How It Works

EdgeIQ Methodology

Five independent sport engines. Each picks from the same transparent architecture — Elo ratings, a three-component vote, confidence capped between 50% and 78%, and line movement tracked at DraftKings, FanDuel, and BetMGM. No black boxes. No secret sauce. Every doc below shows exactly what each engine reads, how it votes, and where the model draws limits.

Every pick from every engine starts from the same architecture: per-team Elo ratings, a three-component vote (each sub-model weighted equally), confidence clamped between 50% and 78%, line movement captured at three books across a six-hour pre-tipoff window, and no pick published as a lock. Sport-specific inputs and calibration adjust the model per league — NFL leans EPA regression, NHL leans xG regression, NBA layers schedule density, and so on.