The Sports Gene: Inside the Science of Extraordinary Athletic Performance

by David Epstein

  Still, Albert Pujols and his All-Star peers see—and crush—95-mph fastballs for a living. So why are they transmogrified into Little Leaguers when faced with 68-mph softballs? It’s because the only way to hit a ball traveling at high speed is to be able to see into the future, and when a baseball player faces a softball pitcher, he is stripped of his crystal ball.

  •

  Nearly forty years ago, before Janet Starkes became one of the most influential sports expertise researchers in the world, she was a 5'2" point guard who spent one summer with the Canadian national team. Her lasting influence on sports, though, would come off the court, from the work she started as a graduate student at the University of Waterloo. Her research was to try to figure out why good athletes are, well, good.

  Tests of innate physical “hardware”—qualities that an athlete is apparently born with, like simple reaction time—had done astonishingly little to help explain expert performance in sports. The reaction times of elite athletes always hovered around one fifth of a second, the same as the reaction times when random people were tested.

  So Starkes looked elsewhere. She had heard of research on air traffic controllers that used “signal detection tests” to gauge how quickly an expert controller can sift through visual information to determine the presence or absence of critical signals. And she decided that conducting studies like these, of perceptual cognitive skills that are learned through practice, might prove fruitful. So, in 1975, as part of her graduate work at Waterloo, Starkes invented the modern sports “occlusion” test.

  She gathered thousands of photographs of women’s volleyball games and made slides of pictures where the volleyball was in the frame and others where the ball had just left the frame. In many photos, the orientation and action of players’ bodies were nearly identical regardless of whether the ball was in the frame, since little had changed in the instant when the ball had just exited the picture.

  Starkes then connected a scope to a slide projector and asked competitive volleyball players to look at the slides for a fraction of a second and decide whether the ball was or was not in the frame that had just flashed before their eyes. The brief glance was too quick for the viewer actually to see the ball, so the idea was to determine whether players were seeing the entire court and the body language of players in a different way from the average person that allowed them to figure out whether the ball was present.

  The results of the first occlusion tests astounded Starkes. Unlike in the results of reaction time tests, the difference between top volleyball players and novices was enormous. For the elite players, a fraction of a second glance was all they needed to determine whether the ball was present. And the better the player, the more quickly she could extract pertinent information from each slide.

  In one instance, Starkes tested members of the Canadian national volleyball team, which at the time included one of the best setters in the world. The setter was able to deduce whether the volleyball was present in a picture that was flashed before her eyes for sixteen thousandths of a second. “That’s a very difficult task,” Starkes told me. “For people who don’t know volleyball, in sixteen milliseconds all they see is a flash of light.”

  Not only did the world-class setter detect the presence or absence of the ball in sixteen milliseconds, she gleaned enough visual information to know when and where the picture was taken. “After each slide she would say ‘yes’ or ‘no,’ whether the ball was there,” Starkes says, “and then sometimes she would say, ‘That was the Sherbrooke team after they got their new uniforms, so the picture must have been taken at such and such a time.’” One woman’s blink of light was another woman’s fully formed narrative. It was a strong clue that one key difference between expert and novice athletes was in the way they had learned to perceive the game, rather than the raw ability to react quickly.

  Shortly after she received her Ph.D., Starkes joined the faculty at McMaster University and continued her occlusion work with the Canadian national field hockey team. At the time, the coaching orthodoxy in field hockey favored the idea that innate reflexes were of primary importance. Conversely, the idea that learned, perceptual skills were a hallmark of expert performance was, as Starkes put it, “heretical.”

  In 1979, when Starkes began helping the Canadian national field hockey team gear up for the 1980 Olympics, she was dismayed to find that the national coaches were relying on outdated ideas to choose and arrange the team. “They thought everybody saw the field the same way,” she says. “They were using simple reaction time tests for selection, and they thought it would be a good determinant of who would be the best goalies or strikers. I was astounded that they had no idea that reaction time might not be predictive of anything.”

  Starkes, of course, knew better. In her occlusion tests of field hockey players, she found just what she had found in volleyball players, and more. Not only were elite field hockey players able to tell faster than the blink of an eye whether a ball was in the frame, they could accurately reconstruct the playing field after just a fleeting glance. This held true from basketball to soccer. It was as if every elite athlete miraculously had a photographic memory when it came to her sport. The question, then, is how important these perceptual abilities are to top athletes and whether they are the result of genetic gifts.

  There is no better place to look for an answer than in a type of competition where the action is slow, deliberate, and devoid of the constraints of muscle and sinew.

  •

  In the early 1940s, Dutch chess master and psychologist Adriaan de Groot began drilling for the core of chess expertise. De Groot would test chess players of various skill levels and attempt to dissect what made a grandmaster better than an average professional, and the average professional far superior to a club player.

  The common wisdom of the time was that highly skilled chess players thought further ahead in the game than did less skilled players. This is true when skilled players are compared with complete novices. But when de Groot asked both grandmasters and merely strong players to narrate their decision making in the face of an unfamiliar game situation, he found that players of disparate skill levels mulled over the same number of pieces and proposed essentially the same array of possible moves. Why then, he wondered, do the grandmasters end up making better moves?

  De Groot assembled a panel of four chess players as representatives of their varying skill echelons: a grandmaster and world champion; a master; a city champion; and an average club player.

  De Groot enlisted another master to come up with different chess arrangements taken from obscure games, and then did something very similar to what Starkes would do with athletes thirty years later: he flashed the chessboards in front of the players for a matter of seconds and then asked them to reconstruct the scenario on a blank board. What emerged were differences between the skill levels, particularly the two masters and the two nonmasters, “so large and unambiguous that they hardly need further support,” de Groot wrote.

  In four of the trials, the grandmaster re-created an entire board after viewing it for three seconds. The master was able to accomplish the same feat twice. Neither of the lesser players was able to reproduce any boards with complete accuracy. Overall, the grandmaster and master accurately replaced more than 90 percent of the pieces in the trials, while the city champion managed around 70 percent, and the club player only about 50 percent. In five seconds, the grandmaster understood more of the game situation than the club player did in fifteen minutes. In these tests, de Groot wrote, “it is evident that experience is the foundation of the superior achievements of the masters.” But it would be three decades before confirmation would come that what de Groot saw was indeed an acquired skill, and not the product of innately miraculous memory.

  In a seminal study published in 1973, Carnegie Mellon University psychologists William G. Chase and Herbert A. Simon—a future Nobel Prize winner—repeated the de Groot exper
iment, and added a twist: they tested the players’ recall for chessboards that contained random arrangements of pieces that could never occur in a game. When the players were given five seconds to study the random assortments and then asked to re-create them, the recall advantages of the masters disappeared. Suddenly, their memories were just like those of average players.

  In order to explain what they saw, Chase and Simon proposed a “chunking theory” of expertise, a pivotal idea in the study of games like chess, but also in sports, that helps explain what Janet Starkes found in her work with field hockey and volleyball players.

  Chess masters and elite athletes alike “chunk” information on the board or the field. In other words, rather than grappling with a large number of individual pieces, experts unconsciously group information into a smaller number of meaningful chunks based on patterns that they have seen before. Whereas the average club player in de Groot’s study was scanning and attempting to remember the arrangement of twenty individual chess pieces, the grandmaster needed to remember only a few chunks of several pieces each, because the relationships between the pieces had great meaning for him.*

  A grandmaster is fluent in the language of chess and has a mental database of millions of arrangements of pieces that are broken down into at least 300,000 meaningful chunks, which are in turn grouped into mental “templates,” large arrangements of pieces (or players, in the case of athletes) within which some pieces can be moved around without rendering the entire arrangement unrecognizable. Where the novice is overwhelmed by new information and randomness, the master sees familiar order and structure that allows him to home in on information that is critical for the decision at hand. “What was once accomplished by slow, conscious deductive reasoning is now arrived at by fast, unconscious perceptual processing,” Chase and Simon wrote. “It is no mistake of language for the chess master to say that he ‘sees’ the right move.”

  Studies that track the eye movements of experienced performers, whether chess players, pianists, surgeons, or athletes, have found that as experts gain experience they are quicker to sift through visual information and separate the wheat from the chaff. Experts swiftly move their attention away from irrelevant input and cut to the data that is most important to determining their next move. While novices dwell on individual pieces or players, experts focus more attention on the spaces between pieces or players that are relevant to the unifying relationship of parts in the whole.

  Most important in sports, perceiving order allows elite athletes to extract critical information from the arrangement of players or from subtle changes in an opponent’s body movements in order to make unconscious predictions about what will happen next.

  •

  Bruce Abernethy was an undergraduate at the University of Queensland in the late 1970s and an avid cricket player when he began to expand on Janet Starkes’s occlusion methods. Abernethy started out using Super 8mm film to capture video of cricket bowlers. He would show batters the video but cut it off before the throw and have them attempt to predict where the ball was headed. Unsurprisingly, expert players were better at predicting the path of the ball than novice players.

  In the decades since, Abernethy, now associate dean for research at Queensland, has become exceedingly sophisticated at using occlusion tests to illuminate the basis of perceptual expertise in sports. Abernethy has moved his studies from the video screen to the field and the court. He has equipped tennis players with goggles that go opaque just as an opponent is about to strike the ball, and he has outfitted cricket batters with contact lenses with varied levels of blurriness.

  The theme of Abernethy’s findings is that elite athletes need less time and less visual information to know what will happen in the future, and, without knowing it, they zero in on critical visual information, just like expert chess players. Elite athletes chunk information about bodies and player arrangements the way that grandmasters do with rooks and bishops. “We’ve tested expert batters in cricket where all they see is the ball, the hand and wrist, and down to the elbow, and they still do better than random chance,” Abernethy says. “It looks bizarre, but there’s significant information between the hand and arm where experts get cues for making judgments.”

  Top tennis players, Abernethy found, could discern from the minuscule pre-serve shifts of an opponent’s torso whether a shot was going to their forehand or backhand, whereas average players had to wait to see the motion of the racket, costing invaluable response time. (In badminton, if Abernethy hides the racket and entire forearm, it transforms elite players back into near novices, an indication that information from the lower arm is critical in that sport.)

  Pro boxers have a similar skill. A Muhammad Ali jab took a mere forty milliseconds to arrive at the face of a victim standing a foot and a half away. Without anticipation based on body movements, Ali’s opponents would have been beaten down in round one, hit flush by every punch. (Ali’s skill at disguising the trajectory of a punch, and thus confounding the opponent’s anticipation, often meant they were finished a few rounds later anyway.)

  Even skills that appear to be purely instinctive—jumping to rebound a basketball after a missed shot—are grounded in learned perceptual expertise and a database of knowledge on how subtle shifts of a shooter’s body will alter the trajectory of the ball. It’s a database that can be built only through rigorous practice.*

  Without that database, every athlete is a chess master facing a random board, or Albert Pujols facing Jennie Finch, stripped of the information that allows him to predict the future.* Since Pujols had no mental database of Finch’s body movements, her pitch tendencies, or even the spin of a softball to predict what might be coming, he was always left reacting at the last moment. And Pujols’s simple reaction speed is downright quotidian.

  When scientists at Washington University in St. Louis tested him, Pujols, the greatest hitter of an era, was in the sixty-sixth percentile for simple reaction time compared with a random sample of college students.

  •

  No one is born with the anticipatory skills required of an elite athlete. When Abernethy studied the eye movement patterns of elite and novice badminton players, he saw that the novices were already looking at the correct area of the opponent’s body, they just did not have the cognitive database needed to extract information from it. “If they did,” Abernethy says, “it would be a hell of a lot easier to coach them to become an expert. You could just say, ‘Look at the arm. Or for a baseball batter the real advice wouldn’t be ‘keep your eye on the ball,’ it would be ‘watch the shoulder.’ But actually, if you tell them that, it makes good players worse.”

  As an individual practices a skill, whether it be hitting, throwing, or learning to drive a car, the mental processes involved in executing the skill move from the higher conscious areas of the brain in the frontal lobe, back to more primitive areas that control automated processes, or skills that you can execute “without thinking.”

  In sports, brain automation is hyperspecific to the practiced skill, so specific that brain-imaging studies of athletes who train in a particular task show that activity in the frontal lobe is turned down only when they do that exact task. When runners are put on bicycles or arm bikes (where the pedals are moved with hands instead of feet) their frontal lobe activity increases compared with when they are running, even though cycling or arm cycling wouldn’t seem to require much conscious thought. The physical activity that one trains in is very specifically automated in the brain. To return to Abernethy’s point, “thinking” about an action is the sign of a novice in sports, or a key to transforming an expert back into an amateur. (University of Chicago psychologist Sian Beilock has shown that a golfer can overcome pressure-induced choking in putting—paralysis by analysis, she calls it—by singing to himself, and thus preoccupying the higher conscious areas of the brain.)

  Chunking and automation travel together on the march toward expertise. It is only
by recognizing body cues and patterns with the rapidity of an unconscious process that Albert Pujols can determine whether he should swing at a ball when it has barely left the pitcher’s hand. The same goes for quarterback Peyton Manning. He cannot stop in the face of blitzing linebackers and consciously sort through the defensive alignments and patterns he learned in hours and years of practicing and studying game film. He has seconds to scan the field and throw. He is a grandmaster playing speed chess, only with linebackers and safeties in place of knights and pawns. (At the same time, NFL defensive coordinators are shuffling their players in an attempt to present Manning with a chessboard that looks misleading or random.)

  The result of expertise study, from de Groot to Abernethy, can be summarized in a single phrase that played like a broken record in my interviews with psychologists who research expertise: “It’s software, not hardware.” That is, the perceptual sports skills that separate experts from dilettantes are learned, or downloaded (like software), via practice. They don’t come standard as part of the human machine. That fact helped spawn the most well-known theory in modern sports expertise, and one that has no place for genes.

  —

  It started with musicians.

  For a 1993 study, three psychologists turned to the Music Academy of West Berlin, which had a global reputation for producing world-class violinists.

  The academy professors helped the psychologists identify ten of the “best” violin students, those who could become international soloists; ten students who were “good” and could make a living in a symphony orchestra; and ten lesser students they categorized as “music teachers,” because that would be their likely career path.