A Treatise on Sabermetrics

Apologies for the lack of activity on this blog of late. I've had a lot on my plate, both school-related and otherwise, which has take precedence. I've got a big Matt Leinart post in the works, but for now I just wanted to clarify my position on one of the most controversial topics in the blogosphere: Sabermetrics. Some of you may have noticed the presence of several bizarre acronyms (WAR and ERA+) in my Josh Byrnes piece, so you probably already think you know how I feel about new statistics. But for the sake of argument, walk through the thought process with me.

Many prominent, old-line newspaper reporters make no secret of their disdain for "statistics," and the blogs that perpetuate them. However, almost to a man, they are lying about hating stats. I have never watched a Joe Morgan broadcast or read a Murray Chass article that didn't use certain statistics (usually Wins or Batting Average) to back up a point about a player. It's rule 1 of journalism: provide evidence for your claims. So what they are actually saying is that they dislike statistics that are new, or overly complicated. And there's some legitimacy to their claim that, if stats like RBI and Win/Loss records have worked for a century and a half, why would we feel the need to change now?

Except that such a claim is not dissimilar from saying that, because typewriters worked for an entire century, we should not have switched to computers. The invention of the computer provided a more effective way of printing words on a page, so it began to replace the typewriter. If we have a more effective way of analyzing the game of baseball, why not let it replace the system already in place? But suppose you believe, for example, that batting average still tells you more about a hitter than, say, on-base plus slugging percentage (OPS). In response to that belief, I would argue that simply watching the game should tell you otherwise, as there are two very clear problems with batting average.

The first problem is that walks are not factored into batting average. Therefore, a batter could potentially have four plate appearances in a game, walk all four times, and not have his batting average change at all despite his reaching base four times. Contrast this with a batter who goes 3 for 4, with all of his hits being singles. The second batter has reached base fewer times, and created more outs than the first batter, yet, assuming that his batting average is below .750, batter #2 will have raised his batting average while batter #1 sees his stay the same. Secondly, batting average does not weight different types of hits, meaning it treats a single with the same amount of importance as a home run. The problems with this should be clear: a player who hits 4 home runs in a game has clearly helped his team more than a player who hits four singles in a game. On-base plus slugging percentage was a concerted effort to avoid these problems, and while it has its own problems, it has proven to have a better correlation with team wins than batting average.

Pitcher wins and losses are an even more foolish and arbitrary way of evaluating performance. For those who don't know, a starting pitcher is awarded a win if the team is winning a game when he exits, and proceeds to hold onto the lead until the end. The obvious problem with this system is that pitchers have almost no control over their own team's offense, which is half the battle for gaining/maintaining a lead. This is obviously favorable to pitchers that happen to pitch for teams with great offenses. For example, if Chien-Ming Wang of the Yankees goes five innings and gives up nineteen runs, but the Yankees' offense scores twenty runs in that same span and holds the lead, Chien-Ming Wang would be awarded a "win." However, if Brandon Webb of the Diamondbacks goes nine innings and allows one unearned run, he would be awarded a "loss" if his offense does not score a single run. Based off of any other stat, Webb was clearly the more productive pitcher in this hypothetical scenario. But going just by pitcher W/L, Wang would appear to have had the better day, despite having given up eighteen more runs.

This website is not going to become fangraphs by any measure. My major is English, not mathematics, and I am not nearly intelligent enough to create mathematically-sound statistics a la Bill James. Statistics do not measure everything on the baseball diamond, and sabermetrics is still a developing field. However, as a writer who will be posting his opinions for the Internet to read, I owe it to you, the reader, to support my subjective assessments with proof in the form of statistics. The last thing that I want as a sportswriter is for my readers to consider me an idiot because of my methodology. I'd much prefer if you consider me an idiot for my points instead.