Steve Waugh, the Physics student, gives a Statistical Mechanics exam and while checking on his corrected answer-sheet founds that he has received zero marks for a certain problem (related to the entropy of a gas in grand-canonical ensemble). He walks to the instructor and inquires what’s wrong. Instructor goes through Steve’s solution, ponders for a few seconds and then replies,

“Mate, you have just dropped a Gibbs factor!”

(a person with bagpack outside Physics dept.)

If someone had time and energy, it would be interesting to collect large sample of mistakes from, say internet, and try to reproduce the keyboard modulo some group, say, reflection about the middle row. If one has the additional knowledge of the user being a right/left hander, one could keep this group size smaller.