProjections, lines, prices and quote timing.
• 07Membership and pricingWhat membership includes and what it costs.

• September 30, 2026·Model vs market

  A Projection Is an Average, Not a Line

  An FDI player projection is an expected value; a sportsbook line is a number chosen to balance two prices. Why the two are different objects, why the difference between them is arithmetic rather than a signal, and how to read them together.

• September 30, 2026·Methodology

  Why a Model Probability Is Not a Recommendation

  A model can estimate the chance a player finishes above a specific number. Here is what that estimate means, why FDI shows it only when the current line matches a modeled threshold exactly, and why it is never turned into an instruction.

• September 30, 2026·Methodology

  Why Market Freshness Matters

  A sportsbook line is a snapshot, not a fixed fact. Why every FDI market figure carries the time it was observed, what changes when a quote goes stale, and why FDI will not compare a projection with a line that is no longer current.

View all insights

FDI organizes what the market says and what governed models say about the same game, side by side, and is explicit about where evidence is thin. Members see:

• 01

  Market context

  Full-game total and moneyline consensus across the books FDI tracks, how the line has moved, how many books are pricing it, and how close the total sits to a key number.

• 02

  Governed model context

  Where an eligible read exists: a projected game total, a moneyline win probability, and a Moneyline Lean when its policy qualifies. A Lean is directional model context, not an instruction. Many games have none of these.

• 03

  Intel Score

  Where an eligible one exists, a 0–100 measure of how strongly the available evidence supports the way a read was built.

• 04

  Player Prop Intelligence

  Player projections for receptions, each frozen two hours before kickoff and shown beside the same player’s current line, with a dossier per player.

• 05

  Honest gaps

  When a view is not available it stays empty. Nothing is estimated just to fill a card, and every market price shows when it was observed.

FDI does not manufacture a read for every game. When the evidence is not there, NO READ says so plainly: no qualified FDI model read is being presented for that game. Insufficient evidence stays insufficient — it is not dressed up as a number, and NO READ is not an instruction to act or to stand aside.

Read more about NO READ

Where FDI tracks a model prospectively, the order of events is fixed:

1. 1.Pregame information is gathered — only what existed before the game.
2. 2.The model output is produced and recorded before kickoff.
3. 3.That record is frozen: its time, inputs and value do not change.
4. 4.The game is played.
5. 5.Postgame evidence arrives: the result, and who actually played.
6. 6.The result is joined back to the original record, which is never rewritten.

A later result never edits the earlier prediction. The record shows what the model said when it said it.

How prospective tracking works

FDI doesn't sell picks, promises no financial return, and makes no claim that its model outputs beat the market or the closing line.