How statistics have changed sports

Last night’s news about the film version of Moneyball being cancelled caused me to ponder how the recent increase in statistical analysis has affected the way we view sports. The effects certainly aren’t universal, as approaches have ranged from flat-out resistance (see Murray Chass and Harold Reynolds (no, not that blogger, Buzz) for a few examples) to full-on adoption and the creation of many new statistics (see Baseball Prospectus andFanGraphs, for example). Read on for more thoughts on how this expansion of statistical analysis has changed the way we view sports.

The above examples only look at baseball, where the battle has been most pronounced thus far, especially in the wake of the 2003 publication of Moneyball (pictured, right). However, similar conflicts have popped up in sports from basketball (see Bill Simmons v. David Berri) to football (Aaron Schatz, KC Joyner and others versus the multitudes who don’t like stats) and hockey (Gabe Desjardins and co. versus some of the traditional media).

The interesting thing is that very few advocate abandoning statistical analysis completely. Indeed, even some of the types many would describe as anti-stats use their own statistics to try and make their points (like batting averages, win-loss records for pitchers and quarterbacks, and goals and plus/minus for NHL players). The difference is that these are usually the more traditional statistics which people are used to and comfortable with, and they have their own value as well. They don’t necessarily explain things as well as some of the more complicated ones, but they are easier to use to get a point across, and can be quite powerful if used correctly.

It reminds me a bit of some of my calculus classes (which I mostly hated, but they had a point): you can sometimes choose between using a ridiculously complicated polynomial that represents a graph’s data points exceptionally well, or a linear approximation that doesn’t fit the data quite as well but is much easier to work with and explain. Both have their strengths and weaknesses, and both can be valuable when used correctly. The existence of one doesn’t preclude the relevance of the other.

In my mind, it’s the same way with statistics. There’s a need for both the simple and the complex. That doesn’t mean that the traditional statistics are always the simplest or best explanations (for example, on-base percentage is actually simpler than batting average and tends to have more explanatory power of a player’s worth), but it does mean that both goals need to be considered. It’s like anything in writing: you need to consider your audience. If you’re writing for a hardcore audience of statheads and trying to definitively evaluate players’ fielding capabilities, a statistic like Michael Lichtman’s Ultimate Zone Rating may work very well, but for a more general audience, it may be easier to utilize something like John Dewan’s +/- system, which posesses a fair bit of explanatory power while remaining simple to explain.

The great thing is that we do have more complex statistics and more information at our fingertips than ever before. This is a huge leap forward for analysis. Many people have spent countless hours working on new formulas and refining statistics, and we can now appreciate the work of those all over the globe thanks to the connective power of the Internet. Moreover, for those of us who like statistics, this new information allows us to be more objective in our analysis than even before. Instead of just saying something along the lines of “Scott Rolen is a great third baseman” or even “Scott Rolen makes a lot of plays going to his left”, we can use various metrics to calculate exactly how well Rolen does defensively, as recently seen over at Ghostrunner on First. That wouldn’t have been possible for the average fan even 15 or 20 years ago.

This doesn’t mean you need statistics to explain everything either, or that you have to try and sell everyone on their merits. As Joe Posnanski pointed out, statistics add whole new levels of enjoyment to sports for money, but they also detract from others’ enjoyment. There are some people who would rather read a player’s quotes after a game than read about his VORP, and there’s nothing wrong with that in my mind.

The unfortunate thing, in my mind, is that the statistics debate has frequently moved beyond subtlety, nuance and tolerance of different viewpoints into the realms of rants and anger. People on both sides have accused the other side of ruining their game, and that’s regrettable. It’s a wide world out there, especially with the expansion of the Internet and the growth of the blogosphere, and there’s room for many different points of view. You can enjoy sports through traditional stories or discussions of VORP and UZR. Neither approach works in a vacuum, and both have their merits. In my mind, that’s something we would do well to realize.

Tags: andrew bucholtz, geekfest, Statistical Analysis
Categories: Daily Specials, Domestics, Non-Alcoholic

This entry was posted on Monday, June 22nd, 2009 at 6:12 pm and is filed under Daily Specials, Domestics, Non-Alcoholic. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.