What the Two-Leg Campaign Actually Returned

Every cell is a run on real traded 1-minute SPY 0DTE option bars — no model, no implied vol, no assumed fills. Returns are expressed as a percentage of the campaign debit. Build a configuration and see what it did.

1 · When to arm

2 · Strike rule

3 · Harvest ladder

4 · Fill assumption & horizon

Why the fill assumption gets its own selector. With no historical quote data, the entry price has to be taken from the traded tape, and the choice between "you got the closing print of your arming minute" and "you lifted the worst print of that minute" is not cosmetic — on this structure it moved one cell from +4.1% to −25.4%, a 30-point swing from a single convention. Any result on a strategy this cheap that quotes one convention without the other has not told you the answer.