That's not even a response to what randomdata said. S/he observed, correctly, that switching strategies until you succeed does not imply that you were not lucky. On the other hand, we can demonstrate extremely strong correlations between very specific applications of effort/skill and success at getting PHDs and marathons.