The converse of this is analysis paralysis. More than once I've agonized over what could go wrong in production, and trying to prevent these scenarios, only to

The converse of this is analysis paralysis. More than once I've agonized over what could go wrong in production, and trying to prevent these scenarios, only to realize in hindsight that even the entire system going down for a day would've not been noticed by or problematic for the particular client.