In the somewhat recent past, I wrote an article about Madison Bumgarner and his proclivity for throwing more pitches up in the zone than previously in his career. An intrepid commenter suggested that an interesting follow-up would be to calculate a zone-based BABIP to see if the particular zone where a pitch was thrown affected BABIP. What follows is an investigation of the suggestion.
The question that the suggestion was trying to answer was if a pitcher's zone distribution percentage could help explain chronically low BABIPs. For example, if a pitcher is constantly living up in the zone, and those parts of the strike zone have lower BABIPs, that might help explain a chronically low BABIP.
I will go further by breaking out the balls in play by their type, i.e., fly ball, line drive, ground ball, and popup according to Baseball Savant. In fact, I'll just state that everything you will see following is from Baseball Savant and from the years 2008-2013. Unfortunately, the data are not broken out by handedness of the batter. For the future, I suppose.
First, here is a replication of Baseball Savant's zone map. Numbers 1-9 are the strike zone, and the map is from the catcher's perspective.
Here is the BABIP, among some other potentially useful information, of all balls in play in each zone. The relative production, or Rel PRD, is calculated by (1.7*BA+SLG)/(1.7*totBA+totSLG), which makes 100 average. Anything above 100 is above average production, and vice versa.
This is probably about what you were expecting. There are more fly balls in the upper parts of the zone, more ground balls in the lower part of the zone. There is some variation in the line drive rate by zone, but not too much. As for the BABIP, it looks to me that zones 1, 2, and 3 have lower BABIPs than the other 6 zones within the strike zone. Overall, it looks like zones 1,2, and 3 yield lower production than the other 6 zones. Naturally, everything is worse in zones 11-14 because those are not within the strike zone.
Here is another table, but for fly balls only.
A similar trend shows up. The lower BABIPs are in the upper zones. There is some variation in the home run rate, but I was surprised to see that the home run rate in the 3 upper zones is not much different than the other 6 zones within the strike zone. It looks like the overall production is lower in the upper 3 zones as well.
Here is the replica of the above table for line drives.
There is some variation, but overall line drives are fairly uniform throughout the zones. The relative production hovers around 100, and the home run rate stays relatively constant.
Now, ground balls.
Zone 9 is the best place for pitchers to get weaker ground balls within the zone; otherwise, there's not much difference between the zones.
In each case of batted ball, there is also a trend across the strike zone going from left to right. Zones 2, 5, and 8, the middle zones, generally sport the highest BABIPs. Zones 1, 4, and 7, inside to righties and outside to lefties, are in the middle. Zones 3, 6, and 9, outside to righties and inside to lefties, have lower BABIPs.
There are a few conclusions from this investigation. One is that pitchers seem to have little control over their line drive rates and the production allowed on those line drives. Minus a very large sample size, I would expect all pitchers to have about average line drive rates and about average production allowed on those line drives. Another is that zone 9, low and away to righties and low and inside to lefties, is a pretty good place to throw baseballs. Finally, throwing up in the zone is capable of generating weaker contact than other parts of the zone. Finally, it does seem like there is a relationship between the zone where the pitch is and the resulting BABIP, although follow up investigations that control for batter and pitcher handedness and selection effects are necessary.
. . .
All statistics courtesy of Baseball Savant.