What Do I Know?

By John Benson

I don't know how many times I've heard fellow public opinion researchers say some variant of this: "Baseball is okay as a sport, I guess, but it's baseball statistics that hooked me as a kid." Let's face it, baseball has the best statistics, and they served as a gateway drug for our future addiction to numbers. No other major sport has the face-to-face confrontation between pitcher and batter that yields reams of data, in a trend file that goes back more than a century... before Gallup! So while I mainly played basketball and soccer and football, I did due data with baseball.

That's why the recent steroid scandal is so maddening to our community, most of whose members never lettered in a sport. Baseball records are sacred, even to people who like the game for its own sake. Batting and earned run averages, RBI and stolen base records have been kept since before your mother was born. And the most sacred statistic of all is the homerun record. So important has it been to the mythology of the sport that when Roger Maris broke Babe Ruth's longstanding record for homeruns in one season, he was pilloried and his record temporarily asterisked because the season was eight games longer in Maris's day than in Ruth's.

It is now clear to all, as it should have been to anyone who applied the most basic skepticism about the data when they were fresh, that the homerun binge of 1998-2003 was fueled by steroids. And if the boom in homeruns had more to do with steroids than with suddenly improved conditioning, what's a data geek to do? Statistically, not much.

We who reveled in the extensive but relatively simple statistics of baseball during our youth recently entered a Golden Age, when new formulas and weights have given us more sophisticated tools for analyzing the sport's data. One of the biggest appeals of baseball stats is that they form a continuous history, an unbroken tradition from the beginning of the industrial age in America up to our postmodern times.

The biggest problem is that players' statistics are affected by a wide variety of factors that have little to do with their abilities. Baseball has had dead-ball and live-ball eras, big and small ballparks, parks that favor lefties or righties, pitchers' mounds of different heights, leagues with and without designated hitters, expansion years when the talent pool was watered down, periods when many of the best players were absent because of war or the color of their skin.

Simple measures like batting averages and RBIs have been giving way to OPS, a combination of on-base and slugging percentages. Bill James (author of The New Bill James Historical Abstract) has developed a formula for calculating Win Shares, which greatly diminish the effect of ballpark size and shape and the era in which a player performed, in an effort to help us compare current players with each other and with players of different eras.

In short, baseball has become a metaphor for our professional experience. We are finding new ways to arrive at better estimates by, in effect, applying weights to the raw data.

But now we are faced with a serious problem of incommensurability because some of the most prominent players may have been taking unfair pharmaceutical advantage. Frankly, there's probably not much we can do to compensate statistically for individual athletes' acts if they are not proven. We cannot simply assume, without proof or confession, that everyone who hit a lot of homeruns was using steroids and banish them from the record books. And it would be unfair to the batters of recent times who did not use steroids to weight down the homerun and run-scoring statistics because other players cheated to bring up those numbers.

John Benson is managing director of the Harvard Opinion Research Program, Harvard School of Public Health, and associate editor of Public Opinion Pros.

Readers who wish to respond to this or other articles appearing in Public Opinion Pros, or to contribute commentary of their own in 800 words or less, should consult our author submission guidelines and editorial policies under" Letters to the editor and op-ed articles." Op-ed submissions must be received at least two weeks before the first of the month for the issue in which they are to appear.