April 28, 2009

Prospectus Hit and Run

On the Rise

by Jay Jaffe

Printer-friendly

The 2009 season is just three weeks old, and already one of the dominant themes of discussion is rising home-run rates. As with many subjects this spring, the new Yankee Stadium has done much to drive the conversation and skew the numbers, with 26 home runs flying out of the yard during the Bronx Bombers' initial six-game homestand, a result that even piqued the interest of idle meteorologists. Elsewhere, anecdotal observations of balls not particularly well struck flying over fences abound, and measure-meister Greg Rybarczyk of Hit Tracker has noted that distances are up. In a staff piece cobbled together by request from our partners at ESPN last Friday, Marc Normandin pointed out that April home-run rates were up from 1.78 per game last year to 2.15 this year. As colleague Will Carroll often asks when passing along data for examination by the more mathematically inclined, "Is this anything?"

Color me unimpressed, at least initially. While home runs are being hit with more frequency than last year, the rate at which balls are flying out of the park is well within the range we've seen since 1994, when home-run rates began climbing. Through play on Saturday, April 25:

Year  HR/PA   HR/TmG
2009   2.79%   1.082
2008   2.60%   1.005
2007   2.63%   1.020
2006   2.86%   1.109
2005   2.69%   1.032
2004   2.89%   1.123
2003   2.78%   1.071
2002   2.71%   1.043
2001   2.92%   1.124
2000   2.99%   1.172
1999   2.91%   1.138
1998   2.69%   1.041
1997   2.64%   1.024
1996   2.80%   1.094
1995   2.60%   1.012
1994   2.66%   1.033

I've expressed the per-game rate as per team per game, so the numbers are half of what Marc reported. That's not only a matter of personal preference, it jibes with our tendency to talk of run-scoring environments in terms of a single team or, say, to normalize offensive statistics to an average of 4.5 runs per game. I've also reported the rate per plate appearance, which some people tend to prefer, but for simplicity's sake, from here on I'll just stick with my per team per game convention.

Excluding the strike-affected 1994-1995 years, the 2009 season ranks seventh out of 14 seasons no matter which rate you use. Within that context, it's a run-of-the mill post-strike season. What's throwing observers is that 2008 featured the lowest home-run rate of that period, and 2007 the second-lowest. The 7.7 percent increase over the previous year, were it to hold, would be the largest climb since 1998-1999 (9.3 percent), just edging out the 2005-2006 increase (7.4 percent).

Of course, we're still dealing with a relatively small sample size here—10.6 percent of the schedule, to be exact—as we haven't even finished the April slate in a season where Opening Day arrived late because of the World Baseball Classic. The question is whether a change observed in the cruelest month will continue to manifest itself over the course of the year. All signs point to yes:

Year    April   Change   Season   Change
1996    1.150     N/A     1.094     N/A
1997    0.944   -17.9%    1.024    -6.4%
1998    0.976     3.4%    1.041     1.7%
1999    1.143    17.1%    1.138     9.3%
2000    1.281    12.1%    1.172     3.0%
2001    1.168   - 8.8%    1.124    -4.1%
2002    0.953   -18.4%    1.043    -7.2%
2003    1.047     9.9%    1.071     2.7%
2004    1.087     3.8%    1.123     4.9%
2005    0.947   -12.9%    1.032    -8.1%
2006    1.154    21.9%    1.109     7.5%
2007    0.920   -20.3%    1.020    -8.0%
2008    0.896   - 2.6%    1.005    -1.5%
2009    1.082    20.8%    1.082     7.7%

Since the post-strike 1995 season didn't start until April 25, we're confined to using 1996 as a cutoff, but the effect is clear: the small samples of April (and March) games can produce swings on the order of 20 percent, and while the magnitudes of such year-to-year changes aren't sustained over the course of the season, an increase or decrease in April home runs has always portended an annual change in the same direction without fail during this era. It's a nearly bulletproof assertion to say that we'll see more home runs hit in 2009 than 2008.

The next question is why. For the better part of this millennium, the conventional wisdom has held that any variation in home-run rates must be connected to the impact of steroid usage; indeed, the entirety of the post-strike period outlined above has been stamped "The Steroid Era." The way this reasoning goes, if home-run rates are on the rise, it's because more players are juicing, or at least more are beating baseball's testing system with human growth hormone or undetectable designer-drug successors to "the Clear" and "the Cream." Similarly, if home run-rates are falling, it's because the testing seems to be working and the game has cleaned up. Particularly with annual rates falling in three out of four years since the beginning of 2005, as Major League Baseball has gained the power to suspend players for testing positive for steroids, it's an easy explanation to invoke.

Except, of course, that this period has also been marked by some of the most drastic changes in the game since the end of the Deadball Era, featuring not only expansion but also interleague play, the redefined strike zone, the three-division format with the unbalanced schedule, and the Wild Card, and the biting with metal teeth, and... wait, scratch that last part. I first studied these factors for a chapter in Will Carroll's The Juice, and if this story sounds familiar, it's because I've covered this ground before. A building boom added a wave of new stadiums—we're up to 25 since 1989, including two temporary facilities that came and went (Mile High and RFK), as well as the two recently opened ballparks in New York—and while the fence distances generally haven't been shorter, intrepid reporters and analysts have revealed instances where the marked distances are less than advertised, an issue that affects even the new Yankee Stadium. Expansion has added high-altitude venues in Colorado and Arizona, and it's possible that the configurations of other new parks, particularly given the HOK-designed tendency toward more field-level seats and shallower upper-deck inclines (as in the shiny Bronx bauble, again) may have produced unanticipated wind patterns that have increased home-run rates even as the outfield dimensions have increased.

Additionally, there is mounting evidence that the ball itself has changed over the years via the increasing size of the "pill" at its center, the introduction of synthetic yarns used to wind the ball, and the incorporation of a synthetic ring unaccounted for in Major League Baseball's official specifications. The oft-cited University of Massachusetts-Amherst study from 2000 commissioned by Rawlings and MLB found that the manufacturing process may produce two balls that fall within the extremes of official tolerances, but differ in flight distance by 49.1 feet, despite being struck under the exact same conditions. One way or another, the ball may have been juiced.

Within the limited sample of this season's games, balls and/or ballparks may be driving what Hit Tracker reports as an increase in both True Distance (actual home-run distance) and Standard Distance (normalized to remove the influence of wind, temperature, and altitude. On both counts, those measures are rebounding from two years of decline:

Year  Avg True Dis   Change    Avg St Dis   Change
2006     398.7        n/a         392.8      n/a
2007     396.8       -0.48%       392.7     -0.03%
2008     396.7       -0.03%       391.4     -0.33%
2009     398.1       +0.35%       396.1     +1.20%

On a percentage basis, those changes might seem trifling, but in a game of inches, they can have a considerable impact. According to Rybarczyk, the drop of eight percent in home runs from 2006 to 2007 resulted from a three-foot reduction in average fly-ball distance (not quite the same as home-run distances). Roughly speaking, Rybarczyk says that every foot of distance added increases homers by about three percent.

While we're unlikely to get smoking gun-level proof regarding any changes in the ball, and while the early flurry of home runs and increased distances may well come out in the wash of a larger sample, we could be witnessing a shift, namely in how this subject is discussed. The absence of any new positive tests or contemporary steroid-related controversies (Alex Rodriguez's admission is soooo 2003) on which one can pin the home-run increase is striking. The ongoing discussion about the impact of the two New York parks, and ballparks in general, plus the entry of Rybarczyk's distance data into the mainstream—he was on ESPN's Baseball Tonight last week (as was our own Nate Silver)—suggests that observers are looking for explanations that refreshingly enough don't involve speculation over which players are sticking needles into their butts. Beyond any fluctuation in home-run rates, that may be the most impressive outcome in this whole situation. While it's premature to declare this the Post-Steroid Era, consider this one more signpost along the way.