The Sports Gene: Inside the Science of Extraordinary Athletic Performance

by David Epstein

  In 1983, the NBA struck a groundbreaking collective bargaining agreement with players that made the athletes partners in the league, entitled to money from licensing agreements, ticket sales, and television contracts. The following year, a rookie Michael Jordan signed a comparably trailblazing contract with Nike that gave him royalties from the sales of sneakers bearing his name.

  Suddenly, the earning potential of professional basketball players shot through the arena roof, and pretty much anyone who could play in the NBA wanted to. At the same time, NBA teams began scouring the globe for giants. In just three years following the new labor agreement, the proportion of seven-footers in the NBA more than doubled, reaching 11 percent, where it has essentially remained ever since. “What it means is that basically everyone in the world who is seven foot tall and can play basketball is part of the game,” Olds says. “We’ve kind of reached a population limit.”

  Reaching it required increasing globalization of the game. The average height of American players in the NBA is about 6'6½", while the average height of foreign players is nearly 6'9". A great many of the foreign players in the NBA are there, it seems, because teams ran low on sufficiently tall players at home. Perhaps it’s no surprise, then, that the non-U.S. countries with stable representation in the NBA—Croatia, Serbia, Lithuania—are among the tallest in the world. Because height is a “normally distributed” human trait (i.e., a bell curve), a tiny difference in the average height in a country means a big difference in the number of people at the far extremes, like seven-footers.

  In terms of freakish height, the Women’s National Basketball Association lags far behind the men’s league. The average height of a WNBA player is between 5'11" and six feet, not as relatively tall compared with an average woman as an NBA player with an average man. The average WNBA player is only about 10 percent taller than the average American woman, compared with the average NBA player who is closer to 15 percent taller than the average man.

  Perhaps it will just take time for more tall women to gravitate to the game. Or perhaps it will take a stronger winner-take-all market. WNBA players make just tens of thousands of dollars per year, while the average NBA player rakes in more than $5 million a year. It is easy to see why many women with the athletic gift of height might be inclined toward other sports that hold more lucrative opportunities for them, like tennis. As rackets have grown lighter and serves more important in tennis, players have gotten taller. At the time of this writing, the top three female tennis players in the world have an average height of 5'11⅔", nearly identical to the average height in the WNBA.

  None of this is to say that shorter men and women can’t succeed in basketball. NBA players like Muggsy Bogues (5'3"), Nate Robinson (a shade under 5'8"), and Spud Webb (5'7", with thick socks) all thrived in the land of giants. But not without abilities that compensated for their stature. Robinson and Webb, two of the shortest players in NBA history, both won the Slam Dunk Contest. Bogues claimed an astonishing forty-four-inch vertical leap, but his tiny hands made it difficult to palm a basketball, so he was content to dunk volleyballs in practice.

  Short people generally don’t make the NBA unless they have extraordinarily anomalous jumping ability. Not necessarily like Bogues, Robinson, and Webb, but consider the all-time grand total number of men drafted into the NBA who were unable to get high enough to grab the rim when tested at a predraft combine: zero. But there’s something else that helps wee NBA ballers play large.

  Leonardo da Vinci’s Vitruvian Man has an arm span equal to his height. So do I. So, probably, do you, or very nearly so. Nate Robinson, on the other hand, is 5'7¾" and his arm span is 6'1". He is, effectively, not as short as he is. Actually, almost none of the players in the NBA are as short as they seem, including the ridiculously tall ones.

  The average arm-span-to-height ratio of an NBA player is 1.063. (For medical context, a ratio of greater than 1.05 is one of the traditional diagnostic criteria for Marfan syndrome, the disorder of the body’s connective tissues that results in elongated limbs.) An average-height NBA player, one who is about 6'7", has a wingspan of seven feet. To fit the Vitruvian NBA player, Leonardo would have needed a rectangle and an ellipse, not his tidy square and circle.

  NBA players who are labeled as “undersized” for the position they play based on stature generally have the extra arm span to make up for it. Elton Brand, the first pick of the 1999 NBA Draft, at 6'8¼" is unremarkable for a power forward. But Brand is actually a giant among power forwards if you consider his 7'5½" of reach. John Wall, the point guard who was the first pick in the 2010 draft, is only 6'2¾" with his shoes off, but has 6'9¼" worth of reach. When the Miami Heat assembled its ballyhooed Big Three—Chris Bosh, LeBron James, and Dwyane Wade—before the 2010–11 season, the team was enlisting 19'9¼" of height, but 21'2½" of wingspan. And it’s no coincidence.

  Based on statistics for players who were on NBA rosters at the beginning of the 2010–11 season, a player’s wingspan influences a number of key statistics. An NBA general manager who wants to increase his team’s blocked shots would be better off signing a player with an extra inch of arm than an inch of height. The New Orleans Pelicans’ Anthony Davis, the shot-swatting first pick of the 2012 draft, is 6'9¼" with a 7'5½" wingspan. A player with Davis’s build will be predicted to get ten more blocks per season than a 7'1" giant who plays an equal number of minutes but has arms that match his height. If the GM wanted offensive rebounds, he would do equally well to sign a player with an extra inch of reach as an extra inch of height. And while height is a slightly better predictor of defensive rebounds than is wingspan, both are important and together account for half of an NBA player’s defensive boards, without even considering characteristics like jumping ability, weight, position, or general rebounding skill.

  Stat-savvy general managers have no doubt noticed. Daryl Morey, the MIT-educated GM of the Houston Rockets, renowned for his Moneyball approach to basketball, has drafted several of the most superficially undersized players in the NBA. (Morey would not comment on whether the Rockets strategically targeted high wingspan-to-height-ratio players in the draft.) Three seasons ago, the Rockets used the shortest starting center in NBA history, Chuck Hayes, who is just 6'5½". Fortunately, his arms are 6'10".

  The bottom line is that not only are NBA players outlandishly tall, they are also preposterously long, even relative to their stature. And when an NBA player does not have the height required to fit into his slot in the athletic body types universe, he nearly always has the arm span to make up for it. In the post–Big Bang of body types era, whether with height or reach, almost no player makes the NBA without a functional size that is typical for his position and often on the fringe of humanity. Only two players from a 2010–11 NBA roster with available official measurements have arms shorter than their height. One is J. J. Redick, the Milwaukee Bucks guard who is 6'4" with a 6'3¼" arm span, downright Tyrannosaurus rex-ian in the NBA.* The other is now-retired Rockets center Yao Ming. But at a height just over 7'5", Yao, whose gargantuan parents were brought together for breeding purposes by the Chinese basketball federation, fit into his niche just fine.

  •

  Repeatedly, studies of families and twins find the heritability of height to be about 80 percent. That means that 80 percent of the difference in height between people in the group that is being studied is attributable to genetics, and around 20 percent to the environment. (In nonindustrialized societies, the heritability of height is lower, as many citizens, like plants in poor soil, are prevented by nutritional deficiencies or infections from reaching their genetic height potential.) So if the tallest 5 percent of citizens in a given population are a foot taller than the shortest 5 percent, genetics will account for about ten inches of the disparity.

  For much of the twentieth century, denizens of industrialized societies were growing taller at a rate of about one centimeter per decade. In the seventeenth century, the average Frenchman wa
s 5'4", which is now the average for an American woman. The first generation of Japanese born to immigrant parents in America, known as the Nisei, famously towered over their parents.

  In the 1960s, growth expert J. M. Tanner examined a set of identical twins that suggested the range of height variability caused by the environment. The identical boys were separated at birth, one brother raised in a nurturing household, and the other reared by a sadistic relative who kept him locked in a darkened room and made him plead for sips of water. In adulthood, the brother from the nurturing household was three inches taller than his identical twin, but many of their body proportions were similar. “The genetic control of shape is more rigorous than that of size,” Tanner wrote in Fetus into Man. The smaller brother was an abuse-shrunken version of the bigger brother.

  Little is known about the actual genes that influence height, however, because the genetics of even outwardly simple traits tend to be very complicated. A 2010 study in Nature Genetics needed 3,925 subjects and 294,831 single nucleotide polymorphisms—spots of DNA where a single letter can vary between people—to account for just 45 percent of the variance in height between adults, and that’s the best any study has done. Finding all the height genes will take much larger and more complex studies than scientists presumed a decade ago.

  Though the genes are difficult to pinpoint, the genetically programmed nature of height is obvious from studies of identical twins. Due to distinct intrauterine conditions, identical twins are often less similar in birth size than fraternal twins. And yet, after birth, the smaller twin of an identical duo quickly catches up with the bigger twin and they will be nearly or exactly the same height as adults. Similarly, female gymnasts delay their growth spurt with furious training, but that does not diminish their ultimate adult height. The genetic programming is also evident in the rate at which children grow. In World Wars I and II, European children were exposed to brief periods of famine during which their growth ground almost to a halt. When food again became plentiful, their bodies put the growth pedal to the metal such that adult height was not curtailed. “The undernourished child slows down and waits for better times,” Tanner wrote. “All young animals have the capacity to do this. . . . Man did not evolve in the supermarket society of today.”

  The permutations of size-determining interactions between nature and nurture are fathomless. Consider that children grow more quickly in spring and summer than in fall and winter, and that this is apparently due to sunlight signals that enter through the eyeballs, since the growth of totally blind children consists of similar fluctuations but are not synchronized with the seasons.

  The height that inhabitants of urban societies gained over the twentieth century came principally from increased leg length. Legs got longer faster than torsos. In developing countries that have gaping nutritional and infection-prevention disparities between the middle class and poor, the difference in height between the comfortable and the afflicted is all in the legs.

  Japan displayed a startling growth trend during its “economic miracle” period following World War II. From 1957 to 1977, the average height of a Japanese man increased by 1.7 inches, and of a woman by an inch. By 1980, the height of Japanese people in Japan had caught up with the height of Japanese people in America. Amazingly, the entire height increase was accounted for by increased leg length. Modern Japanese people are still short compared with Europeans, but not as short as they once were. And they now have more similar proportions.

  •

  There are, however, certain body type differences that have persisted over time and that have attracted the interest of sports anthropometrists. Every study that has examined race differences in body types has documented a disparity between black and white people that remains whether they reside in Africa, Europe, or the Americas. For any given sitting height—that is, the height of one’s head when one is sitting in a chair—Africans or African Americans have longer legs than Europeans. For a sitting height of two feet, an African American boy will tend to have legs that are 2.4 inches longer than a European boy’s. Legs make up a greater proportion of the body in an individual of recent African origin.* And this holds for elite athletes.

  Studies of Olympic athletes are uniformly consistent in finding that Africans and African Americans and African Canadians and Afro-Caribbeans have a more “linear” build than their competitors of Asian and European descent. That is, they tend to have longer legs and more narrow pelvic breadth.

  In their summary of the measurements of 1,265 Olympians from the 1968 Olympics in Mexico City, the scientists state that the successful body types within a sport are much more similar than body types between sports, regardless of ethnicity, but that “the most persistent of these differences” within sports are the narrow hip breadths and longer arms and legs of athletes with recent African ancestry. “They appear in virtually all the events,” the researchers write.

  Modern scientists who have measured athletes mention in their writing, sometimes reluctantly, that these body type differences influence athletic performance. The scientists are often careful to point out that a particular body type is not better overall, but that it may fit more readily into one sports niche than another. “This pattern may, in part, explain the tendency for the linear and relatively long-limbed east Africans to excel in endurance events while the short-limbed eastern Europeans and Asians have a long history of success in weight lifting and gymnastics,” write Norton and Olds, the Big Bang of body types gurus, in their textbook Anthropometrica.

  The limb-length difference manifests in NBA data as well.* In NBA predraft measurements for active players, the average white American NBA player was 6'7½" with a wingspan of 6'10". The average African American NBA player was 6'5½" with a 6'11" wingspan; shorter but longer. Both white and black players in the NBA have wingspan-to-height ratios much greater than the population average, but there’s a sizable gap between white and black players. The average ratio for a white American NBA player is 1.035, and for an African American NBA player 1.071. Still, there is wide variation among players within a given ethnicity. Two white players, Coby Karl (height: 6'3½", wingspan: 6'11") and Cole Aldrich (height: 6'9", wingspan: 7'4¾"), for example, have wingspan-to-height ratios approaching 1.10, but they are significant outliers compared with the other white players in the NBA. No other white players are even close, whereas a number of black players have larger ratios. When I showed this data to a scientist who studies athletes’ bodies, he responded: “So maybe it’s not so much that white men can’t jump. White men just can’t reach high.”*

  In a sense, this is last millennium’s news to scientists who have been studying body forms. In 1877, American zoologist Joel Asaph Allen published a seminal paper in which he noted that the extremities of animals get longer and thinner as one travels closer to the equator. African elephants can be distinguished from Asian elephants by their sail-like floppy ears. This is because the ears, like your skin, act as a radiator to release heat. The greater the surface area of the radiator compared with its volume, the more quickly heat is released. The African elephants, having evolved closer to the equator, have developed larger ears for cooling purposes. “Allen’s rule,” that animals from warmer climates tend to have longer limbs, has been extended to humans by a veritable filing cabinet full of studies.

  A 1998 analysis of hundreds of studies of native populations from around the world found that the higher the average annual temperature of a geographic region, the proportionally longer the legs of the people whose ancestors had historically resided there. Men and women from dozens of native populations on every inhabited continent were included, and when it came to leg length, they grouped by geography. Low-latitude Africans and Australian Aborigines had the proportionally longest legs and shortest torsos. So this is not strictly about ethnicity so much as geography. Or latitude and climate, to be more precise. Africans with ancestry in southern regions of the continent, farther from the equator, do
not necessarily have especially long limbs. But whether an African person in the study was from a population in Nigeria or from a genetically and physically distinct population in Ethiopia, so long as he was from low latitude his legs were likely longer than those of a height-matched European. And certainly longer than those of an Inuit from northern Canada, as Inuit tend to be short and stocky with compact limbs and a wide pelvis.*

  In the nineteenth century, Allen surmised that the long limbs of low-latitude animals were a direct result of a warm climate. In other words, he guessed that if a baby African elephant were adopted by Asian elephant parents and raised at high latitude in Asia, it would have the same smallish ears as Asian elephants. On that point, he was mistaken. Comparisons of human descendants of equatorial Africans and of Europeans who now live in the same country, like England or the United States, show that the limb differences remain. The effect of climate on extremities is therefore primarily through genetic selection over generations. Ancestral humans with shorter limbs had a greater chance of surviving and reproducing in cold northern latitudes because they retained more heat.

  In 2010, a racially diverse research team from Duke and Howard universities confronted the issue of body types as it pertains to ancestry and sports performance. The scientists did a backbend to avoid racial stereotyping. “Our study does not advance the notion of race,” they wrote. In a press release accompanying the study, Edward Jones, a black member of the research team, emphasized that access to sports facilities is critical for athletic development and that while growing up in South Carolina he was discouraged from swimming. Nonetheless, the researchers reported that, compared with white adults of a given height, black adults have a center of mass—approximately the belly button—that is about 3 percent higher. They used engineering models of bodies moving through fluids—air or water—to determine that the 3 percent difference translates into a 1.5 percent running speed advantage for athletes with the higher belly buttons (i.e., black athletes) and a 1.5 percent swimming speed advantage for athletes with a lower belly button (i.e., white athletes).