DataDrivenResults

Methodology

Updated 2026-11-14T15:09:27Z

Props EV

Each row compares the estate's projection model probability (model_p) with the best available sportsbook line devigged to an implied probability (implied_p) across the books scanned that day. Edge = (model_p − implied_p) × 100. Confidence grades (A/B/C) reflect sample depth. Raw inputs: the estate props board, refreshed daily + intraday.

Retail price spreads

For every product the restock rig watches, we take all live listings it currently sees: spread = (highest − lowest) / lowest. This is a cross-listing flip proxy, not a guarantee — fees, shipping, and condition apply (see the calculator's fee input).

Crypto funding spreads

Perpetual funding rates are pulled from public, key-free endpoints (OKX, Gate.io, Hyperliquid). Reported spread = (max − min venue funding per 8h) × 3 × 365, annualized. Execution costs, borrow, and venue risk are NOT modeled — treat as a screening figure.

Standing caveats

Screening figures, not advice. Data surfaces carry as-of stamps; stale cells are dropped rather than guessed. Nothing on this site is a recommendation to transact.