What would be really interesting in light of what you say is repeat performance. Also, it is not inconceivable that even random success may _cause_ repeated success as successful people tend to get more opportunities and are more confident.
Well, then you get fooled by randomness in a completely different way: You begin to believe that the ability to succeed from scratch is a much rarer skill than it actually is, because (a) only a subset of the potentially-successful people ever succeed, by whatever measure, and (b) only a smaller subset of those ever bother to try to succeed again; and (c) only an even smaller subset of subset (b) ever succeed a second time.
But you don't really need to find a serial performer. You only need to find someone who can succeed once, and there are more of the latter than there are of the former, by definition. It's no knock on Bill Gates that, having founded one incredibly successful company, he didn't quit and try to start a different one just to "prove" that he had the knack. Zuckerberg should not be ashamed that, having found himself holding one tiger by the tail, he doesn't feel inclined to shop around for a different tiger.