For those of you that didn't hear, on Saturday morning 2008 first round selection Bradley Holt was released:
Mets Farm Report (@MetsFarmReport) March 30, 2013
It marked the end of a meteoric rise and fall of a pitcher who, at one point in time, looked like a long-term piece of the puzzle for the Mets. To be specific, that point came as a member of the Brooklyn Cyclones back in 2008, where a 21-year-old Holt opened eyes all over the game by posting an outstanding 1.71 ERA in 14 starts, striking out an otherworldly 96 batters in just 72 innings pitched and looking the part of a future ace.
Alas, Brooklyn was to be the apex of his (Mets) career arc as he would go on to post a 5.67 ERA over his next four seasons within the Mets farm system. While he showed flashes, he never again found the same level of success that he featured in Brooklyn.
Aside from the subsequent hand-wringing and countless what ifs that result from such an unfortunate end, the Brad Holt experience (new band name, I call it) has also raised questions among minor league analysts about how to evaluate pitching in Brooklyn. That's because, as we recently observed with a number of the Mets minor league facilities, Brooklyn's MCU Park happens to be a pitcher-friendly environment. With a 0.94 park factor -- as measured in a 2011 study by Baseball Think Factory's Jeff Sackmann and Dan Szymborski -- Brooklyn is one of the tougher run-scoring environments in the entire New York-Penn League.
This question is made even more relevant based on the very timely fact that the Brooklyn Cyclones just happened to feature an absolutely outstanding pitching staff in 2012. As a unit Cyclones pitchers finished first in the New York-Penn League in innings pitched, ERA, and opponent average while ranking second in walks as well as strikeouts (though first in K/BB). The starting rotation, specifically, was flat-out dominant placing four of the NYPL's top fifteen pitchers in ERA, strikeouts, walks, and opponent average.
Guys like Hansel Robles, Luis Mateo, and Gabriel Ynoa have suddenly thrust themselves into the top prospect discussion. However, people are now questioning the predictive value of their collective performance based on context. With the ink of Holt's walking papers still wet, it seems a safer bet to perhaps take Brooklyn stats -- and perhaps all of those short-season pitching stats -- with a larger grain of salt and instead wait a year to determine who is for real.
To me, however, this mentality sort of misses the point of prospecting in the first place. Obviously our first instinct is always to say 'wait until next year to see if if so and so improves.' That's because until a player is a perennial all star at the major league level -- and even then -- he's a constant work in progress. But the point of evaluating these guys is to give our best judgment in the moment.
What's more, by writing off high-level performance in Brooklyn we'll be missing plenty of indicators for future success -- throwing out babies with bathwater. Which babies you ask? Below I've compiled every pitcher in Brooklyn's history (2001 to present) who compiled a sub-three ERA in 60 or more innings pitched (sortable by clicking header):
|Year||Age||Player||Innings Pitched||ERA||Strikeout %||Opponent Average|
Obviously there are a lot of names there -- many of them obscure -- which would tend to hurt my argument in that most of them did not become effective Double-A pitchers, let alone major leaguers. However, if we incorporate a couple more intuitive filters -- to try to gauge overall stuff -- this population begins to make more sense. First, let's say that we'll only look at guys with a strikeout percentage above 20%. Next, we'll trim down only to guys that allowed an opponent average of .235 or lower.
Four of the five regular starters from the 2012 Cyclones rotation -- Robles, Mateo, Ynoa, and Rainy Lara -- made the cut. There were six others who survived the added filters:
- Bobby Parnell
- Dylan Owen
- Brad Holt
- Chris Schwinden
- Brandon Moore
- Collin McHugh
Three of those guys have reached the majors -- which in and of itself is a victory at this level of minor league baseball. Dylan Owen is a well-known tweener who has the ability to consistently retire minor leaguers but not the stuff to contend in the majors. Brandon Moore is something of an exception in that he was one of only two 23-year-olds on the list -- something that already raises a red flag -- and was later suspended for drug use. There were even a handful more familiar names who fit the bill but fell short of the innings cutoff (Bannister: 22.8%; .196 | Petit: 45.1%; .138 | Kazmir: 50%; .193).
And then there's Brad Holt. Obviously Holt's name on this list casts some doubt on the entire pool -- but it shouldn't. At least not in the way that people are thinking about it lately. The fact that Brad Holt busted -- despite excellent primary indicators -- doesn't mean that Brooklyn stats as a whole are not instructive. It means that pitching prospects bust. It happens. Oftentimes for no good reason -- like Holt. Oftentimes for very good reasons. Sometimes after experiencing success in short-season ball. Sometimes after experiencing success in Double-A.
(If you really want to dissect it, Holt's case is an example of the stuff peaking extremely early -- perhaps an indicator that he just didn't have the kind of arm/physicality to hold up over the course of a full season. As we know, like command or control, health is a skillset, and some guys -- despite being gifted with the kind of stuff to rouse major league scouts -- are just not blessed with the kind of durability to do something as unnatural as hurl a horse hide a few thousand times every summer. As we've observed on a much higher level with Tim Lincecum over the past season or so, pitching can take a toll on stuff earlier than we'd expect even if a pitcher is not necessarily hurt.)
In any case, Holt busted. As have many otherwise successful Cyclones pitchers...and B-Mets and Bisons and so on. The point is that projecting pitching prospects is a highly inexact science. But that doesn't mean we can't learn and infer based on pitching stats from Brooklyn -- or any short-season league -- as long as we know what to look for.