When your own incompetence dooms a labmate
Tools for success ≠ Success
When you don't even exceed your own low expectations
Adding too much functionality to the experiment at once
When the experiment breaks while demonstrating it
When you're one set of data away from graduating...