Steroid stats

steriodstats646.jpg

CATEGORY: Academics
TYPE: Articles
RELATED PROGRAMS: Mathematics, Statistics, and Computer Science

Before he received his Macalester diploma, Jeremy Roth ’10 had already published groundbreaking research on steroid use in baseball. This summer his findings with math professor Victor Addona showed up in print again—this time, in ESPN The Magazine.

In a column about measuring the impact of performance enhancing drugs (PEDs), ESPN writer Peter Keating cited Roth and Addona’s 2010 article in the Journal of Quantitative Analysis in Sports on the effects of PEDs in baseball. Keating referenced it online in a blog post that year, then again this summer in the magazine, which reaches a wider audience. Their paper, Keating notes, is the only research he’s found connecting PED use to performance changes in baseball: “Interesting, but just the first droplets of data we need to clean away the effects of doping from MLB stats.”

When Roth and Addona began their research, they, too, were surprised to discover that their findings were among the first—a gap that likely exists because of the combined challenge of classifying players as steroid users and finding the appropriate baseball metric to measure steroids’ effects. “There was so little work done to know if steroids were actually helping players,” Addona says. “Increased strength is perhaps a more obvious consequence, but it doesn’t necessarily translate into increased performance.”

The Macalester researchers’ goal was to document a measurable outcome, and Roth, a baseball fan who had taken three statistics classes from Addona, chose to study fastball velocity for pitchers who’d been linked to PEDs. For the control data they used average fastball velocities for all major league pitchers from 2002 to 2008.

To form a comparison group to test steroid effects, Roth combed through the Mitchell Report, the 409-page investigation of PEDs in baseball led by former Maine senator George Mitchell. “The big issue is, how do you classify players as steroid users?” Addona says. “Whoever did the analysis would run into this problem.”

Because so few players had actually been suspended for PED use, Roth grouped those proven users with players linked to steroid use in the Mitchell Report. He and Addona found that pitchers who had used steroids experienced an average velocity increase of 1.07 miles per hour—one of the first studies to document a quantifiable change.

Through his research, Roth—whose previous perception of PED use in baseball had been limited to well-publicized stories about highly productive players—got a broader view of the issue, especially regarding the players on the cusp of making the major leagues (and thus receiving major league salaries). “There’s a powerful financial incentive for minor league players to turn to PEDs out of desperation to claim a major league roster spot,” he says.

In addition to their work examining fastball velocity, Roth and Addona also published an article on steroid effects on batters, which appeared in the Journal of Recreational Mathematics in 2012.

Roth is now pursuing a PhD in biostatistics at the University of Washington. Addona continues to work on sports topics with Mac students, most recently teaming with Justin Sims ’14 (Bronx, N.Y.) to explore how player age affects the Major League Baseball draft.

PUBLISHED: 11/01/2013