Fundamentally, rules almost always come with compromises — for the sake of making rules understandable by humans, they have to be relatively simple. Simple rules for complex situations will always forbid some amount of good behaviour, and allow some bad behaviour. Many of society's parasites live in the space of "allowable bad behaviour", but there is a lot of value to knowing how to exploit the "forbidden good behaviour" space.

Here's the dangerous way I put it that I only tell senior people: understand why rules were made and make sure the people who made them would be happy.

In any large organization, there are basically two classes of rules: 1) stupid red tape rules that slow everyone down and 2) really important rules that you can never break ever. Effective people learn which rules fall into which group so they can break the red tape rules and get more stuff done.

As a supervisor I didn’t resonate with this until I remembered in some jobs I have communicated the company attendance policy but didn’t enforce it unless someone was a poor performer. I trust adults to manage their own time until they give me a reason to believe otherwise.

For my part, I’d rather trust people’s judgment and intrinsic motivation than enforce the rules. Enforcement is annoying, tedious, and distracting to my mission. However once I decide their judgement can’t be trusted I use rules to extrinsically motivate them.

A more palatable phrasing, "supervisors prefer people that engage with the rules with purpose." That is, choosing to break a rule because you are making a cost call based on what you were able to achieve is not, necessarily, a bad thing.

The "point" where this fails, of course, is where the "cost" call above is such that the supervisor can't agree.