The first version of the list will be alright. It becomes good over a couple of years, one forgotten item at a time.

What the research says

A multilevel meta-analysis found that people with substance use disorders differ from non-using controls in how they value delayed rewards and in how they process feedback, with the effect varying by substance. The feedback part is the relevant bit: learning from what went wrong is a process in its own right, and one that can be propped up from outside. It was a clinical comparison, so it says nothing about lists directly.

Strength of evidence: one line of research. A single study, or a theoretical account.