> Me: > Yeah, but the number of flips isn't chosen arbitrarily. You're a scientist, > you must understand this better than I do - why not apply that knowledge to > baseball? We can look at all of baseball - which is an enormous n - and do > statistical tests on it. The conclusions - that clutch hitting doesn't > exist, and so on, are phenomenally robust. This is true to any reasonable > level of statistical significance. That's what SABRE, or Baseball > Prospectus, or Bill James's Historical Baseball Abstracts, were all about. > Maybe the problem is that the stats are measuring individual performances rather than team performance. I have read the clutch hitting stat and am not sure what to make of it. Kind of like the stats on hot streaks that reveals that no such thing exist; that hot streaks are the predictable results (not when but whether) of the coin toss phenomena and the overall incidence; in other words the overall BA will determine how many hitting streaks and their length for an individual batter. The only important streak that is an outlyer is DiMaggio's 56 game streak, predicting that this record will stand for a long time (either because for that period of time he was much better than anyone else before or since; that is he was "hot" or that he was very lucky; I would actually favor lucky if it weren't for the fact that after having his streak end because of some great fielding by the Indians, he went on a 19 game hiting streak).> > >
> But that's precisely what you're saying about Pedro. He's not as good as Koufax >because he doesn't win big games. My claim was weaker than that. I was saying that you can't tell whether Pedro is the best ever because comparing individual stats across time does not work in my opinion. Pedro is obviously a great pitcher, unique. But all individuals are products of their times and it is not possible to judge individuals in different times with the same tools. I think it easier to compare how the individual did relative to his/her contemporaries. Was he/she the best by far in that era? Apart from the fact that this is empirically not true - his performance over the last two years in big games was dazzling. So you're arguing that Pedro was an inferior pitcher to Koufax because Pedro's teammates couldn't hit as well as Koufax's_even though Pedro's statistical performance was superior to that of Koufax. Like I said it is my memory that Kofax was dominant. The series against the Yankees were all low run or no run games. His teammates only needed one or two runs. That is all they gave him. Maybe it isn't fair to say that Pedro's team did not win key games against the Yankees. But that is the reality. > > Oh yes, one other thing - I quite distinctly remember Koufax being quoted as > saying that Pedro was better than he was. Kofax is famously modest. But I would accept this as reliable if he meant it. > > Gautam
