of course, my methodology, which is infinitely more precise than miket or death's, still is seeing wild swings in the value of PC every time i analyze a new pitcher.
you really do need a huge sample for this to work. i'll save this one in google docs and supplement it on a rainy day