Hit Batters and Home Runs

Yesterday’s New York Times has an article by Stuart Miller on the rise and decline of hit batters in baseball titled Plunking Parallel: Steroid Use and Hit Batsmen.

The so-called steroids era produced plenty of pumped-up statistics. Everyone talked about the home runs, but few noticed the soaring number of batters hit by pitches.

I did, and I even discussed this in The New York Times. So, I was curious about what’s new on the subject. The article looks at the recent rise in home runs and notices the co-movement of home runs and hit batters. It is true, the variables appear to be somewhat correlated, which could mean that one (home runs) is causing the other (hit batters) to move with it. In this case, Miller posits that one explanation for the recent moderate drop in HBPs is that decline of steroid use by players has caused power to decline.

Most significant was the artificially enhanced power surge. To prevent batters from extending their arms and launching moon shots, pitchers came inside with a vengeance.

“You were just trying to survive,” said Orel Hershiser, who hit more than 11 batters a year from 1996 to 2000, more than double his average from 1984 to 1988.

Hershiser, now an ESPN analyst, said pitchers traditionally worked outside because mistakes inside usually equaled a double or homer. But once players became so strong that “the outside mistake was of equal value” because beefed-up hitters could flick it over the fence, “people started pitching to both sides of the plate equally.”

Although many beanballs were not intentional, he said, “I’d think, I am really mad that this guy is so strong, so I am just going to bury the ball inside.” With guys crowding the plate — and so armored they would stay in longer — chances increased for an H.B.P.

Since 2006, however, the trend has reversed — with H.B.P. per team dropping 13 percent, to 59, then 56, then 53 last year. The shift mirrors the supposed decline in steroid use and home runs.

HBPs may be down since 2006, but they are still 40-percent above the 1992 HBP rate. The graph belows shows the home run and hbp-allowed rates (per batter faced) in the modern era.

HR and HBP

The first vertical dashed-line marks 1993, the year when home runs jumped in baseball; and I have argued that this dramatic rise is evidence that steroids are not the main cause of baseball’s present high-offense era. Players just didn’t get together and start using all at once.

Drug testing with suspensions for first offenses began in 2005 (also represented by a vertical line on the graph). Yet, the decline in home runs and hit batters peaked around the turn of the century. Home runs and hit batters were declining before testing came onto the scene. Furthermore, after testing, the observed decline in home runs is difficult to distinguish from random fluctuations. And even we grant that testing has reduced power it is very clear that the 1990s-2000s power surge hasn’t come close to returning its previous level.

Home runs and hit batters do move together, but it’s difficult to know if one causes the other or if a third factor influences both variables while they move independently of each other. (Indeed, I have spent much of the morning boning up on cointegration, which has left me with a terrible headache.) Given simultaneous jump in home runs and hit batters in 1993, the peak several years prior to testing, and the minimal decline after testing, I think there is another factor at work here explaining their movements. The article does mention a few of them, but only steroids makes it into the title, and I find this disappointing. A change in the ball, expansion, or a widening strike zone are all potential explanations. If we want to understand how hit batter and home run rates vary, we need to explain the big change in the early 1990s before analyzing the smaller and more recent fluctuations.

3 Responses “Hit Batters and Home Runs”

  1. Shek says:

    Your 2006 JSE article with Doug Drinen, while not covering the exact same time period as the above article, does look at the incidence of hit batters within the heart of the “Steroid Era” and found that home runs decreased the probability of a hit batter. While the article explored specifically the DH contribution to HBP, the results are transferrable to the question at hand, although the causality for the counterintuitive result probably lies with the double warning rule. Further complicating matters would be how QuesTec and PitchTrax has impacted HBP on the margin (in addition to how the shock of expansion has worn off).

  2. JC says:

    Agreed, good point Shek. I think there likely is a causal relationship between HR and HBP, but there is something bigger and more important going on here.

  3. Millsy says:

    Just eyeballing it, it seems like while they’re moving together, HBP is lagging behind HR from about 1945 to the most recent jump(s) just before 1990. Might be easier to see with some smoothing here.