Mystery

by Jonah Lehrer

  The first wonderful thing you can do, at least from the perspective of a greedy casino, is advertise slots with huge jackpots. Rather than be limited by the number of mechanical reels, casinos could program the random generator to pick from a number set of any size. Slots could now feature millions of possible outcomes, with each of these outcomes mapped onto a specific set of symbols. Let’s imagine, for instance, a dollar slot offering a $1 million payout. That large payout would be linked to a single reel—say, three sevens across—but that reel would only appear if one winning number was generated. The other millions of outcomes would be mapped onto some combination of losing reels, thus allowing the house to advertise a huge payout while still ensuring a healthy profit margin.V

  This new bit of programming made slots far more appealing. Although the chance of winning the jackpot on a slot machine with hundreds of virtual reels might be as low as 1 in 137,000,000, the gaming device felt far easier to beat. (One early analysis of virtual-reel slots found that, if the machines paid out according to the perceived frequency of their symbols, players would actually come out way ahead.) So players kept inserting coins and credits into the mystery box, chasing a reward they’d never receive.

  It didn’t take long before gambling companies realized how much money they could make with these virtual reels. According to the anthropologist Natasha Dow Schüll, who chronicles the history of the slot machine in Addiction by Design, more than 80 percent of spinning slots used virtual reels by the mid-1990s.15

  However, casinos soon discovered an even more effective use of virtual reels. The key was to manipulate the way players perceived their losses, not just their potential gains. The machines did this through the frequent use of “near misses,” a gambling illusion in which people think they almost won because the reel stops next to a winning symbol. Look, for instance, at the evil genius of slot machine manufacturer Universal, which developed a two-stage process after each spin. The first stage determined whether the player won. If he lost—and the vast majority of spins are losers—the second stage initiated the near-miss effect, setting up the player to believe he had come exceedingly close to a real payout. (There might be two sevens on the main pay line, and then a third seven just below.) Although near misses cost the casinos nothing, they keep gamblers motivated, persuading people to stick with a game that’s stacked against them.

  It’s now possible to understand why near misses are so compelling. In a recent Neuron paper, scientists at Cambridge University showed that near misses on a slot machine task activated the same reward circuitry as an actual win, triggering blood flow to dopamine-rich areas of the brain.16

  Why are we wired this way? One hypothesis is that enjoying near misses helps us persist when learning a difficult new skill. Let’s say we’re practicing a three-point shot in basketball. At first, our shots are going to be all over the place, a seemingly random distribution of bricks and airballs. Yet, as we slowly get better, those shots will get closer to the rim. A few might even go in, which can be pretty thrilling. Near misses, then, keep us motivated as we slowly improve our form. If we only got excited by makes, we’d give up. The brain needs a mechanism to enjoy incremental progress.

  Unfortunately, this practical software is cruelly misled by those gambling machines. There is no skill with these mystery boxes; pulling a lever doesn’t take talent. Nevertheless, those dopamine neurons activated by near misses—the peppy cheerleaders of the brain—urge us onward anyway, insisting that we keep playing because we keep almost winning. Alas, the only thing we’re getting better at is losing money.

  The larger lesson of slot machines is that culture never stays still—it is constantly evolving to better fit the grooves of the human mind. Over time, slots have become mystery boxes of astonishing power.

  The slot machines do this by carefully manipulating randomness to create a tantalizing state of mystery. While pure randomness quickly gets tedious, virtual reels allow the slots to disguise their inner chaos, tricking the brain into seeing subtle patterns. If we keep gambling, we’ll get those triple sevens; all those near misses must mean we’re getting closer. It’s that illusory sense of progress—the promise of a mystery box we might one day open—that makes slots so tragically compelling.

  To explain the power of the slot machine is not to excuse it: the casino delivers squirts of chemical pleasure stripped of all context and meaning. Yet, one can see how the appeal of these gambling gadgets also relies on the same basic psychological mechanisms as the toy egg and Star Wars. (As the critic Dave Hickey wrote, in an astute observation about Las Vegas and America, “What is hidden elsewhere exists here [in Vegas] in quotidian visibility.”)17 We want surprise and suspense, but we also crave order and closure. The artistry of the mystery box is in the balance: give away too much and we’re bored, give away too little and we’re lost.

  This is known as the inverted-U curve theory of curiosity. It was first identified by the psychologist Daniel Berlyne, in a series of classic studies done in the late 1960s.18 Berlyne began the experiments by showing people a collection of simple visual patterns, such as four identical squares, or the outline of a sun. Berlyne then introduced complications, such as asymmetry and irregularity. He added in smudges of randomness and extraneous detail.

  As subjects looked at these images, Berlyne asked them to rate each for “pleasingness,” “interestingness,” and on a scale from ugly to beautiful, which he referred to as “hedonic value.” He also measured how long the people looked at the patterns.

  The results looked like an upside-down U, hence the name of the phenomenon:

  Berlyne found that simple and familiar shapes bored people; nobody wanted to keep staring at a few straight lines. But people also dismissed shapes that were too random and incongruous. Our attention had a hedonic sweet spot, clearly preferring patterns that were unknown but not unknowable. (In Berlyne’s formulation, they were either simple and novel or complex and familiar.)19 We wanted a mystery—a new visual form—but one we could still decipher.VI And as we’ll soon see, the ability to locate the ideal amount of mystery, and to hide it within the right box, helps explain the most popular form of entertainment in the world.

  The Rule Change of 1893

  If an anthropologist from Mars studied our culture, our obsession with sports would mystify her. According to one recent measure, ninety-three of the top one hundred American television programs watched live across a single year have been sports related.21 More people watched the Super Bowl than the Oscars, Emmys, Grammys, Golden Globes, and Tonys combined.

  Yet, as the Martian would surely notice, these contests have no stakes, at least in the real world; it doesn’t matter which team wins. Nevertheless, we lavish vast amounts of attention on these freakishly shaped athletes playing with bouncy balls. We spend a fortune on gigantic stadiums and tickets to those stadiums. The games make us cheer and scream and cry.

  To explain this peculiar human behavior, the alien anthropologist might begin with a search of the scientific literature. She would come across a wide range of potential explanations. These include the tribal theory—teams are like tribes, hijacking our Neolithic social instincts—and the mirror neuron speculation, which holds that we enjoy watching athletes because our brain imitates their perfect physical movements.22 We live vicariously through their grace.

  These theories are nice. They have a logical sheen. But they fail to explain why some sports are so much more popular than others. After all, not every game makes us care. It’s the rare competition that turns us into passionate fans or gets a prime-time spot on national television. So what is it about our most successful sports that makes them successful?

  If the Martian tried to answer these questions, she might eventually run across a largely forgotten paper by Nicholas Christenfeld, a psychologist who spent his career at the University of California, San Diego.23 A wiry man, with a chiseled face, twitchy hands, and sardonic sense of humor, Christenfeld is that rare modern scientist who refuses
to narrow his interests. He has studied the psychology of ums and uhs24—art historians use more of them than chemistsVII—and the biases that determine our choice of bathroom stalls.25 (People are much more likely to choose a middle stall, due to our deep-seated aversion to edges.) Christenfeld has looked at whether dogs resemble their owners (they do),26 if it’s possible to be tickled by a robot (it is),27 and the impact of a given name on mortality (men with “negative” initials, such as PIG or DIE, live 2.8 years less on average than matched controls).28 “If you had to categorize my research, it’s about the social psychology of everyday life,” he says. “But mostly I’m just interested in the same idiosyncratic questions that everyone is interested in. I mean, who hasn’t wondered if dogs look like their owners?”

  Christenfeld’s interest in sports began in an unlikely place, with a question about the novelist Joseph Heller, the author of Catch-22. Simply put, Christenfeld wanted to understand why Heller wrote only one great novel. “Heller wrote other books, sure, but no one thinks they’re better than Catch-22,” Christenfeld says. “Maybe the truth is that Heller only had one great book in him.” But how is that possible? If Heller was capable of one masterpiece, shouldn’t he be able to write a second?

  These questions led Christenfeld to think more generally about the reliability of human achievement. Perhaps Heller never wrote another great novel because creativity is tangled up with luck and contingency. (Even Shakespeare wrote mediocre plays.) “Maybe it wasn’t Heller’s fault,” Christenfeld says. “Perhaps he had the talent, but just didn’t get lucky twice.”

  To explore the role of randomness in the creative process, Christenfeld began looking at one-hit wonders in other fields. Such artists exist in music, of course—Right Said Fred and Vanilla Ice, QED—but Christenfeld also found plenty of scientists whose entire careers depended on a single breakthrough. (They were the empirical version of “Ice Ice Baby.”) This doesn’t mean these researchers weren’t smart. Rather, they just never got lucky enough to hit the achievement jackpot again.

  The problem with this research approach is that success in the arts and sciences is full of confounding variables. Picasso created many second-rate sketches, but they’re still “Picassos” and thus hang in museums. (His reputation distorts our critical judgment.) As Christenfeld puts it, “If you have one hit, it’s often much easier to get that second hit. There’s a non-independence problem.”

  To get around this issue, Christenfeld decided to look at athletic competitions, since the measures of success are far more objective. He began with sprinters. “Take a guy like Usain Bolt,” Christenfeld says, citing the world record holder in the 100 and 200 meters. “He’s the fastest man alive, but what is the standard deviation of his performances? I mean, if you see Bolt is in a race, are you fairly certain that he’s going to win?”

  The answer is an emphatic yes; running speed is far more predictable than success in the music industry, science, or literature. During his ten-year peak, Bolt had a winning percentage at major track events of 84 percent. That’s a success rate Right Said Fred never dreamed of.

  Christenfeld’s curiosity is chronic; every answer only leads to more questions. “The predictability of sprinting got me thinking,” Christenfeld says, in between bites of panini at the local university café. “The fastest guy almost always finishes first, but is that optimal? If you know who’s going to win simply by looking at who’s competing, isn’t that a little boring?” While people clearly want sports that reward talent, Christenfeld knew that we also crave surprise, the thrill of an unlikely upset. “That struck me as an interesting tension. I began to wonder if there’s an ideal level of predictability for these sporting competitions, and if I could find it.”

  This search for the perfect sport led Christenfeld to assess the statistical reliability of the most popular ones, including baseball, hockey, soccer, basketball, and football. (He assumed that their popularity was not an accident.) Christenfeld randomly divided each of their seasons in two segments and then asked a simple question: To what extent did a team’s success in half of its games predict its success in the other half? If a sport is statistically reliable, then it should produce predictable outcomes; the better team should almost always win, just like Usain Bolt. An unreliable sport would be full of one-hit wonders like Joseph Heller and Vanilla Ice, teams whose performances were highly variable and inconsistent.

  The first thing Christenfeld discovered is that different sports generate very different reliabilities per game. Major League Baseball, for instance, produces single-game outcomes that are roughly fourteen times less reliable than those of the NFL. (Put another way, the better football team almost always wins, while the better baseball team can easily lose.) But baseball is not all luck and chance. Instead, Christenfeld points out that the randomness of a single baseball game is balanced out by a regular season of 162 games, or ten times longer than the NFL season. What’s more, Christenfeld found the same pattern in every sport he looked at, so that season length was always inversely related to single-game reliability. “The sports whose single games reliably assess talent have short seasons, while those whose games are largely chance have long ones,” Christenfeld wrote. “Thus these sports, differing enormously in their particulars, converge towards the same reliability in a season.” According to Christenfeld, this means that season length is not an “arbitrary product of historical, meteorological or other such constraints.” Rather, it is rooted in the desire of fans to witness a “proper mix of skill and chance.”

  The skill we know about. Christenfeld’s research highlights the importance of chance. By proving that the most popular sports share a similar level of unpredictability, at least over an entire season, Christenfeld revealed their inherent mysteriousness, which is an essential element of their appeal. “Drama requires uncertainty,” he told me. “It requires, at a basic level, that you don’t know what’s going to happen next.” As a result, the most popular sports have evolved to ensure that the mystery remains: the rules of the game intentionally constrain the talent of the players.VIII “If sports were pure contests of skill, then they’d quickly become genetic tournaments,” Christenfeld says. “But that’s not much fun, is it? The result is way too predictable.”

  What fans crave is what Christenfeld calls an “optimal level of discrepancy.” Although the better team should usually win, the best games are also full of surprises, built around interactions that are inherently unknowable.29 In Christenfeld’s telling, the rules of sports are continually revised to find this ideal balance, the peak of that inverted-U curve. They are mystery boxes, engineered to deliver the right amount of uncertainty just like Star Wars or a slot machine. “You can’t let any single talent get too dominant, because then you’re back to the ‘Usain Bolt always wins’ problem,” Christenfeld says. “The problem is that predictability is boring, even when it’s earned.”

  Just look at baseball. As Christenfeld notes, the mystery of the sport is rooted in its basic mechanics, in which a batter swings a rounded bat at a small ball traveling fast. “The cruel thing about baseball is that the difference between a double down the line and a double play comes down to a few millimeters,” Christenfeld says. “This means there’s a limit to what even the best players can control.”

  The history of baseball is largely the story of a game trying to protect this essential mystery. That’s what happened during the rule change of 1893, which was a desperate attempt to save the young sport. At the time, the problem with baseball seemed obvious: hitters had stopped hitting. Since 1887, the batting average of National League players had plummeted from .269 to .245, while the number of strikeouts recorded by each team had increased by more than 41 percent. The best team in baseball, the Boston Beaneaters, hit 34 home runs combined over a single season.30

  The decline of hitters was caused by the rise of pitchers. In the early 1890s, the fastball got faster—Amos Rusie of the New York Giants was reportedly throwing the ball almost a hundred miles per h
ourIX—while the newly invented curve, or “skewball,” befuddled batters. The result was a predictable sport: the only players who mattered were the ones throwing the ball. If a good pitcher was on the mound, his team was almost sure to win. These boring games soon led to serious business issues. Attendance at ballparks was in free fall; small-market teams were bleeding money. To stay afloat, player salaries were cut by nearly 40 percent before the start of the 1893 season.32 It wasn’t clear how much longer the sport could survive.

  To boost attendance, the owners decided to make a dramatic change to the rules of the game. Pitchers had previously been allowed to pitch from the back of a square box, fifty-six feet away from home plate. The owners decided to push this distance back to sixty feet six inches. Their logic was straightforward: if batters had an additional split second to hit the ball, they might have better luck against the new generation of aces.33

  Nearly everyone criticized the change. Batters said it wouldn’t make a difference: When the speed of a fastball was approaching triple digits, what’s the point of four measly feet? (Besides, the trickery of the skewball didn’t depend on speed.) Pitchers, meanwhile, claimed the rule change was unfair—it wasn’t their fault that batters couldn’t hit their pitches. Traditionalists fretted that the young game was being ruined, and small-market teams wanted to return to the hitter’s paradise of underhand pitching.

  But the new rules worked. Given the biological limits of the human arm and the reaction time of the central nervous system, putting sixty feet between the mound and home plate creates a near-perfect mix of skill and chance. Batters can make contact, but they struggle to control the direction of their hits. While the owners almost certainly weren’t thinking about the statistical variance of baseball in 1893—they just wanted more offense—they stumbled upon the optimal level of discrepancy, squeezing as much surprise as possible from a hurled ball and a swinging stick.