## Probabilistic Pitch Framing (part 3)

This is part three of a three-part series detailing a method of judging pitch framing based on the prior probability of the pitch being called a strike.  In part 1, we motivated the method.  In part 2, we formalized it. Here in part 3, we look at the hitter’s effect on ball and strike calls.

The formula we’ve been using for judging catcher framing is the very simple

```IsCalledStrike - prob(CalledStrike)
```

where IsCalledStrike is simply 1 if the pitch is called a strike, and 0 otherwise.  The second term is the probability that the pitch would have been called a strike, absent any information about any given party’s involvement. We add up these values for every called ball or strike that a catcher receives, and report the resulting number.  In this article we could go ahead and do this for all catchers over the past two years, except (a) Matthew Carruth is already doing this exact thing and (b) I can’t figure out how to match Retrosheet data to my Pitch F/X data to get catcher information anyway.  So instead we’ll look at hitter involvement.  How much can a hitter influence whether a pitch is called a ball or strike?

You probably already have a guess, and your guess is probably “not much.” Let’s see how right you are.  Using the above methodology, let’s see how many extra strikes were called for all hitters in 2012 and 2013.  Remember, what we’re plotting here is the number of extra strikes called on a hitter in 2012 and 2013, keeping in mind that each individual pitch may only contribute a tenth of a strike or something to this total.

Ignore that big blob in the middle; this is a counting statistic so hitters with few plate appearances will have their number hover around 0.  Instead, look at how empty it is in the northwest and southeast corners of the plot.  It is pretty rare for a hitter to get way too many strikes called one year and way too few the next.  And some hitters seem especially good or especially bad at this.

Let’s put some numbers on these dots.  Who were the hitters that turned at least 20 strikes into balls in both 2012 and 2013?

Batter Extra strikes 2012 Extra strikes 2013
Paul Konerko -30.7 -24.9
Shin-Soo Choo -20.2 -27.7
Elvis Andrus -24.7 -20.0
Elvis Andrus -67.0 -23.8

Some pretty good hitters here, as you might expect; it is a pretty significant advantage to be able to turn strikes to balls in this way. The one name here that was a surprise to me is Adam Dunn. The enduring image in my mind of Adam Dunn is him taking a called third strike and glumly walking back to the dugout. Not only that, but he is 6’6″ — you might expect some high strikes to be called for him that aren’t for other batters. It turns out, though, that he somehow has a knack for turning strikes into balls.

All right, how about the “unlucky” hitters, the ones who had at least 20 balls turned into strikes in each of the past two years?

Batter Extra strikes 2012 Extra strikes 2013
Zack Cozart 31.4 32.0
Mike Aviles 36.3 28.7
Paul Goldschmidt 35.4 22.8
Will Middlebrooks 24.3 27.7

Yep, a couple guys with OBPs hovering around .300 and … wait, Paul Goldschmidt? The guy who walked 99 times last year? That is just nuts. If he can figure out how to be just average in this category next year, watch out.

It looks like some hitters have a little bit of influence over called balls and strikes.  I would want more years in the sample to definitively call it a repeatable skill, but we have at least some evidence that points in that direction here.  Any ideas for looking at this moving forward?

Print This Post

The Kudzu Kid is a mathematician and avid baseball fan.

Guest
RKanderson
2 years 4 months ago

The suggestions I would have moving forward is to include pitch location and type into the analysis, but I’m not sure exactly how to do it. Take Dunn for example. Maybe he has an ability to sell the pitch being inside and almost all the extra ball calls he gets are on the inner half of the plate. Or maybe Dunn sees a lot more breaking balls than other players and those are the pitches umpires vary more, in terms of location, in calling consistently?

Guest
The Kudzu Kid
2 years 4 months ago

I like it. I’ll run some numbers for Dunn and if there’s anything interesting I’ll post it.

Guest
The Kudzu Kid
2 years 4 months ago

So Elvis Andrus doesn’t actually appear twice in the first table. The second Elvis should be Carlos Santana.

Guest
The Kudzu Kid
2 years 4 months ago

Well, I officially don’t see anything interesting for Adam Dunn in the numbers. Maybe we’ll have to go to the video for that one.