1. LARGER POOLS AND BETTER TRAINING. In 1900,76 million people inhabited the United States, and only white men could play major league baseball. Our population has since ballooned to 249 million people (1990 census), and men of all colors and nations are welcome. Training and coaching were absent to slapdash in the past, but represent a massive industry today. Players follow rigorous and carefully calculated programs for working out (even, if not especially, during the off season, when their predecessors mostly drank beer and gained weight); they no longer risk careers and records by playing hurt. (Joe DiMaggio once told me that he was batting 0.413 with two weeks to go in the 1939 season. He caught a serious cold, which clouded his left [leading] eye, and he could not adequately see incoming pitches. The Yanks had already clinched the pennant. Any modern counterpart would sit on the bench and preserve his record; DiMaggio played to the last game and fell to 0.381. his highest seasonal average, but below the grand plateau.) No one—neither the players nor the owners—can afford to take risks and fool around today, not with star salaries in the multiple millions for careers that last but a few years at peak value. What possible argument could convince us that a smaller and more restricted pool of indifferently trained men might supply better hitters than our modern massive industry with its maximal monetary rewards: I'll bet on the larger pool, recruitment of men of all races, and better, more careful training any day.
2. SIZE. I don't want to fall into the silly mythology of "bigger is better" (okay for a few things, like brains in the evolution of most mammalian lineages, but irrelevant for many items, like penises and automobiles). Still, _ceteris paribus_ as the Romans said (all other things being equal), larger people tend to be stronger (and I say this as a short man who loved to watch Phil Rizzuto and Fred Patek). If height and weight of ballplayers have augmented through time, then (however roughly) bodily prowess should be increasing.
Pete Palmer, sabermetrician extraordinaire and editor, along with John Thorn, of _Total Baseball_, the best (and fattest) general reference book of baseball stats, sent me his chart (reproduced here as Table 1) of mean heights and weights for pitchers and batters averaged by decades. Note the remarkably steady increase through time. I cannot believe that the larger players of today are worse than their smaller counterparts of decades past.
TABLE 1
DECADAL AVERAGES FOR HEIGHTS AND WEIGHTS OF MAJOR LEAGUE BASEBALL PLAYERS
BATTERS PITCHERS
Height Weight Height Weight
(inches) (pounds)
1870s 69.1 163.7 69.1 161.1
1880s 69.6 171.6 70.2 172.7
1890s 69.8 172.1 70.6 174.1
1900s 69.9 172.6 71.5 180.7
1910s 70.3 170.5 72.1 180.7
1920s 70.4 171.2 72.0 179.8
1930s 71.1 176.8 72.6 184.8
1940s 71.4 180.3 73.0 186.5
1950s 72.0 183.0 73.1 186.1
1960s 72.2 182.7 73.6 189.3
1970s 72.3 182.3 74.1 191.0
1980s 72.5 182.9 74.5 192.2
3. RECORDS IN OTHER SPORTS. All major baseball records are relative—that is, they assess performance against other players in an adversarial role—not absolute as measured by personal achievement, and counted, weighed, or timed by a stopwatch. A 0.400 batting average records degree of relative success against pitchers, whereas a four-minute mile, a nineteen-foot pole vault, or a 250-pound lift is unvarnished you against an unchanging outer world.
Improvements in relative records are ambiguous in permitting several possible (and some diametrically opposed) interpretations: rising batting averages might mean that hitting has improved, but the same increase might also signify that batting has gotten worse while pitching has deteriorated even more sharply (leading to relative advantages for hitters as their absolute skills eroded).
Absolute records, however, have clearer meaning. If leading sprinters are running quicker and vaulters jumping higher . . . well, then they are performing their art better. What else can we say? The breaking of records doesn't tell us why modern athletes are doing better—and a range of diverse reasons might be cited, from better training, better understanding of human physiology, new techniques (the Beamon back flop), to new equipment (fiberglass poles and the immediate, dramatic rise in record heights for pole vaulting)—but I don't think that we can deny the fact of improvement.
Therefore, since the relative records of baseball must be ambiguous in their causes, we should study the absolute records of related sports. If most absolute records have been improving, then shouldn't we assume that athletic prowess has risen in baseball as well: Wouldn't we be denying a general pattern and creating an implausible, ad hoc theory if we attributed the extinction of 0.400 hitting to a decline in batting skills? Shouldn't we be searching for a theory that can interpret the death of 0.400 hitting as a consequence of generally superior athleticism—thus making this most interesting and widely discussed trend in the history of baseball statistics consistent with the pattern and history of almost every other sport?
I don't want to worry a well-understood subject to death, and I don't want to bore you with endless documentation of well-known phenomena. Surely all sports fans recognize the pervasive pattern of improvement in absolute records through time. The first modern Olympic marathon champ, Spiridon Loues, took almost three hours in 1896; more recent winners are nearly down to two. The allure of the four-minute mile challenged runners for decades, while Paavo Nurmi's enticing 4.01 held from 1941 until Roger Bannister's great moment on May 6, 1954. Now, most of the best runners routinely break four minutes nearly every time. By 1972, for the 100-meter freestyle, and by 1964 for the 400 meters, the best women swimmers eclipsed the Olympic records of the 1920s and 1930s, set by the two great Tarzans (both played the role in movies) Buster Crabbe and Johnny Weissmuller. I will let one chart stand for the generality, based on data closest to hand for the most famous of local races at my workplace—the Boston Marathon (see Figure 12). The general pattern is clear, and the few anomalies record changes in distance (the "standard" 26 miles 385 yards has prevailed in most years, but early winners, from 1897 to 1923, ran only 24 miles 1,232 yards for their longer times; with a rise to 26 miles 209 yards from 1924 to 1926; the standard distance from 1927 to 1952; and a shortened 25 miles 958 yards from 1953 to 1956, until reestablishment the standard distance in 1957).
For almost every sport, the improvement in absolute records follows a definite pattern, with presumed causes central to my developing argument about 0.400 hitting. Improvement does not follow a linear path of constant rate. Rather, times and records fall more rapidly early in the sequence and then slow markedly, sometimes reaching a plateau of no further advance (or of minutest measurable increments from old records). In other words, athletes eventually encounter some kind of barrier to future progress, and records stabilize (or at least slow markedly in their frequency and amount of improvement). Statisticians call such a barrier an asymptote; vernacular language might speak of a limit. In the terminology of this book, athletes reach a "right wall" that stymies future improvement.
Since we are considering the world's best performers in these calculations, the probable reasons for such limits or walls should be readily apparent. After all, bodies are physical devices, subject to constraints upon performance set by size, physiology, and the mechanics of muscles and joints. No one will argue that curves of improvement can be extrapolated forever—or else runners would eventually complete the mile in nothing flat (and, finally, in negative time), and pole vaulters would truly match a gentleman of legend and leap tall buildings in a single bound.
We can best test the proposition that physical limits (or right walls) cause the slowing and plateauing of improvements by comparing curves for athletes operating near the extremes in human capacity with performers who probably retain much room for further advance. What conditions might place people far from the right wall, and therefore endow them with great scope for improvement? Consider some potential examples: new sports where athletes have
not yet figured out optimal procedures; new categories of people recently admitted to old sports; records for amateur play. As an example, the Boston Marathon was opened to women only in 1972. Note how much more rapidly women have improved than men from their beginning to the present (Figure 13).
We may generalize this principle by setting up a hierarchy of decreasing improvement (also a ranking of increasing worth in the value systems of some rather old-fashioned and well-heeled folks): women, men, and horses. Winning rimes for major horse races have improved, but ever so slightly over long intervals. For example, between 1840 and 1980, thoroughbred horses in the three great English races of St. Leger, Oaks, and the Derby have shaved twelve, twenty, and eighteen seconds off record times, for a minuscule gain of 0.4 to 0.8 percent per generation (Eckhardt et al., 1988). These gains are tiny even when compared with the other great arena of breeding in domesticated animals: improvement of livestock, where gains of 1 to 3 percent per year are often achieved for features of economic importance.
This limited improvement makes perfect and predictable sense. Thoroughbreds have been rigorously raised from a limited stock for more than two hundred years. Stakes could not be higher, as the slightest improvement may be worth millions. More effort has gone into betterment of this breed than into almost any other biological endeavor of economic importance to humans. We might therefore suppose that the best thoroughbreds have long resided at the genetic right wall for the breed, and that future improvements will be negligible to slow. But since (thank God) we have not yet reached brave new world, we do not breed humans for optimized physical performance, and records for people should therefore show more flexibility—for we have no purposeful purebreds at our right walls.
In most popular and established men's events, we note the pattern of rapid initial improvement followed by flattening of the curve.[4] Exceptions may he found in such events as the marathon, where length and complexity provide great "play" for experimenting with new strategies, and where recent surges in popularity have brought large increases in prestige and participation. (Note that the curve of improvement for the Boston Marathon has remained virtually linear for men. and did not slow before 1990—though the pattern may now be shifting into the usual mode as the world's best runners now compete and improvements begin to abate.)
[4. All bets are off when fundamentally new equipment or procedures enter the field, as in the fiberglass pole, or (God forbid) the aluminum bat. which (we may hope and pray) will never darken the doorstep of major league dugouts. Such innovations will produce sudden blips in curves of improvement. In fact, such innovations are usually better treated statistically as the beginning points of new curves.]
Many commentators have noted that most women's records are both falling faster than men's for the same event, and are not yet flattening, but maintaining a linear pace of improvement. Interestingly (see Whipp and Ward, 1992), most men's running events (200 to 10,000 meters) have improved in the same range of rates regardless of the event's total distance—5.69 to 7.57 meters per minute improvement per decade. (Improvement in the marathon has been greater, at 9.18 meters per minute per decade, thus supporting my claim that this event remains "immature" and still in the category of potentially linear improvement—that is, not near the right wall.) But for women in the same events, rates of improvement run from 14.04 to 17.86 minutes per meter per decade (with a whopping 37.75 meters for the marathon).
These findings have led to all manner of speculation, some rather silly. For example, Whipp and Ward (1992) just extrapolate their curves and then defend the conclusion that women will eventually outrun men in most events, and rather soon for some. (The extrapolated curves for the marathon, for example, cross in 1998 when women should beat men by this argument.)
But extrapolation is a dangerous, generally invalid, and often foolish game. After all, as I said before, extrapolate the linear curve far enough and all distances will be run in aero and then in negative time. (False extrapolation also produces the irresponsible figures often cited for growth of human populations—in a few centuries, for example, humans will form a solid mass equal to the volume of the earth and no escape into outer space will be possible because the rate of increase will cause the diameter of this human sphere to grow at greater than the speed of light, which, as Einstein taught us, sets an upper bound upon rapidity of motion.) Clearly we will never run in negative time, nor will our sphere of solid humanity expand at light speed. Limits or right walls will be reached, and rates of increase will first slow and eventually stop.
Women may outclass men in certain events like ultra-long-distance swimming, where buoyancy and fat distribution favor women's physiques and endurances over men's (women already hold the absolute record for the English Channel and Catalina Island swims). The marathon may also be a possibility. But I doubt that women will ever capture either the 100-meter dash or the heavyweight lifting records. (Many women will always beat most men in any particular event—most women can beat me in virtually anything physical. But remember that we are talking of world records among the very best performers—and here the biomechanics of different construction will play a determining role.)
The basic reason for more rapid gains (and less curve flattening) in women's events seems clear. Sexism is the culprit, and happy reversals of these older injustices the reward. Most of these events have been opened to women only recently. Women have been brought into the world of professionalism, intense training, and stiff competition only in the last few years. Women, not so long ago (and still now for so many), were socialized to regard athletic performance as debarred to their gender—and many of the great women performers of the past, Babe Didrikson in particular, suffered the onus of wide dismissal as overly masculine. In other words, most women's curves are now near the beginning of the sequence—in the early stages of rapid and linear improvement. These curves will flatten as women reach their own right walls—and only then will we know true equality of opportunity. Until then, the steep and linear improvement curves of women's sports stand as a testimony to our past and present inequities.
9
0.400 Hitting Dies as the Right Tail Shrinks
Granting the foregoing argument that hitting must be improving in some absolute sense as the best athletes first rush, and then creep, toward the right wall of biomechanical limits on human performance, only one traditional explanation remains unrefined for viewing the extinction of 0.400 hitting as the deterioration of something at bat—the possibility that, while hitting has improved, other opposing activities (pitching and fielding) have gotten better, even faster, leading to a _relative_ decline in batting performance.
This last holdout of traditionalism fails the simplest and most obvious test of possible validity. If pitching and fielding have slowly won an upper hand over hitting, we should be able to measure this effect as a general decline in batting averages through the twentieth-century history of baseball. It mean batting averages have fallen with time, as pitching and fielding assert increasing domination, then the best hitters (the 0.400 men of yore) get dragged down along with the masses—that is, if the mean batting average were once 0.280, then a best of over 0.400 makes sense as an upper bound, but if the mean has now fallen to, say, 0.230, then 0.4UO might stand too far from this declining mean for even the best to reach.
This entirely sensible explanation fails because, in fact, the mean batting average for everyday players has been rock-stable throughout our century (with interesting exceptions, discussed later, that prove the rule). Table 2 (page 102) presents decadal mean batting averages for all regular players in both leagues during the twentieth century. (I included only those players who averaged more than two at-bats per game for the entire season, thus eliminating weak-hitting pitchers and second-stringers hired for their skills in fielding or running.)[5] The mean batting average began at about 0.260, and has remained there throughout our century. (The sustained, though temporary, rise in the 1920s and 1930s presents a single and sensible excepti
on, for reasons soon to come, but cannot provide an explanation for the subsequent decline of 0.400 hitting for two reasons: first, the greatest age of 0.400 hitting occurred before then, while averages stood at their usual level; second, not a soul hit over 0.400 throughout the 1930s, despite the high league means—I include Bill Terry's 0.401 of 1930 itself in the 1920s calculation.) Thus our paradox only deepens: 0.400 hitting disappeared in the face of preserved constancy in average performance. Why should the best be trimmed, while ordinary Joes continue to perform as ever before? We must conclude that the extinction of 0.400 hitting does not reflect a general decline in batting prowess, either absolute or relative.
[5. The recent disparity between, the two leagues records, in large part, the introduction of the "designated hitter" to the American League alone—a permanent "pinch hitter" for the pitcher. His substitution for the pitcher doesn't affect the decadal average per se, because I don't include pitchers in this calculation. But the designated hitter still provokes a small general rise in the American League mean by introducing another good bat into the lineup, whereas the National League retains more relatively poor hitters in the bottom part of the order. Nonetheless, I remain an adamant opponent of the DH rule—the one vital subject in our culture that permits no middle ground. You gotta either love it or hate it!]