Statistical Blips and Possible Steroid Use
Dugout Central has run a couple of very long, but quite interesting pieces the past two days entitled “Statistical Blips and Possible Steroid Use.” Part I is here. Part II is here. The premise, very crudely put, is that the remarkable thing about the PED-implicated players of the current era is not necessarily the raw number of home runs they hit -- eras and ballparks and equipment change, after all -- but the deviation from the baseline, be it their own established norms or league averages. “Blips” is the term author Bill Wellman uses, as in “Barry Bonds averaged 41 homes a season outside of 2001, so his 73 that year was a “blip.”
What Wellman does is to look at blips throughout modern (or at least post-WWII) baseball history, not just the post-1987 (i.e. Jose Canseco) period. Why? One reason is to figure out how much the numbers really were inflated by steroid use in the generally accepted “steroid era.” The short answer is that, yes, there’s a clear bump post-1987, as you might expect.
But a far more interesting reason for his research is to determine if 1987 and the emergence of Jose Canseco truly represented something new or if, alternatively, there is any evidence suggesting a gradual increase of PED use -- or at least “blips” -- over time. Wellman thinks this is important -- and I do too -- because it seems nonsensical to simply establish a 1987 starting date for “the steroid era” based on when Jose Canseco showed up. Steroids existed before Canseco. We even know of baseball players who used them before him. It stands to reason, therefore, that they were in the game before Canseco arrived.
In light of this, Wellman breaks down the post-war period into three separate eras -- The Improbable Era (1946-90); the Anecdotal Era (1961-86); and the Steroid Era (1987-present) -- and tries to figure out what effect, if any, PEDs may have had on home run totals. he then goes on in Part II and breaks down the blips of multiple famous players, including Hank Greenberg, Carl Yastrzemski and the usual, more modern suspects, in many cases wondering whether or not PED use could have played a role in the blips.
I find the raw data interesting. I’m less sold on some of Wellman’s speculation (though it should be noted that he’s not throwing bombs here). Players aren’t automatons who can be expected to perform at the same level every year. Fluke seasons happen. As do seasons when Major League Baseball quietly experiments with things like lively baseballs, as it certainly did in the 1930s, almost certainly did in 1987 and is rumored to have done post-1993. Weird things happen, so just because something odd occurs doesn’t mean something nefarious is going on.
But the general conversation is a worthy one. While new names continue to emerge, for better or worse, we are still very much in an era where people cite statistical anomalies as evidence of PED use. If we’re going to continue to do that -- and if we’re going to make Hall of Fame decisions based on such anomalies -- we should at least be more thorough in doing so.
Wellman tries todo that here. His work in this regard is not perfect, but it is a start, and for that reason I recommend that anyone interested in the subject take a look.
UPDATE: I had not seen this when it first ran three and a half years ago -- I actually had a life then and wasn’t blogging about baseball yet -- but apparently Baseball Prospectus’ Nate Silver did this sort of thing back in March 2006. And, I might say, did a better job of it.
Still, I think Wellman’s piece is worth reading, if for no other reason than it appears that people decided to ignore Silver’s findings that players back in the day were just as likely as modern players to have fluke seasons, spikes, blips, or what have you.