Last week I talked about why I believe that translating college statistics without scouting context leads to bad data, and is therefore relatively worthless. At that time I focused on the hitting side of things, so let’s move to the mound.
In the hitting piece, I talked about how many hitters are capable of putting up excellent batting average, power and on-base numbers without necessarily having the kind of baseball skills that will lead to a professional career. The same is true for pitchers, but maybe even more so. The problem here, once again, is the disconnect between statistics and raw stuff. There are certain kinds of pitchers that can succeed in college and at the lower levels of the minors, but the manner in which they succeed will not work in the majors. These pitchers, while often lacking plus velocity, have impeccable command and the ability to spin a breaking ball or change speeds well. These types of pitchers can put of the kind of numbers that can fool the stats-only observers into believing that they have a future star on their hands, but in the end, without a true “out pitch” to depend on, these pitchers have little margin for error, and that margin shrinks exponentially with each level they advance.
Before I go on, I’d like to speak for a moment about Jamie Moyer. Every time somebody in the prospect world points out that a pitcher who fits the lefty/command/finesse mold will have trouble succeeding at the higher levels, he gets inevitably compared to Moyer as an argument in favor of the player. Jamie Moyer is what is commonly referred to in the scouting world as a “freak,” and in the statistical analysis world as an “outlier.” That lefty who rips up Double-A while throwing 85-87 mph is not the next Jamie Moyer. Every team in baseball has one or two of these guys in their system, and most don’t make it at all–and those who do are limited to LOOGY status or spot starting at best, while one out of maybe two or three hundred actually turn out to be Moyer.
When looking at hitters, I defined a set of parameters that would identify the top statistical performers. Let’s do this again with pitching. Once again limiting ourselves to the four power conferences in college baseball (ACC, Big 12, Pac-10, SEC), here are the pitchers from 2003-2005 that fell under the following criteria:
- Were among the top 10 in the conference in ERA
- Struck out at least one batter per inning
- Had an K-to-BB ratio of at least 3-1
YEAR PLAYER SCHOOL ERA IP H BB K DRAFT 2005 Cesar Carillo Miami 2.22 126 100 23 127 2005-1st 2005 Max Scherzer Missouri 1.86 106 59 41 131 2006 elig. 2005 Ray LaMotta Baylor 2.15 80 60 21 82 2006 elig. 2005 Jonah Nickerson Oregon St. 2.13 110 88 29 114 2006 elig. 2005 Ian Kennedy S.Calif. 2.54 117 85 38 158 2006 elig. 2005 Jensen Lewis Vanderbilt 2.62 93 74 23 95 2005-3rd 2005 David Price Vanderbilt 2.86 69 51 30 92 2007 elig. 2004 Michael Gross N.C. State 2.24 52 44 12 59 2004-33rd 2004 Andrew Dobies Virginia 3.41 108 102 30 109 2004-3rd 2004 Huston Street Texas 1.58 57 36 13 59 2004-1st (sup) 2004 J.P. Howell Texas 2.13 135 90 53 166 2004-1st (sup) 2004 Mark Alexander Missouri 2.14 55 50 18 59 2004-20th 2004 Ian Kennedy S.Calif 2.91 93 86 31 120 2006 elig. 2004 Derek Tharpe Tennessee 2.01 81 66 19 84 2004-6th 2004 Scot Drucker Tennessee 2.61 59 49 19 65 2004-13th 2003 Joey Devine N.C. State 2.19 66 49 16 78 2005-1st 2003 Kyle Sleeth Wake Forest 2.81 96 78 29 102 2003-1st 2003 Vern Sterry N.C. State 3.25 116 99 39 124 2004-8th 2003 Zane Carlson Baylor 2.61 59 46 14 71 2004-27th 2003 Jeremy Sowers Vanderbilt 2.50 115 94 29 123 2004-1st
Like the list of elite hitters, once again we have a mixed bag here of elite prospects and fringy talents. The pitching version of Jeremy Cleveland, mentioned in last week’s piece is North Carolina State’s Vern Sterry, who had an outstanding two-year career with the Wolfpack that looks like a line from an elite major league starter:
G W L ERA IP H BB SO 35 20 2 2.74 230.2 191 62 230
Impressive raw numbers, but once again, the more important piece of information is not in the achievement, but in the method of achievement. Sterry’s numbers could fool the casual observer into thinking that he’s a dominating arm, but he did this with good command and a plus changeup that he sets up with a fastball that sits only at 85-87 mph. Once again, it’s the kind of stuff that can work against college and low-level pro hitters, but not in the majors. That’s why despite having some of the best statistics in college baseball, Sterry was just an eighth-round pick. In this year’s draft, one in which college pitching is the only area of strength, some of the best numbers are being put up by Rice’s Eddie Degerman, who has a 1.29 ERA in 69.2 innings with 92 strikeouts while limiting hitters to a batting line of .168/.241/.212. Those are remarkable numbers for a pitcher at a bigtime program, yet Degerman’s stuff will keep him from even sniffing the first round.
I got a considerable amount of e-mail in response to my last piece, and a good percentage of it was from readers who disagreed with my assumptions–accusing me of shooting at fish in a barrel, or pointing out the progress the some analysts have already made in this arena. In Baseball Prospectus 2006, I have a quote in Gary Huckaby’s excellent essay on where statistical analysis fails that I’d like to build upon. I’m sure many of our readers, including those who wrote to me, are multi-sport fans. Here are the Heisman Trophy (football) and Naismith Award (basketball) winners for this decade, handed out to the best player in the sport.
YEAR HEISMAN NAISMITH 2005 Reggie Bush Andrew Bogut 2004 Matt Leinart Jameer Nelson 2003 Jason White T.J. Ford 2002 Carson Palmer Jason Williams 2001 Eric Crouch Shane Battier 2000 Chris Weinke Kenyon Martin
Again, a mixed bag filled with stars, role players, and fringy pro talents. While the world of performance analysis in both football and basketball lags well behind that of baseball, it has always surprised me that fans readily accept the fact that the top college football and basketball players will not make the best pro players in their respective sport, yet still insist on searching for a college baseball Rosetta Stone that doesn’t exist. Instead, isn’t it possible that the games are so different, that statistics alone don’t tell us everything we need to know?
So where do we go from here? By no means am I saying that college statistics are not valuable pieces of information, or that anybody working on translating and/or balancing college numbers should abandon the project. However, I think the true objective here would be to find a way to get a base translation and then adjust by finding a way to quantify the subjective side of things. This would be a tremendous challenge, and one that could only be done by a team, as they’re the only ones that have access to professional-level scouting report for every player. Developing coefficients for each aspect would also require a strict adherence to standardized grading, but the value one could find here borders on groundbreaking. This is the true challenge when it comes to making college statistics useful in evaluating future projection, and it’s a monumental task.