Why a result looks odd (and isn't)

A few things come up again and again. Spread width counts strikes on the chain's grid, not dollars, so "1 to 5 wide" on a $2.50 grid means up to $12.50. Rows with a missing quote pass filters on purpose, because unloaded data shouldn't look like a structure that doesn't exist.

Too-good-to-be-true rows are removed at source: quotes that violate no-arbitrage bounds are rejected, and near-free "edges" smaller than commissions are dropped. And the win-rate figure is a modelled probability, meaningful for comparing structures but not a forecast.