Tuesday, April 13, 2010

On Sabermetrics

As many readers of Cubs Stats may know, I recently joined the club baseball team at my school of postgraduate study. This proved to be the consummation of many years of yearning to play baseball, having played little more than sandlot ball for the last decade.

To my delight, the opportunity has allowed me several great experiences: I have been able to encounter first-hand the difficulties of a typical ball player (catching up with an inside fastball, playing on after committing an error, battling the sun for a fly ball, etc.); I have seen the disastrous effects of disparities in talent (suffice it to say we are 0-2); and I have had lively dugout discussions concerning statistics.

It is this last topic I want to examine and explore because many of my brethren can become quite heated when I propose something such as, "Walks can mean more than hits," or, "Sacrifice bunts tend to do more harm than good." It's granted that such statements already carry satchels of caveats across their broad shoulders, but our difference of opinion is vast nonetheless.

The treatise continues after the jump.

Let's examine a particular instance from our most recent game (a 21-1 drubbing). In that game, I -- as I usually and nerdily do -- was keeping track of the score. In the previous game, I had tried to keep track of hits, walks, strikeouts, doubles, triples, stolen bases, caught stealings, errors, runs batted in, runs, and, oh yeah, the score -- all to very disastrous effect (by the end of the game, no one was sure what the score was; we settled on 15-6). So, per the "recommendation" of the coach, I was tracking only tracking the score for our latest contest (I was also a designated hitter, so my attention was divided to say the least).

Anyway, somewhere around the 5th inning, just as things were getting really out of hand, the second baseman trotted over to me and my clipboard and asked the score. I think it was only 9-0 at this point. Then he asked: "What are the hits?"

"I don't know," I said shrugging, only half paying attention.

"What?!" he almost yelled. "You are so terrible at this! You've got to keep track of the hits!"

"Why?" I asked, still only half paying attention (I was watching a baseball game, so it was only natural he received the Wife Treatment).

"So we can keep track of batting averages!" he effused.

Ah! He had my interest now. I turned to him and lowered the clipboard. Cracking a smile I retorted: "Batting averages contain no valuable information." (Indeed, I talk like that.)

This fellow, who knew I had more than a crush on sports statistics and sabermetrics, understood my intention.

"Don't give me that sabermetrics crap," he said. "If sabermetrics really worked, then why hasn't Oakland ever won the World Series."

"The playoffs are a coin flip," I said, miming Billy Beane's famous claim. Of course, if my heart was really in the debate, I might have mentioned the ever-so-successful Red Sox (who run a highly sabermetric front office) or the incredible odds that small market teams such as the Rays and the A's face or the massively unsuccessful teams that avoid sabermetrics like the plague (i.e. the Royals).

The core of our divide seems to be these "old" statistics, the stats that enthusiastic and ill-informed fathers feed their tee-ball playing sons, or what I will call: The Legacy of the Limy. I believe Michael Lewis' analysis of old statistics -- as seen in Moneyball -- still rings true (from the bottom of page 69 through page 71):

In short, the old school stats were made by a limy who didn't properly appreciate our wonderful sport. Every time an athlete rolls his eyes at me insisting batting average is useless or kicks himself because his excellent range led to a hard-luck error, the athlete is giving credit to a man who deserves instead blame. The athlete is defying tested, proven (and American) statistical analysis and instead listening to the lies of 150 year-old formulas.

This second baseman, who's ever guilty of kicking himself for range-induced errors, wanted his batting average. We have only three games this season. If those two sentences don't raise a thousand stat-head red flags, then the world is upside down. What good would it have been to the second baseman to know he was hitting .350 or .100? Would that have changed his approach in our final game? Would that have affected his ability in the coming at bat?

No. He just wanted to let the Limy kick him.