I've known PhDs who wrote very high level design docs full of hand-waved features. When they implemented anything, half of the features were segfaulting and the

I've known PhDs who wrote very high level design docs full of hand-waved features. When they implemented anything, half of the features were segfaulting and the code base was a mess. None of this changed their self confidence in the slightest. Curiously, people pushing for formal methods, unit tests and sane design practices were all non-PhDs.