The problem with this type of evidence is that it assumes the single change in the game was the presence or absence of steroids. To learn more from the data, check out the references I included at the bottom. The TL;DR is that there is no compelling statistical evidence that suggests steroids impacted homers, or offense in general.
The facile argument fails to address the fact that the game underwent expansion (weakening the average pitcher). The strike zone was changed a number of times (if you normalize for strikeout rate, the power in today's MLB is the same as in the 90s). Umpires got a lot better once PitchFX was able provide them feedback. (Lots were let go as well). Colorado started using a humidor, reducing home runs at the greatest hitters park of all time.
There's not really any way of knowing who did or did not use. MLB and mainstream sports media "strongly suspected" PED usage is closer to HUAC findings than real evidence. Also, it doesn't acknowledge the fact that a lot of pitchers were found using PEDs.
In short, it's incredibly facile to look at some superficial stats and hints of suspicion in order to reach the "steroids = homers" conclusion. But when you look at the underlying component numbers and adjust for the changing environment in the game, there is no evidence to support that conclusion. (That does not mean steroids had no effect, just that we cannot detect them with our best efforts.)
Here's an incredibly strong counter argument from Joe Sheehan's paywalled newsletter [I added it to pastebin for reference, I doubt Joe will mind]. http://pastebin.com/DXW0HSSt
Dan Szymborski is one of the top baseball data analysts in the country. From [0] (paywall)
"Despite the rhetoric surrounding PEDs, players caught for steroid/testosterone use do not show a pattern of overperforming their projections in the years leading up to the drug suspension or a pattern of underperforming their projections in the years after a drug suspension."
In layman's terms, if all you knew about players the past 10-15 years was their past OPS+ and whether they were busted for steroids -- now or at any time in the past -- it appears the PEDs had no noticeable effect on the projection of their future OPS+.
What this means is that, even with the knowledge of what outliers such as Barry Bonds accomplished while allegedly using performance-enhancing drugs, as a whole, there's extremely limited evidence of a significant effect on statistics of the drug users as whole. And without double-blind research studies of PED use among major leaguers and/or detailed information of what players are using, all we have to go by so far is the bottom-line results.
Now, none of this should be taken as endorsing the idea that MLB should simply open the floodgates and allow players to do whatever they want. Instituting drug testing is a very good thing for the sport -- but that improvement is for reasons other than the record books, such as the long-term health of players and the public trust. As far as the record books being tainted by PED use, well, it appears there isn't much evidence of that."
The facile argument fails to address the fact that the game underwent expansion (weakening the average pitcher). The strike zone was changed a number of times (if you normalize for strikeout rate, the power in today's MLB is the same as in the 90s). Umpires got a lot better once PitchFX was able provide them feedback. (Lots were let go as well). Colorado started using a humidor, reducing home runs at the greatest hitters park of all time.
There's not really any way of knowing who did or did not use. MLB and mainstream sports media "strongly suspected" PED usage is closer to HUAC findings than real evidence. Also, it doesn't acknowledge the fact that a lot of pitchers were found using PEDs.
In short, it's incredibly facile to look at some superficial stats and hints of suspicion in order to reach the "steroids = homers" conclusion. But when you look at the underlying component numbers and adjust for the changing environment in the game, there is no evidence to support that conclusion. (That does not mean steroids had no effect, just that we cannot detect them with our best efforts.)
Here's an incredibly strong counter argument from Joe Sheehan's paywalled newsletter [I added it to pastebin for reference, I doubt Joe will mind]. http://pastebin.com/DXW0HSSt
Dan Szymborski is one of the top baseball data analysts in the country. From [0] (paywall) "Despite the rhetoric surrounding PEDs, players caught for steroid/testosterone use do not show a pattern of overperforming their projections in the years leading up to the drug suspension or a pattern of underperforming their projections in the years after a drug suspension."
In layman's terms, if all you knew about players the past 10-15 years was their past OPS+ and whether they were busted for steroids -- now or at any time in the past -- it appears the PEDs had no noticeable effect on the projection of their future OPS+.
What this means is that, even with the knowledge of what outliers such as Barry Bonds accomplished while allegedly using performance-enhancing drugs, as a whole, there's extremely limited evidence of a significant effect on statistics of the drug users as whole. And without double-blind research studies of PED use among major leaguers and/or detailed information of what players are using, all we have to go by so far is the bottom-line results.
Now, none of this should be taken as endorsing the idea that MLB should simply open the floodgates and allow players to do whatever they want. Instituting drug testing is a very good thing for the sport -- but that improvement is for reasons other than the record books, such as the long-term health of players and the public trust. As far as the record books being tainted by PED use, well, it appears there isn't much evidence of that."
[0] http://insider.espn.go.com/mlb/insider/story/_/id/10922627/m...