I. Extra credit for attending a marginally relevant lecture by a famous visitor? Sure! Except: those points will interact with the existing homework rules so baroquely that if you can figure out that you're in bad enough shape to benefit significantly from the freebies, then you actually deserve some academic remuneration after all.
II. We're back at that decision-making/risk/imperfect-information/(ir)rationality point again, in my "pssst! wanna fulfill some distribution requirements, punk?" course. After we discuss some medical examples, an overconfident student comes up. "Isn't it true for pregnancy tests that like, if it's negative, then that might or might not be true, but if it's positive, then you're in trouble?"
I involuntarily take a deep breath, stare into space for a moment, then answer, "Actually, I know way too much about that example." Even this student isn't so oafish as to ask another question as I furiously gather my notes.