The Cheater’s Guide to Baseball, Barry Bonds’ Steroid Cycles in 2002

Baseball’s history is rife with cheating, many say it is part of the game. In The Cheater’s Guide to Baseball, author Greg Zumsteg outlines all forms of cheating in baseball over its history. From gambling to corked bats to stealing signs, Zumsteg not only describes the well known ways of cheating but also devotes chapters to lesser know ways of manipulating the game like intimidating umpires to get calls and creative grounds keeping.

Of particular note to this site is the chapter on steroids. Zumsteg playfully lays out the history, the ramifications, and moral issues concerning steroids in baseball before taking a more analytical approach. In an attempt to understand the statistical affects of being on steroids, Zumsteg maps out Bonds’ 2002 season in a series of hypothetical four week cycles of steroid use.

When using steroids it’s important to cycle on and off the drugs to prevent testicular atrophy. According to Game of Shadows, Bonds used steroids for three weeks then took a week off before starting a new cycle.

“During a three-week cycle, Bonds was injected with human growth hormone every other day. Between injections, he alternately used (BALCO President, Victor) Conte’s two undetectable steroids, The Clear and The Cream. At cycle’s end, Bonds took the prescription drug Clomid; doctors prescribe it to women for infertility, but Conte thought it helped his clients recover their natural ability to produce testosterone, which was suppressed by steroid use. Conte recommended a week off between cycles.”

Again according to Game of Shadows, Bonds complained that between cycles he felt weak and that it affected his play.

“He could feel the drop in energy that came when he was cycling off the performance enhancers, and he could see it, too, in the distance the ball traveled. He was mindful of the distance of his home runs and his splash hits, and when his power started to decline he would tell (Greg) Anderson to start him on another drug cycle, according to a source familiar with Bonds. Anderson kept the calendar that tracked his cycles.”

So Zumsteg set out to determine if Bonds statistics in 2002 followed a pattern similar to that of cycles of steroids. He took every day in April, the first month of the season, and mapped out Bonds’ statistics with three weeks “on” and then one week “off.” With each day of April a hypothetical start date for a cycle Zumsteg

Zumsteg hypothesized.

“What we’d expect to see, if a player isn’t using or doesn’t perform better while on steroids, is that the “on” weeks would look, more or less, like the “off” weeks. If they were on, we’d expect to see not only an increase in performance during the “on” weeks but that the dates we suspect the cycles start on would be fairly tightly grouped.”

Zumsteg surmises that April 1 is the most fitting start date. Bonds’ slugging percentage during “on” weeks starting that day was an astonishing 345 points higher than the “off” weeks. The next most likely starting date according to Zumsteg was April 28th, because the cycle would repeat after 28 days. He found similar patterns in Bonds’ 2001 statistics.

To test his analysis Zumsteg had Baseball Prospectus run some stats.

“To test this schedule, I went to Keith Woolner at Baseball Prospectus and had him run it for other Bonds seasons using the detailed game logs. In 1997 and 1998, before Bonds supposedly decided to take up steroids, we find that there are no strong patterns; the difference from even the best candidate dates show about half the difference in power as found in 2002.”

So Bonds’ stats seemed to mirror his cycles of steroids. Could statisticians reverse engineer such patterns in other players over the last 15 years or so? Almost certainly. But there are many factors that would keep it murky at best. Some players presumably used steroids just in the off-season. Some used sophisticated cocktails of steroids and human growth hormone, while others just experimented. The list goes on. Nevertheless baseball is a game of statistics and statistics can be normalized with the proper algorithm.

Once we understand performance-enhancing drugs’ effects on the game we can go about putting these players in their proper place in history. This is a step in the right direction.