Heh, I wonder if the quizzes just shuffled the questions between students, how many would get it wrong because their option A) was other people's option B)
No. It wasn't like that. I built software that would generate unique completely random test questions. The software would eval() the students response and would give 100% if correct, 0 if not.