Skip to main content
Menu
Steroids Probably Aren’t Causing Baseball’s Power Surge

Ever since the steroid era, baseball fans have been cautious about reading too much into unexpected leaps in player performance. Today’s jaded rooters assume that performance-enhancing drugs are to blame, somehow, whenever a player experiences a breakout or the league undergoes a transformation. So it’s no surprise that, with home runs flying out of parks at nearly an all-time high in 2016, many have reacted by assigning the responsibility to some undetectable new PED.

After exhaustive consideration, my former FiveThirtyEight colleague Ben Lindbergh and I concluded that the most likely culprit for 2016’s home run explosion is a change to the construction of the ball. Still, given MLB’s lengthy history with PEDs, I thought it was worth revisiting whether a chemical explanation could possibly shed light on the most recent offensive uptick. And there are some similarities between the steroid era and the present — but 2016’s home run explosion is also missing a few key characteristics that defined the steroid era.

One of main reasons PEDs seem an unlikely explanation for baseball’s recent offensive surge is the suddenness with which home runs increased. The league’s rate of homers per game began climbing around the 2015 All-Star break, and this year it’s risen more than 30 percent compared with 2014. In theory, PEDs make their way into the game slowly, with knowledge being passed between players over the course of years, but changes to the game’s equipment could drive a more rapid increase. A new supply of slightly altered game balls would affect all players in the league at once, so even a small modification to the ball’s properties could produce a massive statistical change across MLB. Given the quick, drastic shift we’ve seen in home run rates, a change to the ball appears to require fewer leaps in logic: players take time to become juiced, but balls can become juiced immediately.

To further investigate whether the ball — and not PEDs — explains MLB’s quick home run increase these past two seasons, I compared the steroid era with a handful of other times in history when the ball’s construction is known to have changed. Although the build of the modern ball has been nominally consistent since 1976, MLB has an extensive history of openly altering the ball in earlier epochs. The first alteration in baseball’s modern era (since 1901) was the introduction of the cork-core ball in 1910, after which the league’s batting average jumped 17 points. In 1943, wartime rubber shortages forced manufacturers to make baseballs with a substitute called balata, which had different elastic properties. Runs per game fell that season by 0.17, before rebounding by 0.26 runs per game in 1945, when the balls reverted to rubber cores. And in 1974, the surface of the ball changed from horsehide to cowhide, without much discernible impact on the league’s overall statistics.

In addition to the times in MLB’s history in which the league has admitted to altering the ball, conspiracy theorists have posited numerous other instances in which the league may have modified the sphere in secret. Perhaps the most intriguing parallel to the current home-run increase came in the late 1980s, when Rawlings, MLB’s official baseball manufacturer, shifted production from Haiti to Costa Rica after the collapse of Haitian dictator Jean-Claude Duvalier’s regime. From 1988 to 1990, baseballs were produced in both countries before the Costa Rican facility took over completely. Over that period, slugging percentage immediately fell by nearly 40 points, taking seven years to creep back back toward its 1987 level. This production switch is especially intriguing because of shake-ups at Rawlings in the middle of 2015: although it didn’t change where the balls are manufactured, it did move part of its factory operations and laid off 200 employees.

To get a better sense of whether MLB’s current offensive spike more closely resembles the steroid era or one of the times the ball changed, I charted home runs per plate appearance around a few of the historical instances of known ball-tinkering, as well as the midpoint of the steroid era.

arthur-steroidera-1

Most historical instances of the ball changing were accompanied by rather dramatic shifts in HR/PA, which helps confirm that the makeup of the baseball can indeed influence power rates.1 And 2016 currently holds the second-largest two-year shift in HR/PA since 1901. Then again, the largest change came in 1994, at the very beginning of the steroid era. (It’s impossible to precisely date the onset of PEDs in the game, but the massive increase in home runs, coupled with changes in the productivity of older hitters and an increase in outlier seasons at the same time, suggests that 1994 is as reasonable a guess for its beginning as any.)2 But although that means PEDs could also explain the current offensive leap, the start of the steroid era came with a number of other statistical changes that aren’t being repeated in 2015 and 2016.

In addition to the dramatic rise in home runs per plate appearance, one of the hallmarks of the PED era was a jump in the average age of hitters. Instead of withering away, many older hitters remained productive into their late 30s and early 40s, in some cases putting up their best seasons toward the end of their careers. (We’re looking at you, Barry.) MLB’s average plate appearance-weighted age in 2005 was 29.3, the second-highest history behind 1945, when many young would-be players were at war. Most of the top 10 years for weighted player age, whether you weight by wins above replacement or by plate appearances, are either within the steroid era or around World War II.

By contrast, the average age of hitters hasn’t undergone much of an increase between 2014 and 2016. If anything, it’s gone down: last year featured the lowest WAR-weighted age since 1990, and while that number has ticked upward slightly in 2016, it’s still only level with 2013 and below 2014.3 The 1994 season, which saw the largest jump in HR/PA, also saw a 0.75-year increase in WAR-weighted age relative to 1992. At least so far, there has been no significant increase in older players’ value as there was in the steroid era.

Finally, we can look at parity between players, since one of the PED era’s defining features was the profound imbalance between the chemically-altered juggernauts like Barry Bonds and the average player. Not only did the rate of home runs increase across the league, but the top players in particular saw their dingers increase by an astounding degree. Record-breaking outlier seasons, like the famous 1998 single-season home run chase between Sammy Sosa and Mark McGwire, became almost common. Roger Maris’s all-time record of 61 homers, which had stood for 37 years, was exceeded six times from 1998-2001.

To measure this outlier effect in the steroid era, I calculated HR/PA for the top five home-run hitters in baseball each season.4 I’m not necessarily assuming that all of the top five were PED-users, but if there were artificially-enhanced players in the top five, they acted to drive the highest HR/PA rates even higher.

arthur-steroidera-2

The gap between the best home run hitters in the league and the average was never wider than in 1998, the midpoint of the PED era in terms of average hitter age, according to our definition above. Looking at the list of hitters that year, it’s not hard to tell why: Three of the top five hitters in home runs per plate appearance (minimum 300 PAs) were Mark McGwire, Sammy Sosa and Jose Canseco, all of whom later admitted or were alleged to have used PEDs. The second largest gap occurs in 2001, and features Bonds, McGwire and Sosa at the top of the list.

Meanwhile, contemporary HR rates for the best hitters have increased in lockstep with the MLB average. The difference between the top five current hitters’ HR/PA and the average in the last two years has been about the same as the norm across MLB’s history.5 And while the rate of HR/PA across the league has never been higher than today, the rate of the top five lags far behind that of the steroid era. MLB’s recent offensive explosion has seen the average hitter perform significantly better without creating a new wave of outliers.

That extraordinary evenness with which MLB’s latest HR surge has affected all players is maybe the best reason to discard the steroid explanation. The start of the steroid era was associated with a massive jump in home runs, but it affected some hitters more than others. Even most steroid users didn’t turn into musclebound hulks, but for those who did, the results were sometimes extraordinary. The recent offensive surge, on the other hand, has been both sudden and uniform across the league, resembling previous times in MLB’s history when the ball changed. Although historical comparisons like this cannot definitively prove that the ball is different now, they do suggest that whatever is causing the ball to fly farther is affecting all hitters equally.

Check out our latest MLB predictions.

Footnotes

  1. It’s worth noting that many of the largest jumps had dips immediately preceding them, suggesting either efforts by MLB to correct the ball’s performance or simple regression to the mean.
  2. Also, both 1994 and 2016 may have slightly elevated HR rates due to our data not including September, when home runs tend to be slightly less frequent. Even so, the addition of September would be expected to change the home run frequency by less than 1 percent, so any adjustment wouldn’t meaningfully affect the results.
  3. Probably some of the blame for the year-over-year increase lies with the slowing pace of MLB games, a factor which disproportionately aids older hitters.
  4. There’s nothing special about the top five, but I got similar results when I used the top 10, 15, 20 and 25 hitters each year.
  5. Specifically, the top five from 2015 ranked in the 55th percentile of all years, and 2016 ranks in the 25th percentile.

Rob Arthur is FiveThirtyEight’s baseball columnist and also writes about crime.

Comments