And there is Feynman's point that it is less embarrassing to prove yourself wrong, especially if you can do it before you publish.