Quote:
|
Originally Posted by IatricSB
On the other hand, the more I read and here it "explained", the less I believe in it. To me it seems more useful when looking at pitchers as a whole rather than the individual.
|
Yes, Iantric, that is precisely true. But, read the quote from MGL's article that Skipaway posted above (and which, iirc, an OotP poster responded). I will reproduce what is for me the important part:
Quote:
|
Originally Posted by MGL
As I explained before, no matter how small the sample correlation is for x number of BIP, as long as there is any true correlation at all, given a large enough sample size, the correlation will eventually be 1 (assuming that whatever talent a pitcher does have for preventing hits is fairly constant over time). However, given a true correlation of .036 for 500 BIP samples, it would take more than a 13,000 BIP sample (4,200 innings or around 20 years of pitching) to have a correlation of .500. In other words, of you wanted to estimate a pitcher?s true $H from a year?s worth of his sample data (~500 BIP), you would regress his sample $H over 96% towards the league average; if you wanted to estimate a pitcher?s true $H from his 20-year history (~13,000 BIP), you would regress his sample $H only 50% toward the mean (the regression amount always equals 1-r).
|
So, as been pointed out before, if you agree DIPS works wholesale, then you agree that DIPS means you get 95%+ there individually. Afterall, what OotP really is doing is esimating -- or stating -- the pitcher's "true H$" -- his actual talent. What Rob and others want is to go that extra 5%. But, that sounds good to me, but just remember it is that extra 5%, not a wholesale revision.