Quantifying Plate Discipline: Balls out of the Zone

Many baseball fans can collectively agree that Joey Votto is one of, if not the most disciplined hitters at the plate. However, what metric really shows that he's a disciplined hitter? Is it his absurdly low chase rate? His high number of "Good Decisions"? Or maybe it's his high walk rate? The fact of the matter is there are way too many statistics that go into evaluating a players plate discipline. If a player has a low chase rate coupled with a low walk rate, what is valued higher? What makes one player have a higher plate discipline than another? These are all questions that can't be answered under the current conditions, which leads to a more subjective (eye-test) view of plate discipline.

Over this 3 part series of blogs, I plan to evaluate the different aspects of plate discipline and narrow it down to one singular statistic to better aid teams and fans alike in evaluating this complex metric of baseball. In this first part, I want to focus on balls that are out of the strike zone. On part two, I will look at balls that are in the strike zone, and in part three, I will combine everything together to create one singular Plate Discipline statistic.

Accumulation of Data

In order to get an accurate sample of data, I required hitters to have seen at least 2,000 balls in their career from 2008-2020 (baseball savant doesn't have any data before then). This ultimately left the sample to 384 players.

Q-Score

The first part of the statistic is something called a Q-Score (Quality of Take Score). I broke down the percentage of takes that a player had that were in the "Shadow Region", the "Chase Region", and the "Waste Region". I then calculated a weighted average of all of the percentages and developed the Q-Score. I weighted the different regions as shown in the image above. Shadow pitches were only given 10% of the weight because many of these pitches are borderline strikes that depend on the umpire, whether they are called balls or strikes. Chase pitches were given 80% of the weight because those most accurately reflect a hitters ability to take a good pitch. The waste region was also given 10% because they have very little correlation with a hitters ability to distinguish a ball or strike.

Chase % and P/PA

The other two factors of the statistic are chase % and P/PA. Chase percentage is a good indicator of a hitters ability to distinguish balls and strikes and I felt that it had to be included. I also included pitches per plate appearance as players who see more pitches in an at bat typically have better plate discipline (on pitches out of the zone).

Creating PDBOZ (Plate Discipline on Balls Out of the Zone)

With these 3 baseline metrics, we can now develop a singular statistic. I combined the Q-Score, Chase %, and P/PA to get a single number. From there, I took the raw numbers and converted them into a percentile to give the statistic a meaningful and understandable range.

Results

In order to find the true results of this statistic, I ran a regression analysis and that compared our new PDBOZ statistic with true walk %. The results of this statistic came back with a much higher correlation than any of the other statistics that were utilized in this report. Thus, this statistic can be labeled as a success for narrowing down plate discipline on balls out of the zone to make the evaluation process much easier. With a 0.61 R Square value, we don't get perfect information, which is impossible to expect when evaluating stats because there will always be some element of randomness to the game. However, this R Square (and a p-value of 1.91x10-80) gives us enough confidence that this statistic is significant is measuring plate discipline.

Now for the fun part; How does this statistic rank current and former MLB players? Well, let's have a look.

On the left hand side, the top 20 hitters of PDBOZ are listed and on the right side the bottom 20 hitters of PDBOZ are listed. This information backs up our common knowledge of the game that hitters such as Joey Votto (4) and Juan Soto (13) are elite at deciphering pitches that are out of the zone. This also confirms that hitters such as Javier Baez (379) and Pablo Sandoval (381) are some of the worst hitters in the league at identifying bad pitches to lay off of.

Limitations

To wrap this article up, I want to quickly go over some limitations of this research and how those could be improved. First of all, this statistic has a hard time incorporating the "star factor". The biggest example of this is Bryce Harper. Bryce Harper's PDBOZ is an average 61.00 which ranks him at 150 out of 384 players. However, his career true walk % sits at 17%, which is one of the highest in the league. This shows that this statistic doesn't quite compensate for players that get pitched around, or don't have anything near the strike zone to hit. Another limitation to this statistic is that it can only be used for MLB players or for college programs that have access to a stadium Trackman. It would be very difficult to have any confidence with this statistic if Synergy, for example, was the main way that pitch location data was being gathered.

Linked below is the excel file that I used for all research in this blog. Feel free to look around the document for your favorite player or anything else you're interested in.

Plate Discipline on Balls Out of Zone (PDBOZ).xlsx

Download XLSX • 125KB