Thanks for the explanation. I was a bit confused about that.

]]>Park factors are typically expressed in terms of runs per out, and can be used for anything that is in those terms (like ERA, RC27 and so on). Another way of putting it is, a ten percent increase in the measure means a ten percent increase in runs – the linear relationship has a 1:1 slope.

That is not true of measures like wOBA, OPS, etc. They all tend to have a relationship closer to 2:1. So you need to adjust the park factor accordingly. Taking the square root is an approximation of the “correct” adjustment.

]]>