35
BASEBALL’S GREAT HITTING BARRAGE OF THE 1990S (AND BEYOND) REEXAMINED
By Benjamin G. Rader and Kenneth J. Winkle
In an article published in NINE in 2002, we examined what we called “Baseball’s Great Hitting Barrage of the 1990s.”[1] In addition to offering statistical support for the claim that there was an unusual amount of offensive productivity in the 1994 through 1999 seasons, we also considered explanations for why the hitting revolution had occurred. With regard to the latter, we questioned some of the popular theories for the offensive outburst—namely the “juiced-ball” hypothesis, the belief that the ballparks were cozier in the late 1990s than they had been earlier, and the role of league expansion in diluting the quality of pitching. But, at the same time, we lent support to the arguments that lighter bats, physically stronger hitters, and a new style of hitting (with the assistance of a smaller de facto strike zone) contributed significantly to the great hitting barrage of the late 1990s.
Now is an especially opportune time to reexamine and update our earlier findings. Not only do we presently enjoy the benefit of a longer historical perspective on the 1990s, but we are also able to extend our analysis from the 2000 through the 2006 seasons. Furthermore, recent disclosures of the widespread use of performance-enhancing drugs by the players and the implementation of a major-league drug testing program in 2003 make it possible to employ statistics to speculate more knowingly about the effects of drugs on the great offensive barrage. Equally if not more important to a reconsideration of the recent offensive outburst was the decision of the major leagues (beginning in 2001) to enlarge the de facto (that called by the umpires) strike zone and to try to impose on the umpires a more uniform strike zone.
We reach three major conclusions. First, that the great hitting barrage peaked during the 1999 and 2000 seasons. While remaining far above the two-divisional era (1969-1993 seasons) in offensive productivity, the 2001 through 2006 seasons fell below the peak achieved in 1999 and 2000. Based on batting averages, runs per game, home runs per game, and on-base percentage plus slugging percentage, we posit three eras of offense in recent baseball history: (1) the two-divisional era of low productivity (1969-1993), (2) the great offensive barrage (1994-2000 seasons), and (3) the new equilibrium (2001-2006 seasons). Second, while it is impossible to offer quantifiably direct evidence of the relationship between drug use and the offensive explosion, we conclude that player use of performance-enhancing drugs did contribute to the hitting barrage. When the threat of exposure and then drug testing increased, some measures of offensive productivity began to decline, though not approaching the depths of the 1969-1993 era. Third, it is possible to offer more quantifiably direct evidence of the relationship between the strike zone and the offensive explosion than it is the relationship between drugs and offense. We conclude that the size of the de facto (that called by the umpires) strike zone was an equal and perhaps even more important variable in explaining the coming of the hitting revolution as well as its modest decline after the 2000 season. When the major leagues decided to try to impose a more uniform strike zone on the umpires in the 2001 season, seasonal batting average and runs per game (but not home runs) fell, though not back to earlier levels.
THE CASE FOR THREE ERAS OF OFFENSIVE PRODUCTIVITY
[FIGURE 1, FIGURE 2, FIGURE 3, and FIGURE 4 inserted approximately here]
Figures 1-4 provide the basic statistical support for dividing the offensive productivity of recent baseball history into three eras. In Figure 1, notice the sharp ascent in major league batting averages for the 1993 and 1994 seasons; averages leveled off slightly for the 1995 through 1998 seasons, and then they lurched upward again for the 1999-2000 seasons. In 2001, batting averages began a sharp descent and remained below the 1994-2000 seasons until the 2006 season. Figures 2 and 4, which show runs per game and on base percentage plus slugging percentage (OPS) respectively, follow a similar season-to-season trajectory.
In Figure 3, notice that home runs per game also jumped sharply in 1994 and continued upward to their 1999-2000 record highs. In terms of fan involvement in the hitting revolution, nothing offered more convincing evidence of a new era in baseball history than the home run. In the four seasons of 1998 through 2001, the 60 home-run plateau was broken no fewer than six times. During every season from 1995 through 2002, at least one player hit 50 home runs, including four players in 1998. Home runs began a descent after the 2000 season though not as sharp as that of batting averages or runs scored. Indeed, the 2004 and 2006 seasons exceeded every earlier season in home runs per game except those of 1999, 2000, and 2001. Also 2006 marked the return of the 50-homer season when two players eclipsed this mark. In terms of home-run productivity, then, it is safe to say that the hitting revolution of the late 1990s continued essentially unabated through the 2006 season.
Offensive productivity in the AAA Pacific Coast and International Leagues follows a pattern similar to the major leagues. But, while the upturn in hitting in the majors began in the 1994 season, it did not fully reach the AAA minors until the 1996 season. As with the majors, hitting in the minor leagues has decreased slightly in the early twenty-first century. Indeed, batting averages and runs for the 2000 through 2006 seasons are below those for the 1990 through 1995 seasons. If measured in home runs, however, the minor league offensive barrage of the late-1990s continued through the 2006 season.
Table I. Three Eras of Minor League Offensive Productivity in the AAA Pacific Coast and International Leagues
______
Seasons Hits/100 at Bats Runs/100 at Bats Home Runs/100 at Bats
______
1990-95 27.4 14.6 2.2
1996-2000 27.5 15.3 2.9
2000-06 27.1 14.0 2.7
______
EXPLANATIONS
Given the above data, we sought to zero in on the determinants of the rise and the modest decline in baseball’s offensive productivity. In doing this, we found no new data or evidence to question our earlier rejection of a “juiced” ball[2] or a dilution in pitching talent as important determinants of the hitting barrage of the late 1990s. Furthermore, we have no reason to believe that either of these factors explain what we call the new equilibrium in the early twenty-first century. Neither do we have reason to question our conclusion that lighter bats may have contributed to the offensive revolution of the late 1990s. So, in this article, we limit our inquiry to a reexamination of how the use of performance-enhancing drugs, a changing strike zone, and the construction of new baseball parks may have contributed to the recent rise and modest decline in some important measures of baseball’s offensive productivity.
The Size of Ballparks. Sportswriters and fans alike frequently suggest or declare (without statistical proof) that change in the size of the ballparks has been a significant determinant of the increase in offensive productivity. In our 2002 article, we found that, while the franchises frequently did tinker with the distances from home plate to the outfield walls as well as with the height of fences, the average distance to the fences and their height remained virtually unchanged between the1990 and 1998 seasons. However, from the 1988 through the 2006 seasons, more than half of the franchises have built new parks. Following the example of Baltimore in 1992, nearly all the new stadia are “retroparks.” While in most instances the new parks put the fans closer to the action than their multipurpose predecessors (thereby encouraging the illusion that the fences are closer to home plate), we have found that the distances to the fences and the height of fences have in fact changed little between 1990 and 2006. Nor have these measurements changed much between 1998 and 2006. (See Table II). Employing regression analysis, statistical cognoscenti will be interested in learning that the slight increase in average fence distance between the 1998 and 2006 seasons (about 2 feet) “explains” about 6.5 percent of the slight decline in home-run productivity for this period.[3]
Table II. Average Distances to Fences of Major League Parks, 1990, 1998, and 2006
______
Season Left Foul Line Left Field Center Field Right Field Right Foul Line Average
______
1990 330 378 406 375 326 363
1998 331 376 406 374 328 363
2006 331 380 404 378 329 365
______
As we indicated in our 2002 article, the effects of Denver’s extraordinary high altitude on offensive productivity requires separate attention. According to Yale physicist, Robert Adair, batted balls travel 9 percent farther in Denver than a ball struck similarly at sea level. And we found in our 2002 article that “the addition of the Colorado franchise in Denver—first with Mile-High Stadium in 1993 and 1994 and then the new Coors Field in 1995—provided an 11 percent boost to the overall offensive outburst in the NL [to 1999], but the advantages to the hitters playing in Denver apparently arose from elevation rather than the size of the park.”[4] Indeed, during the late 1990s offensive totals soared in Denver. In the 1999 season, the Rockies and their opponents struck 303 home runs in Coors Field, the most ever hit in one season at one venue with a single tenant. The average 1999 score in Coors Field was 8-7. With the outburst of hitting in the new park, sportswriters began to label it “Coors Light Field.” When the Cincinnati Reds defeated the Rockies by a score of 24-12 on May 19, 1999, Reds first baseman Sean Casey exclaimed that “It was like a beer league [slow-pitch softball] game. We should have walked away with the keg.”[5]
With the 2002 season, Denver officials began an experiment (initially in secret) designed to retard offensive productivity in Coors Field. Based on the fact that Denver’s prevailing low humidity dried out the balls, thus making them more resilient than the balls used in the low-altitude parks, the Rockies proceeded to store game balls in a humidity- and temperature-controlled room—a “humidor”—built near the Coors Field clubhouses.[6] Statistics offer strong support for the conclusion that the presumably damper and slightly heavier balls stored in the humidor reduced offensive productivity in Coors Field (See Table 3). Indeed, with the use of the humidor, runs fell by more than 2 and home runs by 1 per hundred times at bat. The sharp decline in offensive productivity at Coors Field beginning with the 2002 season “explains” 14.9 percent of the overall decline in home runs, 7.8 percent of the runs, and 6.5 percent of the hits in all of major league baseball for the 2002 through 2006 seasons. In the parlance of regression analysis, this is a “significant” relationship. Impressed with the possibility that temperature and humidity control could provide a ball in every park that would behave uniformly, baseball’s general managers discussed in 2006 the possibility of installing humidors in all the parks.[7]
TABLE 3. Offensive Output at Coors Field (Denver) for both the home team and visitors before and after the installation of the Humidor
______
Seasons Batting Avg. Runs/100 at bats HRs/100 at bats
______
Before the humidor (1995-2001) .310 18.5 4.5
With the humidor (2002-2006) .292 16.3 3.5
______
Because of Rockies general manager Dan O’Dowd’s change in strategy beginning with the 2000 season, there are those who argue that the decline in offensive productivity in Coors Field was due to more than the humidior. The Rockies had led the league in runs scored and home runs for three consecutive seasons (1997-1999)-- before the arrival of O’Dowd—but the team had played above .500 only once (1997) in those seasons. To turn around the team’s fortunes, O’Dowd quickly set out to alter the character of the ball club. He sought to improve its speed, defense, and pitching even if it meant some sacrifice of hitting prowess. Entering the player market with an enthusiasm unequalled in the past fourteen years of big-league history, before the 2000 season commenced, he jettisoned twelve players while acquiring fifteen replacements. In subsequent years, he continued to be an unusually active trader; and, in 2001, he sought to improve the Rockies’ fortunes by signing two veteran starting pitchers, Mike Hampton and Denny Neagle.[8]
The seasons of 2000 and 2001 (before the installation of the humidor) permit the most direct evidence of how O’Dowd’s emphasis on pitching and fielding affected offense in Coor’s field. In the aggregate, the new ploy did not work: except for batting averages, the offenses of both the Rockies and visitors in the 2000 and 2001 seasons were above their averages for the 1995-1999 seasons! (See Table 4). Neither has the strategy been effective in improving the Rockies’ performance relative to the visitors since the installation of the humidor. (See Tables 3 & 5).
TABLE 4. Offensive Output, Colorado Rockies and Opponents at Coors Field before and after O’Dowd Became General Manager of the Rockies, but before the Installation of the Humidor
Batting average Runs/100 at bats Hrs/100/ at bats
Rockies Opponents Rockies Opponents Rockies Opponents
______
1995-1999 .326 .295 19.8 16.8 4.7 3.9
2000-2001 .332 .288 20.6 17.7 5.4 4.6
TABLE 5. Offensive Output, Rockies and Opponents at Coors Field before and after O’Dowd Became General Manager of the Rockies
______
Batting Average Runs/100 at bats Hrs/100 at bats
Rockies Opponents Rockies Opponents Rockies Opponents
______