Earlier this week, I discussed my mission to figure a way to quantify defense in a meaningful manner, and I came out with something I dubbed pitching independent defense, which basically relates a team’s relationship to their expected BABIP against, which is calculated as a function of line drive percentage.
But does it mean anything? My immediate instinct was to see if the pitching rotation was tangibly improved by a team having a better PID number. I regressed the difference between ERA and FIP versus the PID metric that I came up with.
There is one immediate, recognizable concern. FIP is a statistic that tries to suss out a pitcher’s true ability by eliminating luck factors, such as the defensive skill. BABIP and the difference between FIP and ERA are going to be closely correlated, and they are.
That said, BABIP SHOULD be close across the league, but there is variability across the league. The first thing I wanted to do was make sure that raw line drive percentage wasn’t a factor in anything, and as it turns out, the relationship between LD% and DiffERA/FIP was very low.
Naturally, BABIP should generally correlate with the PID if BABIP is supposed to be similar across the league. PID accounts for variance away from BABIP… well, an expected BABIP. There was about a 64% correlation between BABIP and PID. This is a good start.
Now, the important test. Does a pitching staff ERA respond to a good PID? Essentially, does my estimation of a good defense bear out in a fruitful benefit to the pitching staff? In short, yes. There was an r^2 of about .53, which is pretty good, nearly as good as the relationship to BABIP.
The good news is, we now have an effective tool for evaluating team defense. We know that a good team defense will result in better results for the pitching staff. This is all using data assembled from last year. Tomorrow evening, we will see what teams have assembled the best team defenses THIS year.
Add The Sports Daily to your Google News Feed!
