Ranked decisions, exact limits.
The production board will publish only source-grounded, exact-line NFL decisions that clear the calibrated threshold.
Nine inputs. One decision.
Wager Tape evaluates the distribution, line, payout, and uncertainty before qualifying a decision.
Recent production adjusted for role and sample quality.
Expected snaps, routes, touches, and attempts.
Unit performance adjusted for scheme and personnel.
Expected assignment and alignment.
Play-calling and substitution tendencies.
Applied only when conditions materially affect the market.
Availability changes and redistributed opportunity.
Pace, spread, rest, travel, and likely script.
Current line, sharp baseline, and movement.
More data is not the edge. Knowing what matters is.
Each market uses a purpose-built model. A signal enters only after demonstrating predictive value on unseen games.
Sharp-market baseline
Every proprietary model must beat a price-derived baseline on unseen data.
Full outcome distributions
Tail outcomes, role changes, and altered scripts remain in the distribution.
Evidence-tested inputs
Weather, coaching, injuries, and matchups enter only after testing.
Line-specific decisions
A grade applies to one direction at one exact line.
Accountability before performance claims.
There is not yet a meaningful public sample. Every future result will be preserved.
A+ through B− publishes. C is research only.
A grade belongs to one direction at one exact line, not to a player. Numerical thresholds publish only after unseen-data calibration.