DIPS and handedness

“And [Jesus] answered and said, He that dippeth his hand with me in the dish, the same shall betray me.” – Matthew 26:23 (KJV).
Nothing like starting off a Sabermetric blog post with a quote from the Bible, eh? You know you’ve been doing Sabermetrics too long when even your preacher’s sermons give you ideas for a study. (My pastor actually reads StatSpeak too.)
By this point, I presume that everyone in the room is familiar with DIPS theory. It says that pitchers have very little to no control over what happens to the ball once it is hit as to whether it will become an out or not. A little while ago, I noted that certain types of batted balls (i.e. grounders, line drives, etc.) were more in the pitcher’s control than others. I decided to take a look to see whether handedness made any difference. The pitcher can throw with either his left or his right arm, and the batter can stand in one of two rectangles drawn for his convenience. This creates four possible combinations of pitcher and hitter handedness.
I took my trusty Retrosheet 2000-2006 database and selected out all the balls in play, then coded whether the pitcher (whose hand preference was obviously fixed) was facing a right-handed batter or a left-handed batter. (Switch hitters were coded as whatever they were batting in that particular plate appearance, usually opposite to the pitcher.) To qualify for these analyses, the pitcher had to have at least 50 balls in play against the batter handedness in question (so if Larry had 52 balls in play against righties and 48 against lefties, only his data against righties was retained.)
I calculated BABIP (yes, that’s the stat for me! If you get that joke, you win a cookie!)for each pitcher-year separately for plate appearances against lefties and righties. Those of you who know my statistical style well know what’s coming next. I took the log of the odds ratio for the BABIP, then the AR1 intraclass correlation coefficient.
Right-handed pitcher vs.
Right-handed batter: .181
Left-handed batter: .105
Left-handed pitcher vs.
Right-handed batter: .190
Left-handed batter: -.025, which in this case, basically means nil.
Pitchers have more control over what happens when a right-handed batter is in the box (“more” being a relative matter; those r-squared values are still south of 4%), and lefty pitchers are at the mercy of the league average when a left-handed hitter makes contact against them.
My theory on this? Well, it’s the same reason with two different interpretations. Pitchers see fewer left-handed batters than right-handed batters. There are simply fewer lefties in the world and in baseball than there are righties. This could have two effects. The more statistical explanation is that pitchers see fewer lefties, which means a smaller sample size per season relative to righties. Smaller sample sizes are more volatile, which is hard on a year-to-year correlational method or an ICC.
The other possibility is that there really is some small thing that the pitcher can do to spin the ball in a certain way or throw it in a certain way that actually makes a batted ball more likely to go for an out. Pitchers, obviously, would want to develop this skill as much as they could. DIPS tells us that the effect of this skill is minimal, but any advantage is welcome. Well, the spin of the ball affects right handers and left handers differently. A ball that tails away from a righty breaks in on a lefty, so pitchers must plan accordingly. And they have more chances to practice this skill on right handed hitters.
Or it’s just statistical noise.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: