I realize Ken is a real student of the quirks of PB stats, but I think on this one he is wrong.

Personally, I do NOT believe the game would treat the hypothetical John Dow any differently than it treats Helton if they had identical stats.

I think Ken and others incorrectly interpret the meaning of "park effects".

Of course, I realize I might be the one who is wrong. This is just my "hunch".

I strongly suspect the park effects occur ONLY based on the park where a PB game is played. If the game is in Coors, then ALL players will be more likely to get hits and hit HR. But, players who played their games at COORS will NOT have their stats treated any differently than players who played their games at LA.

My recommendation:
Don't draft Colorado pitchers.
Do draft Colorado hitters.

Without seeing the actual program logic, obviously this is just a guess based on how I've seen players perform in PB. I've had Todd Helton hit over .400 for me in a non-hitters park. I've had a pitcher who gave up 27 HR at COORS give up 37 in a pithcers park in PB. These are just a couple of instances and perhaps extreme examples.

The manual Ken quoted talks about adjusting the park index for a 3 year average. Now imagine doing that for each player when many players played for 3 or more different teams in different parks over that 3 year period. Even in a given year some players play for multiple teams.