Tuesday, July 28, 2009

Sports & Steroids

Your opinion, how much of baseball is shooting steroids into their arms (or legs, or butts, etc.)?

In most debates I take the low side, most people seem to think the number over half, and some argue that it's "everybody". Clearly something seems to be driving up the numbers of the "long ball" over the past decade or so, but how much is steroids responsible for it?

Plenty of other factors could provide the same statistical boost. Better training, changes to rules, new ballparks that are much more friendly to hitters, etc. Most of these should affect all players equally though, so I've made a couple of assumptions to take a look at the numbers.

Assumption 1: The top long ball hitters in the league will be doping at a different rate than the rest of the league.
The dramatic performance increase of the upper echelon (Barry Bonds, Mark McGuire, etc.) in recent years is what brought about most of the talk of steroids' role in baseball. On the other hand there is the argument that it's the lower ranked players that need it more and are thus more likely to dope. Either of these is possible, as long as one is true we should see an effect if we compare the two. If we see no effect then we'll know this is a bad assumption.

Assumption 2: The "steroid effect" outweighs other factors.
This is much harder to justify. I began using this assumption solely due to the common perception that steroids were the driving factor behind the home run rates increasing so significantly. Once the data was in though, this assumption was partially justified.

Here is how the slugging numbers compare between the top 10 hitters (averaged) and the entire league since 1986.

Note that both groups increased their slugging percentage through the past two decades, but the "top 10" did so at a much faster pace. If our assumptions are correct we could say that the upper echelon of baseball is definitely doping at a higher rate (or at least is more effective with it's usage).

More interesting though is the recent drop-off that the top players have shown while the league average has leveled off. This provides some support for our second assumption, as the league crackdown on drugs would be expected to reduce usage, particularly within the high-profile players since they draw the most scrutiny and have the most to lose.

How far has it dropped? We'll divide the two values to get Harx's Baseball Doping Index.

This shows that the top echelon is now in relation to the league average at about the same point it was in the late-1980's (considered very early in or before the steroids era). This doesn't mean that steroids are out of baseball, just that it is no longer providing the same distinct advantage to the top players.

So to the question, how many players dope? At the peak there was about a 50% disparity between the two classes, while during "pre-steroids baseball" the rate was closer to 40%. That doesn't tell us much as far as number of players, but it does tell us that around 10% of the performance of top players (at the peak) was due to doping.

So though I can't answer my original question, I can at least say that it appears the league's efforts to contain the problem have been effective. A better answer than I had originally hoped for.

No comments: