In Part 1, we looked at Barry Larkin's defensive efficiency on ground balls. In Part 2, we considered his double play efficiency. Today, it's balls in the air- pop-ups, bloopers and line-drives. And it will be much shorter.
The reason it will be shorter is simple- we do not have sufficient data. The retrosheet event file contains some information about balls in the air:
Unfortunately, important data is not listed in the events file for determining "ball in the air efficiency":
The consequence of the absent data is that while we can measure how many plays the shortstop made on balls in the air, we have no reliable information on the number of opportunities he had to make such plays.
Let's look at each type of ball in the air separately. Most true pop-ups caught by the shortstop are routine. A small number are not. Occasionally a shortstop will make a play on a popup in fair territory right down the third base line between the third baseman and the leftfielder.
A larger number of these plays are made in foul territory. We do not treat balls in the stands as opportunities for fielders (home runs to left are not counted as opportunities for the left-fielder), and the same logic applies to balls in the stands in foul territory. Unfortunately, we do not have the data to separate that balls that land in foul territory from ones that land in the stands.
Many of the most difficult plays made by shortstops are made on bloopers, and the retrosheet classification system (pop-ups vs. line-drives) does not allow for a consistent evaluation of a shortstop's opportunities. To give a concrete example, a line-drive to short left-centre is almost never caught by the shortstop and a pop-up there will be caught most often by the shortstop with little difficulty. If a blooper there is caught and classified as a pop-up, the meaningfulness of the data is lost.
For line drives, the question is also simple. Which balls are realistically "opportunities" for the shortstop? A ball directly at him and 12 feet high is not realistically an opportunity, whereas one 8 feet high is. If we count all line drives to left centre as opportunities, we capture both.
In short, the data set does not allow for conclusions with respect to Larkin's defensive ability on balls in the air.
The reason it will be shorter is simple- we do not have sufficient data. The retrosheet event file contains some information about balls in the air:
-pop-up vs. line-drive
-location of balls caught (foul and fair) and location of hits (fair only)
Unfortunately, important data is not listed in the events file for determining "ball in the air efficiency":
-no hang-time measurement,
so bloopers do not fit into either pop-up or line-drive class reliably
-no height measurement for line drives
-no recording of foul pop-ups that are not caught
(i.e. whether they are in the stands or hit the ground).
The consequence of the absent data is that while we can measure how many plays the shortstop made on balls in the air, we have no reliable information on the number of opportunities he had to make such plays.
Let's look at each type of ball in the air separately. Most true pop-ups caught by the shortstop are routine. A small number are not. Occasionally a shortstop will make a play on a popup in fair territory right down the third base line between the third baseman and the leftfielder.
A larger number of these plays are made in foul territory. We do not treat balls in the stands as opportunities for fielders (home runs to left are not counted as opportunities for the left-fielder), and the same logic applies to balls in the stands in foul territory. Unfortunately, we do not have the data to separate that balls that land in foul territory from ones that land in the stands.
Many of the most difficult plays made by shortstops are made on bloopers, and the retrosheet classification system (pop-ups vs. line-drives) does not allow for a consistent evaluation of a shortstop's opportunities. To give a concrete example, a line-drive to short left-centre is almost never caught by the shortstop and a pop-up there will be caught most often by the shortstop with little difficulty. If a blooper there is caught and classified as a pop-up, the meaningfulness of the data is lost.
For line drives, the question is also simple. Which balls are realistically "opportunities" for the shortstop? A ball directly at him and 12 feet high is not realistically an opportunity, whereas one 8 feet high is. If we count all line drives to left centre as opportunities, we capture both.
In short, the data set does not allow for conclusions with respect to Larkin's defensive ability on balls in the air.
