Using Zone% to Find Breakout Minor Leaguers

Last May, Rob Arthur of Baseball of Prospectus unveiled new research implying that PITCHf/x data can be used to forecast hitter performance. Essentially, he found that hitters who saw fewer pitches down the heart of the plate late in the year often outperformed their projections the following season. The theory behind this phenomenon is pretty straightforward. A pitcher knows a good hitter when he sees one, and chooses to approach him with caution. So when pitcher’s change their approach, it’s often an early indicator of improved talent from a hitter.

This finding gives us yet another tool to use in forecasting hitter performance. As great as statistical projections are, they can sometimes be a bit slow to recognize a breakout when it happens. Simply put, it’s not easy to identify the one breakout from the sea of small sample size flukes. However, Arthur’s research hints that incorporating pitch locations can help us get ahead of the curve.

This past January, Eno Sarris wrote a piece for JABO where he applied Arthur’s findings to 2014 data in an attempt to identify breakout candidates for 2015. Eno noted that this type of analysis is most useful for players who are relatively new to the league. For this reason, he limited his analysis to hitters with fewer than 800 big plate appearances.

I decided to dig a little deeper. To my knowledge, all applications of Arthur’s research have looked exclusively at big league hitters. However, in what follows, I make my best attempt at expanding this practice to hitters in the minor leagues.

Admittedly, this is an imperfect exercise. Arthur’s analysis was based on distances from the heart of the plate. Simply put, replicating these calculations just isn’t possible without PITCHf/x data, which isn’t publicly available for minor league games.

But while we don’t have minor league PITCHf/x data for minor leaguers, we do have something similar. Using data scraped from minor league play-by-play data, Minor League Central publishes pitch-specific data on minor league hitters. These are the same metrics featured on the PITCHf/x Plate Discipline page here at FanGraphs: O-Swing%, Z-Contact%, Zone%, etc.

The metric I’m focusing on today is Zone% — the ratio of pitches thrown to a hitter that fall within the strike zone. If we assume Arthur’s theory applies to hitters in the minor leaguers, we’d expect to be able to find some of 2015’s breakout players by searching for hitters whose Zone% tailed off towards the end of last year.

Now, MLC’s Zone% data isn’t gospel. They’re recorded by pitch stringers, who manually record every single pitch as being either inside or outside of the strike zone. Each is subject to biases, and they surely mis-classify a pitch or three every now and then. This data is far from perfect.

But imperfect data and all, let’s see what we can find. In this analysis, I considered all hitters who saw at least 1,000 pitches from April-July, and at least 300 more from August 1st on. Then, I simply looked for guys whose Zone% were significantly lower in the season’s final month and change. This data is only available for the Triple-A levels (the International and Pacific Coast Leagues) and two of the three Double-A levels (the Southern and Texas Leagues, but not the Eastern League).

Some of the players who rank highly on these lists aren’t prospects at all, but look more like minor league lifers. Nonetheless, I decided to list the entire top 5, just in case. If Todd Cunningham or Brady Shoemaker somehow winds up being big league relevant this year, I don’t want to miss out the opportunity to say, “I told you so!” Anyway, on to the Zone% decliners.

International League

Rank	Name	April-July Zone%	August-September Zone%	Change
1	Todd Cunningham	62.3%	53.7%	-8.6%
2	Matt Tuiasosopo	53.8%	46.5%	-7.3%
3	Joey Terdoslavich	56.7%	51.0%	-5.7%
4	Quintin Berry	57.1%	51.6%	-5.5%
5	Andy Wilkins	61.7%	56.3%	-5.4%

None of the players in this quintet are particularly exciting, as each is at least 26-years-old. Even so, each of them is on the cusp of the major leagues, and at least in theory, is only a small improvement away from being worthy of a call up. The most interesting player here is probably Andy Wilkins, who is a first baseman in the Blue Jays system.  Wilkins clubbed 30 homers in Triple-A last year, and has mashed at just about every minor league stop in his minor league career.

Pacific Coast League

Rank	Name	April-July Zone%	August-September Zone%	Change
1	Nick Buss	64.7%	59.9%	-4.8%
2	Domingo Santana	57.0%	52.4%	-4.6%
3	Tony Campana	62.9%	58.9%	-3.9%
4	Andy Marte	56.7%	53.6%	-3.1%
5	Tyler Greene	61.5%	58.7%	-2.7%

Here, the name that stands out most is Domingo Santana, who’s the only real prospect of the bunch. Santana had a rough go of it in a big league cameo last year, but the power is definitely there. Otherwise, Nick Buss is likely the most interesting name on this list, and there’s very little that’s interesting about Nick Buss.

Southern League

Rank	Name	April-July Zone%	August-September Zone%	Change
1	Richard Shaffer	61.8%	53.2%	-8.6%
2	Brady Shoemaker	54.5%	47.5%	-7.0%
3	Gerson Montilla	58.2%	52.3%	-5.9%
4	Nick Ramirez	56.5%	50.9%	-5.6%
5	Scott Schebler	59.5%	53.9%	-5.6%

Rays prospect Richard Shaffer might be the most intriguing player mentioned in this article. Shaffer was a first round pick back in 2012, but hasn’t hit particularly well in his two plus years in the minors. Shaffer’s Zone% dropped nearly 9% over the season’s final month, and his drop Zone% isn’t the only reason to think he may have turned a corner.

From Kiley McDaniel’s writeup:

[Shaffer is] another stalled former 1st rounder with plus bat speed and raw power has been so bad I wasn’t going to mention him at all until Rays personnel raved about his offensive adjustments in instructs; I’ll give him one more season.

From a conversation David Laurila recently had with Shaffer:

I’ve made adjustments and am more direct to the ball with my hands, and more under control with my lower half. I’m mechanically sound and more physical this year.

Scott Schebler doesn’t have Shaffer’s pedigree, but he’s also worth keeping an eye on. As a 23-year-old in Double-A last year, he crushed opposing pitching to the tune of .280/.36/.556. This showing earned him a 40 FV rating from Kiley in his write-up of the Dodgers organization.

Texas League

Rank	Name	April-July Zone%	August-September Zone%	Change
1	Brian Hernandez	57.9%	50.2%	-7.7%
2	Ethan Chapman	58.8%	51.9%	-6.8%
3	Rafael Ortega	58.1%	53.0%	-5.1%
4	Jorge Bonifacio	56.7%	51.9%	-4.8%
5	Delino DeShields	54.7%	50.3%	-4.4%

Jorge Bonifacio and Delino DeShields are the names to remember from this list. Bonafacio is a highly-regarded outfield prospect in the Royals system, who revieved a FV of 45 from Kiley this winter. Although he had a down year in 2014, there’s still plenty to like about the toolsy 21-year-old.

DeShields is also coming off of a mediocre 2014 campaign in Double-A. Nonetheless, the Texas Rangers plucked him from the Houston Astros in last winter’s rule 5 draft. DeShields’s carrying tool is his speed — He stole 256 bases in roughly twice as many minor league games. Up to this point, his bat has left a lot to be desired, but perhaps the drop in Zone% could be an early indicator of an improvement in that department.

Just outside the top five, we find a couple of more rule 5 picks who managed to crack their clubs’ opening day rosters — Taylor Featherston of the Angels and Odubel Herrera of the Phillies. This duo placed 8th and 10th respectively with decreases between 3% and 4%.

Most of these players won’t perform noticeably better than expected in the coming year, and it’s entirely possible that none of them will. Zone% is far from a perfect mechanism for predicting a player’s future performance. Even so, the data suggests Zone% can be of some value in forecasting future performance, and if it can identify one prospect who’s on the precipice of a breakout, it might be worth exploring.