Greeting

A Baseball Blog - Scientific and Speculative Thoughts from Third Base

Monday, June 27, 2011

A Late Scorecasting Mini-Review

Okay, I’m back from a trip, and can resume regular updates now. Over time, though, hopefully more of my old articles will find their way to Hot Corner Harbor. Anyway, on to today’s main article.



One of the things I managed to do on my trip was finish Scorecasting by Tobias J. Moskowitz and L. Jon Wertheim. I had been meaning to read it for a long time, and started it months ago...I just hadn’t gotten enough time to finish it. I have to say, it is a very compelling book. As a baseball stat fan (obviously), some of the data was less absolute than I would have liked, but seeing how the topics dealt with most or all sports, it is understandable, I suppose. The book feels uneven in parts, as you can tell the authors had more to say on some topics than others (for example, the chapters on home field advantage are twenty-six and thirty-two pages, but the chapters immediately surrounding them are only five pages each). The authors do offer support their arguments for the most part, and you can usually follow their work up to their conclusions. They foresee some of the criticisms that may be offered, and manage to account for some of them (I found several instances where I was about to dispute a claim only for them to include something on the case I had thought of).

There is one problem I had though. There are several instances where their conclusions feel...off. I can’t think how to describe them (and it doesn’t help that I read some of the sections in question so long ago). Most of the time, the conclusions in question aren’t the main point of the chapter, but are rather a detail that the authors felt compelled to share. I don’t know if these details were not as well researched, or if the authors were trying to reach a pre-determined solution, but it felt like they missed things. The first example that comes to mind is the section on drugs in sports, specifically steroids.

Moskowitz and Wertheim analyze the pervasiveness of steroids in baseball, noting its greater prominence in the minor leagues (particularly in the higher levels) and in foreign born players (particularly from Latin America). They do a good job of refraining from pointing fingers on the issue or issuing blame, noting the economic incentives that may drive these results. It is, for the most part, a good, well-thought-out chapter. They point out the lack of drug laws in the aforementioned countries and the greater financial rewards in reaching the Majors when starting from poorer nations. They also hypothesize that the higher minors see more positive tests for steroids because players on the border are seeking an edge to push them over (thereby earning a larger salary, as they mention that MLB players get much higher paychecks than AAA players). All in all, this does make sense. However, towards the end, they try to determine whether steroids actually do help players. More informed people than I have tackled this issue (just one example; Joe Posnanski’s summary is here), and the data isn’t as clear as one would think. My thinking is that it’s a complex issue, and at the moment, there isn’t really a definite answer. There may never be.

However, the authors of Scorecasting seem quite sure of their result. They determined that 60% of positive testers reach the next level of play, and therefore, steroids help. My biggest problem with this is that it tries to oversimplify things in order to reach an easy conclusion, even though there might be no conclusion at all. Even though the ones most likely the ones taking steroids are, in theory, the ones already on the borderline of getting called up, and there are several other scenarios that may lead to the call-up, Scorecasting seems more than happy to wave this as proof. We don’t know how long these players achieved the next level for, or whether other factors (like an injury at the next level, or a hot streak) necessitated the call-up. The results themselves weren’t much better than a coin flip as is, considering these borderline players should be somewhere around 50/50 in their chance of a call-up to start with (and even then, I would think that it’s rare borderline players aren’t given some sort of a chance at the next level).

Again, overall, Scorecasting is a good book, and the writers did say they weren’t trying to provide absolute answer for everything, only cause discussion. I’m just worried that their matter-of-fact delivery of such a vague and difficult to quantify subject will do the exact opposite.




Now that that’s out of the way, I have big things planned. Hopefully, I’ll have a one shot article up later in the week, followed by the start of a series that I’ve been researching for some time.

1 comment:

  1. As a side note, I had been purposely been holding off on Phil Birnbaum's review of the book to prevent myself from reading any spoilers. I just read it, and he touched on some of the earlier sections of the book, and reached a similar conclusion to mine:

    http://www.baseballprospectus.com/article.php?articleid=13003

    ReplyDelete