Add new comment

The important part to remember is that the calculation compares Seattle hitters (home) vs. Seattle Hitters (away).
In that respect, it is (mostly) irrelevant whether your hitters are good or bad. It is how much better or worse they hit at home versus on the road.
Say Seattle hitters OPS .750 at home and .800 on the road. The calculation is looking at the SKEW (that 50 point swing), not whether they are good or bad.
Same goes with pitchers. If Seattle pitchers accumulate a 3.50 team ERA at home and go 4.00 on the road, the skew is what matters.
So, the visiting teams are composited, meaning they are 50% of the rating ... but of course every opposing hitter is facing a Seattle pitcher for the calculation (and vice versa).
The one mitigating factor in all this is if a team manages to construct itself to take particularly good advantage of its park (or constructed horribly to be disadvantaged), then talent *could* skew the park effect. But that is part of the reason they calculate both a pitcher and hitter park effect as two numbers. That can potentially reveal any outlier results where ONLY your offense or defense is seriously impacted by the park.
While I don't know the entire calculation, the basic gist of it looks at the home/road skews. The "official" park adjustments are calculated using 3 years of data, as it is felt single season values are not as reliable (prone to random fluctuations). That's what I call the weather effect -- if one year the wind happens to be blowing in for 60 games at Wrigley and out for the other 21 ... and then the next year it is 40/40 ... that's going to impact the park effect. The multi-year method attempts to factor out the random fluctuations to reveal the "static" impact. (Of course, when you move the fences around that kind of throws the whole 3 year calculation out the window. In that regard, it will be 2015 before the Safeco park effects will become "vested".