Okay, so Jesse has been on a "NL = cozy park" theme of late. Where cozy parks = many HRs.
This one I had an opinion on - not a stong one, but an opinion and my opinion was the AL had the cozy parks (camden yards, texas, comiskey, Fenway - even though I KNOW the numbers say otherwise) and the ptichers parks are in the NL (LA, Pac Bell, SD, FLA).
So again. Lets go to the numbers. Park affect statistics for 2004.
Based on runs scored.
White Sox 1.139
Chicago (NL) 1.123
New York (NL) 0.974
St. Louis 0.936
New York (AL) 0.922
Clearly the NL has the better pitchers parks. But this does not equal cozy parks in terms of HRs - after all, the aforementioned Fenway is a hitters park (with 12 % more runs scored there) but it has 2.4% less HR than the average park.
So, pure rankings for HR.
White Sox 1.402
Chicago (NL) 1.329
NY (AL) 1.029
NY (NL) 0.804
Two things jumped out at me.
1. Colorado was not number one?
2. Jesse was right!!
Colorado is a good place for HR as we all know, but their hitters hit them both at home (111) and on the road (91) and their pitchers give them up in both places (110/88). Whereas the White Sox hit 145 HR at home (and pitchers gave up 127) (interesting in that there were far more gross HRs hit in CWS than Coors) and hit only 97 (and allowed 97) on the Road. Led by Konerko with a 29/12 H/R split and Frank Thomas 14/4.
And while the sum in each league does equal 1.00 thus effectively negating the effects of all park affects in the aggregate (yes it sounds stupid but is true), there is a large enough gap in the standard deviation to make a difference for individual hitters (i.e., Bonds and Beltre while playing mid to tough parks get to play almost 25% of their games in AZ and Colorado), let alone the slugger (say Thome)who plays half his games in Philly and 20 more in ATL.
One last thing I did was divide Park affect runs by park affect HRs, which would yield parks with a disproportionate share of their runs being scored via the HR. That analysis gave no real trend by league and I had trouble drawing a correlation about what it meant.
Good call Jesse.