> Me again:
> But that is, in its very essence, a statistical question, and one that can
> only be answered statistically. What is Piazza's chance of throwing out a
> runner of Jeter's speed? What is the number of outs? How good are the
> other players in the Yankee bullpen? How good are the pitchers in the Mets
> bullpen? All of these are pieces of data which you can get, and then you
> can evaluate whether it's a good idea or not. Managers do that, and a good
> manager (a Davey Johnson or Joe Torre) is one who has an instinctual feel
> for the game so good that he can easily duplicate the statistical
> calculations in his head. But it's still a statistical question. How
> valuable is the stolen base? What are the odds that Jeter would score
> anyways? If it was Posada instead of Jeter, would you send him? Things
> like that all go into the equation. What about the Babe Ruth rule - don't
> make the last out of the game on the basepaths. Is that correct? I don't
> think it is - I think Babe was right to try and steal that base. Again,
> it's a statistical question that can be answered.
>
Ah we are coming together I think. The specifics could be answered statistically with
enough data but there is never enough data. Because in a single situation there are
too many subtle variables (Jeter had a big date last night and he is a bit uh
sluggish). So the manager has to have a core of statistical data to provide for his
overall stategy and a feel for how to apply that data in a specific situation. It is
that feel that is in part what makes a champion (not just for a year but for many
years).
