Baseball Analysis.... dead?

When I was a wee lad, I collected baseball cards and poured over the statistics on the back side, creating a fascination with numbers and numerical calculations that has propelled me into a statistics degree and a job as a statistician. But even as I work in a retail setting, my first love of numbers always will be baseball statistics. It is why I joined SABR, and why I have continued and growing interest in sabermetrics.

With this in mind, one can imagine what my initial reaction was to reading these words: "Baseball analysis is dead." Now, I need not detail my direct disagreement with the statement made by Mr. Huckabay, as it does not add to what has been stated by many a person already (see Rob Neyer's ESPN blog if you have access or the response by Dave Studeman on Hardball Times). Rather, I feel that it is not a funeral march that was written, but a challenge to the entire sabermetric/baseball analysis community.

Mr. Neyer and Nate Silver of Mr. Huckabay's Baseball Prospectus, as noted by Mr. Studeman, both cite that sabermetrics is an analytical science, albeit covering a much more finite realm than a more traditional science like biology. As with any science, discovery can be classified into two broad groups: ideas with practical application and ideas that foster new ways of thinking about the subject. Mr. Huckabay's view is focused almost exclusively on the former, and it is to this end he is challenging the sabermetric community to reach for. He is mindful that baseball organizations have far more data than is available publicly (even on his own website), particularly scouting data, and while the use of sabermetrics like WARP, Pythagorean records, and FIP is informative to the fans of the game, it is the organizations who stand to profit the most from the science. If a sabermetrician does not strive for discovery which can immediately impact or lead to another discovery that impacts front office thinking, then the practice of the science is in vain.

Perhaps my favorite example of this latter idea is only double Nobel laureate in physics, and a genius by any standard. John Bardeen, while at the University of Illinois in Chambana, IL (that's Champaign-Urbana for those unfamiliar), made two extremely important discoveries that modern society now takes for granted as a part of everyday life. He invented the transistor and, later on, the superconductor. Both of those discoveries are enabling the PC you are currently using to read this to even be in existence, as they are the fundamental building blocks of the computer, like protons and nuetrons in chemistry. Now, did he know then what his discoveries would lead to? Not in full, I'm sure. But his discoveries lead to a revolution in how society interacts and operates. This is what sabermetrics must do as well: make discoveries that lead to impact, directly or indirectly.

Going back to BP or Hardball Times or any other publication that includes sabermetric articles, and this much is clear: those meta-discoveries that can lead to the discoveries of direct impact are still being made.