Steroids Before Testing

July 29, 2015



I recently had an exchange with a sportswriter and Hall of Fame voter regarding Barry Bonds and the steroid era. His opinion that Bonds deserves enshrinement wasn’t all that novel or intriguing, but his reasoning was. He was certain that Bonds used steroids, but felt that if a player was never suspended for using, then it doesn’t matter if he ever tested positive. It’s an interesting point of view, putting the entire value of the behavior on the sentence and none on the conviction. Given the history of our country’s mandatory minimum sentencing laws, the flaws in that line of thinking should be pretty self-evident so I won’t spend any time on it. But he also felt that Bonds was compelled to use steroids because, in his words, “everyone was using”. He later clarified that he meant that everyone was using before the 2003 MLB survey testing and then cited the Mitchell Report, several books and his own observations of big players and home run records as his proof.


So I took it as a challenge to see if I could discover if these sentiments contained any truth. First of all, I’ve read a number of those same books and while they make many broad claims, very few offered names. Jose Canseco’s books were probably the most noteworthy in this regard, and positive tests have since proved many of his claims true. But the number of players he exposed was relatively small. Ken Caminiti once estimated that 80% of the players were using, but a player guesstimating how many other players are secretly guilty of the same offense as he without having any evidence is about as useful as a creationist estimating how old the Earth is.  Gossip is rarely proof. And of course he has an incentive to suggest that the vast majority were using. Doing something wrong doesn’t seem as bad if he was one of many rather than one of only a few.


Since major league testing began in 2003, there have been roughly 860 positive tests returned. That includes all players in the minor leagues as well. So of the roughly 19,000 players who have played professional baseball in the US and Canada who have been subjected to the testing program instituted by Major League Baseball in the majors and minors, and that includes 3898 major league players, only about 4.5% have tested positive. Of those positive tests, 15% were for drugs of abuse like marijuana which aren’t considered performance enhancing. Does the testing catch everyone? Of course not. There will always be some who manage to avoid detection. But the testing is always evolving and it is always catching players who underestimated its ability. So unless one believes that testing is so far behind as to be pointless, the percentage of those caught is a useful measure.


That’s since 2003. However, I believe there is evidence to at least cause reasonable doubt regarding the commonly held notion that prior to 2003 everyone was using.


At least one team, the Baltimore Orioles, had mandatory testing clauses in the contracts of the majority of their players as early as 1986. This was disputed by the Players Association as an illegal invasion of privacy.  The Orioles stopped the practice but I’m not sure if it had gone to court whether or not their contention would hold up as numerous companies have similar testing policies. The owners historically have won most such cases over what they are and aren’t allowed to do.  Nevertheless it proves that some in the sport were trying to do something about the issue well before testing was instituted.


In 2001, the minor leagues instituted a new drug program, administered in two rounds of random testing to be taken during the season. It netted 9.1% positive results.  The following year, the number of positive tests fell to 4.8%. In 2003 it was 4%. By 2005, only 1.78% tested positive, and that number dropped to 0.36% in 2006. At least in one respect that first year of testing was significantly different than the first major league survey test: there was no half-year grace period between the announcement and the implementation of the testing in which players could get clean. So the 9.1% is probably a fairly accurate reading of the overall usage. It's also reasonable to assume that two years of testing in the minors had some impact on the number of players in the majors using even before the 2003 survey test. 


Some other things to note: minor league players don’t make much money. The average minor league salary is below the federal poverty level. So those players can’t afford to spend a ton on complex stacking supplements or masking agents. Despite the incentive to do everything possible to get to the big payday in the majors, only 9.1% were identified as choosing the drug route. Of course, their paltry income means that not a lot of them can afford to cheat, literally. But what about the players who can, namely major leaguers?


The Mitchell Report is commonly cited as proof that it was an epidemic in baseball. Kurt Radomski was just one distributor named in the report and his client list included 53 major league players. Surely there must have been dozens of others just like him. The Mitchell Report actually lists 87 players, but according to the report, 24 of those players used exclusively after survey testing began and only 36 were confirmed to have been using before the 2003 survey test. Of those, only 12 continued to use in 2003 or later. Additionally, the Report, which investigated three distribution sources (not just Radomski) plus several players not tied to one distributor, covered players using as early as 1993.  From that date until the survey test, 3128 players played Major League Baseball, which means the Mitchell Report, which admittedly was an incomplete investigation, proved only 1.1% of the players were using before survey testing. There were 27 other players named in which there was no date associated with their usage, but even including them moves the needle very little. And Radomski was considered a fairly big fish, distributing to players on a number of teams. There would need to be nearly a hundred guys just like him to get the percentage of users to 50%.


A more likely scenario is like the 36 players who were named as using before testing: two thirds of them stopped once the survey testing began. Assuming that ratio is representative of the rest of the players, since only about 4% of the players have since tested positive that figures to be only about 12-13% using before testing began. This is essentially how polling is done and for the most part, historically speaking, it has been accurate.


It only takes one dullard driving slowly in the middle lane of an interstate to cause a traffic jam behind him/her. It only takes one multi-million dollar lawsuit over spilled coffee where the actual details were not widely disclosed to make everyone believe that our justice system is nothing more than a lottery. There have been even fewer details revealed about “the steroid era” thanks to the lackadaisical job the sportswriters did in investigating it. I can think of only two – Jon Saraceno and Christine Brennan – who even made a legitimate effort to raise the issue. The point is that it did not require “everyone” in baseball to use steroids in order for some players to begin to voice their concern and force the sport to begin testing. It didn’t even take a majority using. It only took a few players, like Bonds, Canseco and Caminiti and those in the Mitchell Report, to change the way the game was perceived, how it was played and it’s relevance in history.


Baseball is a game of both statistics and impressions. If you attend 10 games over the course of a season and the same guy gets two hits in each of those ten games, there’s a decent chance that he’s a pretty good hitter. It’s certainly possible that he’s a bad hitter and you just happened to be there on his good days. But the more you see him and the more times he keeps getting two hits, the more it looks like those impressions represent the truth. Unfortunately with the steroid era we don’t have a complete record. But the more one looks into it and the more data that comes to light, the more it looks like the usage was less widespread than was previously thought and the excuse that “everyone was using” becomes less and less reasonable.