If one believes Barry Bonds when he asserts that he did not take steroids until after getting upstaged by Mark McGwire and Sammy Sosa in 1998,* one might revisit the 1998 National League Most Valuable Player race and write a story about the man who later became the face of the steroids era, a tale filled to the brim with irony and pathos—well, okay, unless you are F. Scott Fitzgerald it may be hard to drum up much pathos for your protagonist, but the irony is there and you might at least turn him into a sort of tragic hero.
But instead of focusing on what might have been, I want to write about what actually was. And it turns out the reality was probably quite a bit different than what I (and almost everyone else) thought it was at the time. Last week I asserted that a certain type of player—the Shin-Soo Choo type—whose contributions were spread out over several areas tends to be undervalued. Before diving into a broader historical survey, today I want to start by narrowing the focus considerably to the National League in 1998. It may be that the story of Barry Bonds was not that he was unable to compete with the juiced bats of McGwire and Sosa; it may be the that he was able to compete but hardly anyone noticed.
In part, this is the story of how statistical analysis is still evolving. At the time, I ranted quite loudly about how McGwire had been robbed. Looking at his advantage over Sosa in OPS+ (216 to 151) and runs created (193 to 149) it seemed clear that McGwire should have been the MVP.** But if we look at the race in terms of wins above replacement, it seems that McGwire probably did not deserve the award either:
Not only did Bonds significantly outpace McGwire, but Vladimir Guerrero rates a close third in WAR. Not surprisingly, the largest contributing factor to the new outlook is defense. Not only did Bonds and Guerrero add some value defensively but McGwire was actually a liability by that point in his career,*** according to Sean Smith’s fielding runs.
But maybe none of this is really too surprising. Even at the time most of us had an inkling that we were willfully deluding ourselves. But given the choice between knowing the truth and watching Mark McGwire hit some dingers…well, is that even a choice?
* And this part of his story sounds very plausible. A fun exercise I used to give my students when we were studying polynomial curves was to plot baseball players’ trajectories over the course of their careers. Typically, a quadratic model works fairly well. But for fun I would throw in Barry Bonds, for whom a cubic model actually fit better.
** What made the choice of Sosa even more dubious is that it took the questionable practice of determining and individual award based on the team’s finish to the ludicrous extreme: Sosa’s groundswell of support was due to the fact that his team finished with the fourth best record in the league, versus McGwire’s finishing seventh.