“Never make a calculation unless you already know the answer”
I don’t recall when I heard this first, but I’m sure I heard is several times as a physics undergrad at Carnegie Mellon. John Wheeler http://en.wikipedia.org/wiki/John_Archibald_Wheeler was a notable physicist at Princeton (and Richard Feynman’s thesis advisor.) Needless to say, he was held in very high esteem. I still have the copy of his “Space-Time Physics” that the Carnegie Mellon Physics Department Faculty gave me as a graduation gift.
Essentially, Wheeler advises that if you don’t have an expected answer, you cannot be surprised by an interesting result. “It is what it is”, is a useless tautology because it can’t be wrong! If you can’t be wrong, being right isn’t very interesting or useful. You don’t have to be precise, but you should have at least a rough estimation.
Many great discoveries have made when someone noted something to the effect “hmm, that’s peculiar”. Think, for example, of Wilhelm Roentgen’s discovery of X-Rays. An unexpected result is an opportunity to learn something. If, on the other hand, you get the answer you expected, your understanding receives a strong confirmation.
Whenever you try to improve a process, you should have some idea of what the results should be. Otherwise, how will you know if it worked or by how much? If you can’t make an estimate, you don’t know enough to make an intelligent change to the process.