Your 0DTE stop should be based on the SPY level where your trade idea is wrong — translated into a premium and a dollar amount — not an arbitrary percentage. Placing the stop at the right level is what makes it protect you instead of shaking you out.

Anchor the stop to your thesis

Every scalp has a point where it's invalid — a SPY level that, if broken, means your idea was wrong (below the support you're bouncing from, back inside the range you broke out of). That level is where your stop belongs, because it's where you actually want to be out. A stop placed there is meaningful; a stop placed at a random distance just gets hit by noise or lets losses run too far. Start from “where am I wrong?”, not “how much am I willing to lose on the option?”

Translate it to the option

Since your stop lives on the option's premium, translate the SPY level to a premium via delta, and size the position so hitting that stop loses your intended dollar amount. This ties the stop to both your thesis (the level) and your risk budget (the dollars) — far better than a crude “50% of the option” rule that varies risk randomly.

A good stop answers “where is my idea wrong?” — not “how much of the option can I stomach losing?” Anchor it to the level, size it to your dollars.

The quick takeaway

Put your 0DTE stop at the SPY level that invalidates the trade, translated to a premium, and size so that loss equals your risk budget. Anchor to the thesis, not a random percent. NoVo attaches a level-based, dollar-risked stop automatically on entry — and it should be a real (hard) stop, not one you plan to watch.