1. Fix a mistake in one mammoth calculation, another mistake elsewhere is inevitable (mathematical whack-a-mole law).
2. If you base your results on the work of others, a flaw in one of those works will be the worst possible for your work.
3. The longer your paper, the more likely you are to forget where you started.
4. Clear results are more quickly rejected. (Corollary: journals have a bias for borderline results.)
5. If a result seems too good to be true, it is. (Exception: if you are very famous, it doesn’t matter.)
6. Your most startling new theorem will turn out to be valid only in the trivial case.
7. If you hold a seminar on your new work, no one will understand it except an opinionated asshole who happens to be your grant review officer.
8. If you used fudge factors for years and no one caught them; they will be like chum for sharks at your tenure committee meetings.
9. If you chose a hot topic as a graduate student, it will be dead by the time you get your Ph.D.
10. If you discover an interesting model, Feynman will already have lectured about its possibility.