Someone correct me if I'm misremembering, but I think Donald Knuth wrote somewhere that when he'd be assigned the odd-numbered problems, he'd always do them, and do the even-numbered problems as well.