Now let me introduce you to two popular sabermetric statistics that FJMers are particularly fond of. The first is OPS+. This is on base percentage plus slugging percentage, adjusted for the park and the league. The second is EqA, equivalent average. This also adjusts for park and league. Both of these stats allow for historical perspective on statistics.
Why do I bring this up? Well, such league-adjusted stats allow sabermetricians to make objective analysis of players from different eras. Thus a player with a .900 OPS in 2006 is not as impressive as a player with a .900 OPS in 1976. Why? Because the league averages are higher in 2006 than they were in 1976.
So why are they higher? Well first off, it doesn't matter from statistical standpoint. You can adjust regardless of the variances. But from a practical standpoint, ballparks are smaller now, pitching is more diluted because of expansion, and players are bigger. Yep, that's right. Players are bigger, and part of that is because of steroids.
In other words, if steroids affect the entire league, then their effects can be statistically quantified and steroids users are "statistically punished." The crux of this argument is that steroids affect the entire league. So let's assume they don't, i.e. steroid is use rare.
Well most reasonable people think that players like Mark McGwire, Sammy Sosa, and Barry Bonds have all used steroids. We pretty much know that Jose Canseco, Jason Giambi, and Rafael Palmeiro used steroids. So if steroids use is rare, then it seems like is a sure fire way to generate huge offensive numbers. Think about, if only a handful of players were using steroids, then there is a high correlation between steroid use and offensive production. So it would seem like a simple equation "use steroids, hit 40+ home runs." So then wouldn't a lot of players use steroids? But wait, our assumption was that few used them. So that had to be wrong. A lot of players use steroids, including a lot of bad players.
So that means stats like EqA do a very good job of adjusting for steroid use. So steroid using players can easily be compared in a historical context using sabermetrics. The key is abandoning raw numbers like hit/home run/RBI totals, and using contextual numbers instead. You would think this would appeal to any sabermetric loving baseball fan.