At one point in my career, I was caught in the middle of a contentious change to an app's user interface. The app allowed users to communicate with each other about products via a chat feature. One side of the team wanted to add a disclaimer that said something to the effect of 'Please don't use Feature X to ask about blah blah blah. Use Feature Y instead." The other side of the team didn't want to add this disclaimer because 'people won't read the disclaimer and it clutters up the UI'.
I was on the side that didn't want to add the disclaimer. In my view, the reason that some users misused the chat feature (and therefore annoyed other users) is because the user interface did not make the correct choice obvious to users. Adding the disclaimer felt like we would be admonishing users for our design mistakes.
The 'add a disclaimer' side of the house was not convinced. The debate raged on for quite some time until I was able to break the stalemate with a hypothetical situation. I told my colleagues (paraphrasing here):
Imagine that you went to a restaurant for dinner. As you approach the door, you notice a sign that says "Please don't poop on the tables". Would you eat at that restaurant after reading that sign?
I suppose that the proprietor of that hypothetical establishment might have had a perfectly good reason to put up such a warning sign. Likewise, the folks that wanted to add a disclaimer had a perfectly good reason in mind. Unfortunately, the intended effect (avoiding unwanted behavior) is overshadowed by an undesirable effect (scolding users when they haven't done anything wrong).
Folks, we all know that users probably won't read the manual. The best way to handle undesirable behavior is to guide users towards the correct behavior with an excellent user interface. Avoid telling people to not poop on the tables.