I was watching this, uh . . . “interesting” video from the MLB Network the other day, and the responses to the role of statistics in the game were—well, for the most part, exactly what I expected. The panel comprised of Matt Vasgersian, Harold Reynolds, Barry Larkin and Sean Casey. A quick summary: Reynolds has a strong distaste for statistics and believe that they should not play an important role in player evaluation, Sean Casey believes that numbers cannot tell you who a “winner” is and who plays the game with heart, and Barry Larkin said the following:
“There is a place in the game for stats. But there certainly is a touch and a feel that gets lost in the translation when you simply start talking about stats. Stats are to compare a guy; you say you’re going to pay him this amount of money, why are you going to pay him this amount of money, because he compares favorably or less favorably to this guy that came before him. So, there is a position for stats. But Baseball in the clubhouses, in the trenches, is all about a touch and feel. […]”
I wholeheartedly agree, and I think he has the most fair and balanced (eat your heart out, Fox News) opinion of the panel. He’s absolutely right—there is a part of the game that numbers simply cannot quantify, and to try and say that advanced statistics encapsulates everything is just plain silly. Aspects of the game can be measured, yes. We can say with a reasonable amount of certainty that Player ‘A’ is better than Player ‘B,’ based on certain measurements that we believe accurately and objectively model reality. Oftentimes, the numbers support our observations—but not always. This means that either we’re hopelessly biased or that there is a flaw in the methodology. I’d like to think that it’s a mix of the two.
There are things that the numbers can’t (and most likely never will) tell us that are vitally important: the manner in which the player conducts himself on and off the field, and the knowledge he can impart on others in the clubhouse. Additionally, when looking at younger, less developed players, numbers can’t project the type of ceiling the player has. All of these things are absolutely essential when doing a total evaluation of a player, be it an established Major Leaguer or a young up and coming prospect. Numbers can help guide evaluation, but they certainly are not the end-all, be-all that many would like us to believe. The majority of (if not all) advanced statistics are theoretical, and there really is no one statistic that can tell you absolutely everything about a player’s contributions to his team. Advanced statistics aren’t always accurate, and they can be quite misleading (see: “Improving VORP”).
On the other hand, it is absolutely imperative to have at least a fundamental understanding of advanced baseball statistics if you intend to win games on a consistent basis. You can build a team full of “gamers,” but that means absolutely nothing unless they produce. Statistics give you a means with which you can maximize your chances of winning, and that’s the goal of all organizations. It’s what brings in money. At the professional level, baseball is not just a game—it’s a business.
Is there a place for statistics in baseball? Absolutely. And that’s the direction quite a few organizations are heading towards in their efforts to be perennial contenders. You’re all aware of the success of the money-deprived Athletics, and the annual dominance by the Red Sox (they’re not just a product of a huge payroll). Now teams like Seattle and Tampa Bay are doing the same (Seattle’s just beginning, while everyone witnessed the Rays’ quick rise to postseason glory). There are tons of statistics that are being cited as gospel, and some of them are terribly flawed to the point where they’re practically useless. And yet we still cite them as if they’re a definitive description of a player’s skills. Statistics are far from perfect—especially when we’re using flawed models. So while we don’t necessarily have to “tip toe” while using statistics, we need to pick and choose the more accurate metrics.
I love sabermetrics. I love reading the numbers, calculating them, learning new things about them, tweaking them—but I’m very much aware of their misgivings, and a lot of the statistical material that I present on here (although not all) are more theoretical than absolute, and that’s how they should be interpreted. I strive for objective analysis, and I feel that numbers are the best way to go about that. But in terms of predicting future success, mechanics and character, I’ll leave that to the gray-haired man in the stands with a radar gun every single time.