Can't Just Stop

by Sharon Begley

  Four years after Rovio Entertainment unleashed that game on the world, it had been downloaded two billion times, as people couldn’t resist the chance to use a virtual slingshot to fling vexed avians at egg-stealing green pigs. Why? There are lots of explanations for why the game is fun—it’s easy, there’s no learning curve, and landing a direct hit on a pig makes it blow up, which everyone’s inner seven-year-old loves. But the reason the game is compelling lies deeper. When an action is typically followed by a reward, the way that flinging a bird is followed by a pig exploding, it triggers the brain’s dopamine system. Although once believed only to create the subjective feeling of reward or pleasure, in fact the dopamine system is more sophisticated: it calculates the likelihood that an action will bring a reward and sets the brain’s expectation module accordingly. “Dopamine’s presence signals the brain that there is a reward coming, like glass-and-wood houses deliciously flying apart,” psychologist Michael Chorost wrote in Psychology Today in 2011 (he deleted Angry Birds from his phone to stop his compulsive playing). “But the brain doesn’t know how good the reward will be. Will the bird just glance off the top, or will it score a glorious direct hit? That uncertainty creates a tension, and the brain craves release. It makes you want to do whatever it is that creates the release,” such as pulling the virtual slingshot again and again and again.

  No wonder many people who can’t stop playing Bejeweled or even FreeCell describe the experience as not all that pleasurable. They feel stuck, unable to extricate themselves from the clutches of the game, compelled to keep playing but deriving very little enjoyment from it except in those rare moments of success. Somehow, video games tap into deep psychological tendencies that make us expect pleasure, deliver an experience that’s frustrating, and make us crave the same experience over and over even though we know the disappointing, aggravating outcome will in all likelihood be the same. The reason games can be compelling without being much fun is that designers take advantage of two common psychological quirks: variable reinforcement and intermittent reinforcement.

  Intermittent reinforcement refers to a varying probability of reaping a reward; sometimes you are rewarded (with game loot or leveling up, for example) for an accomplishment and other times the same action yields you . . . nothing. Variable reinforcement describes a system in which the value of rewards for a given accomplishment changes. Slot machines are the epitome of variable and intermittent reinforcement. The only action a player executes is pulling the arm of the one-armed bandit—or, now that mechanics has been replaced with electronics, pressing a button. Sometimes you win, sometimes you win big, and most times you lose. The input is identical, the output as varied as a jackpot and a bust. No wonder the stereotypical image of a slot machine player is someone staring at the machine as if hypnotized, feeding in quarters robotically, compulsively, until her cup of coins is empty and she gets back on the bus.

  Like a slot machine, “Diablo 3 uses variable rewards, which is one of the things that make it so compelling,” Mikros told me. For the uninitiated, Diablo 3 is the 2012 release of a game franchise that was launched by Blizzard Entertainment in late 1996. All three iterations belong to the genre known as action role-playing hack-and-slash games. A player/hero navigates through the Kingdom of Khanduras, battling ghouls and other enemies in order to rid the world of Diablo, the Lord of Terror. If she can beat sixteen dungeon levels and reach Hell, she faces Diablo in the ultimate battle. Along the way the player casts spells, acquires weapons and other loot, and interacts with characters including a warrior, a rogue, and a sorcerer.

  At the start of the game, rewards are essentially fixed: kill the monster and something good happens, such as leveling up or winning “experience” (essentially, power). As you progress through the game, however, the likelihood of a drop (being rewarded with a powerful new weapon or other tool that will help you survive and advance) decreases, but its value increases. “You’re still looking forward to it but you’re not necessarily getting it all the time,” Mikros said. “You’ve already established the connection that killing this demon or this monster is going to give you something good, like gold or a special sword or a special bow. But now, because you don’t know if you’re going to get anything, you’re waiting expectantly, even anxiously.”

  Game designers call this a “compulsion loop.” It has a brain basis, one that gets to the heart of the hybrid nature of gaming: like other electronic catnip such as email and texting, video games are perhaps the paradigmatic example of activities where addiction and compulsion morph into one another like shape-shifting demons.

  Enter Dopamine

  Addictions run on a desperate need for another hedonic hit. That’s because addictions are born in pleasure; the initial experience is rewarding, exciting, fun, risky. Those feelings are produced by what’s been called the brain’s reward circuitry, which is activated when we experience something pleasurable and consists of neurons running on dopamine. “Run on” means that when an electrical signal reaches the end of one neuron, it jumps across a synapse to the next by triggering the release of dopamine from the first neuron. The dopamine flows across the gap between the two neurons and is gathered in by the receiving neuron like the International Space Station gathering in a Soyuz supply ship. In neurons, the docking port is called a dopamine receptor. The act of docking triggers the propagation of an electrical signal down the length of the receiving neuron, on and on until the activity is eventually registered, subjectively, as a feeling of reward that we feel from food, sex, alcohol, nicotine, cocaine, and killing monsters in Diablo 3—making them all deeply, joyously, deliriously, euphorically reinforcing.

  Nothing in the brain is as simple as scientists initially think, however, including the dopamine story. Reward circuitry is actually better understood as an expectation machine: it makes predictions about how rewarding an experience will be.

  For a deeper explanation of how video game designers exploit the dopamine system, I called Jamie Madigan, who has a Ph.D. in psychology and worked for years at a gaming company. Madigan has made a name for himself in the gaming world through his website “The Psychology of Video Games,” where he posts essays on precisely that, including a “dopamine freak-out” that he said he had just barely survived while playing Diablo 3.

  At the end of Diablo 3, “You’ve completed the story line” of defeating ghouls and monsters to reach and battle Diablo, he told me, “and are just acquiring more and better loot in order to kill more monsters and get still better loot.” There are more than a dozen levels, “and the monsters get tougher to beat unless you’ve acquired better equipment. It never ends. Eventually, I did realize that I was doing the same thing over and over for three hours a night and it was no longer fun. If someone like me, who knows what elements of a game make it compelling, can feel this pull . . .” His voice drifted off.

  But what are those compulsion-making elements? Like Diablo, Nik Mikros’s paradigmatic example of a game that has aced the variable/intermittent reward structure, the immensely popular World of Warcraft is also renowned for drawing players in compulsively by offering unpredictable and unexpected windfalls. A massively multiplayer online role-playing game (MMORPG) released in 2004, it has more than 10 million subscribers, each choosing a character and moving through multiple levels of the virtual world on a quest. In World of Warcraft, players choose a profession such as blacksmithing or mining, and can master any of four secondary skills (archaeology, cooking, fishing, or first aid). They band together to accomplish tasks, either ad hoc or through enduring associations called guilds, by inviting others to join via player-to-player instant messaging, group-wide “text channels,” or, in some games, voice systems. Guilds give players access to features that can help them on quests, the missions that form the narrative spine of the game and earn players experience points, useful objects, skills, and money. World of Warcraft and other MMORPGs also offer the escapism of entering a fully formed, complex, intriguing world without hectoring parents or ab
usive bosses or unappreciative spouses. And they exploit our drive to achieve even when the achievement—conquering foes, killing monsters, rescuing princesses, accumulating wealth or status, and advancing to higher levels—is not quite real.

  But that’s the benign, psychologically obvious appeal of multiplayer games. Jamie Madigan fell victim to another. One day, he was merrily killing World of Warcraft bandits, which earned him a chance at one or more pieces of armor, weapons, or other supplies—“loot”—that would help him in subsequent battles and quests. Loot varies in quality, with the color of the text identifying it indicating its worth: gray is the lowest, white a little more valuable, on up through green, blue, purple, and orange. The character a player chooses as his avatar also has a place in the pecking order: “classes” such as monks, rogues, shamans, warriors, and druids each have a specific adventuring style, defined by the weapons and defenses they can deploy and the skills, powers, and magic they can gain by achieving various milestones. Since his character was fairly unimpressive, and therefore unlikely to score valuable loot, Madigan was “shocked by the loot drop: a rare pair of ‘blue’ gloves that perfectly fit my class’s needs at the time,” he recalled. For a lowly character “to find a blue item on a random enemy was actually very rare, and I experienced a huge rush from it,” he said. “But more importantly, with that came an acute desire to keep playing the game and to murder more bandits.”

  Intermittent reinforcement in the form of random loot drops stokes the brain in ways that expected and predictable rewards do not. “It’s incredibly effective at making people keep playing because of how the dopamine-based reward circuitry works,” Madigan said. Remember that dopamine neurons predict the pleasure hit that will flood the brain from an enjoyable experience, firing even before the reward arrives (such as when the microwave beep alerts you that your yummy Hot Pocket is cooked, Madigan offers). “But this is only part of what makes loot-based games work so well,” he continued. “The real key is that while dopamine neurons fire once your brain has figured out how to predict an event, they really go nuts when an unexpected, unpredicted gush of dopamine shows up, giving you an even bigger rush. It’s like, DUDE! UNEXPECTED HOT POCKET! KEEP DOING WHAT YOU’RE DOING UNTIL WE FIGURE OUT HOW TO MAKE IT HAPPEN AGAIN! So you keep playing.”

  And don’t count on the rational brain to tell you, enough! When your emotions are firing on all cylinders, as when you mow down bad guys in an online shooter game or careen through the diabolical racecourses of Gran Turismo stoked by squealing-wheel sound effects, you don’t remember that you have to get dinner on the table, or prepare for a presentation at work tomorrow, or finish that term paper. “Despite any intentions born of rational thought, you’re just not thinking with the same brain after some infuriating punk has bested you in a shooter or you’ve just pulled off some thrilling act of derring-do in some other game,” Madigan explained. “Rationality gets elbowed aside and you look up to realize that it’s a quarter to three on a weekday morning. And yet you’re still muttering ‘Okay, just one more . . .’ ”

  Nik Mikros wasn’t happy that video game designers had tapped into the dopamine system. While it seems as if half the basements in Brooklyn are populated by game designers who program compulsion loops into their creations, not all designers are proud of what their field has nailed with such devious effectiveness. “What makes my skin crawl is games that are completely driven by Skinner boxes,” Mikros told me as I was gathering up my stuff. “It’s not why I want to make games, just to give people pellets—hit the bar, get the pellets. I don’t think that’s improving humanity.”

  Home-Brew Neuroscience

  I was beginning to feel as if I were on a quest for Easter eggs in Halo 3, hoping the next room I entered—the next expert I interviewed—would reveal more secrets of compulsive gaming. My next stop was New York University’s Game Center.

  Its home at MetroTech Center in Brooklyn was still so new that after its director, Frank Lantz, let me in from the elevator vestibule, the card key for his office didn’t work. (A graduate student came to the rescue.) The common room’s game screens were still covered in plastic wrap; packing boxes were everywhere. Founded in 2008 as part of NYU’s Tisch School of the Arts, the Game Center offers two-year master of fine arts degrees in the design of games.

  Lantz is a legend in the gaming world. He co-founded Area/Code (acquired by Zynga in 2011), which created such Facebook games as CSI: Crime City and Power Planets (“control the fate of your own miniature planet. Construct buildings to make your inhabitants happy and . . . construct power sources to keep your civilization running”). He has created numerous iPhone games, including Drop7. Sharkrunners, which he made for the Discovery Channel’s 2007 Shark Week, lets players pretend to be marine biologists, interacting with living sharks fitted with GPS units that fed telemetry data to the game.

  Settling into the chair behind his nearly bare desk (most of his stuff was still in moving boxes), Lantz said that game design was finally being recognized as a legitimate discipline, especially as it incorporated ideas from fields as different as architecture and literature. “Most game designers have creative goals, and that drives them more than the goal of creating a game that people can’t stop playing,” he told me. But while many designers are driven by aesthetic and other lofty considerations, companies that sell games are more strongly invested in making a game compelling than ever before. Years ago a teenager would lay out $59.95 for Gran Turismo, and that was the last revenue Sony would see from him (until Gran Turismo 2 came out). It didn’t matter if players lost interest.

  In the 2000s, however, a new business model took hold: instead of paying up front, players acquired free-to-play games at no cost, often as a download to a mobile device, but then incurred “microcharges.” In Farmville, for instance, you can pay one dollar to magically “unwither” your crops (which are dying because you weren’t paying attention; damn homework) or speed up a harvest (allowing you to reap your vegetables before bedtime). Farmville keeps drawing players back to their simulated agricultural fields because it has a timer system: your crops die if you do not tend to them often enough. Many people hate to lose what they have gained, a phenomenon so strong psychologists have a name for it: loss aversion.

  Other games invite you to pay a dollar or two to leap over an obstacle, access a more exotic part of the game’s world, buy cooler clothes for your avatar, or purchase virtual food and drink for your virtual residents of CityVille. With the microcharge model, stickiness—being so attractive that players can’t put the game down—is all. “The commercial transaction is in the game,” Lantz said. “That’s causing a really deep discussion about game design, because some of the techniques feel manipulative. They’re not intended to make the game experience better or fulfill the vision of the designer, but to get you to run up these microcharges. I don’t think game developers are coldly applying behavioral psychology to guarantee players stick with their game. Very few game designers think they’re designing a compulsion engine. They want people to look back on their experience and say, ‘Hey, that was cool, that was fun.’ But yeah, they understand that this body of knowledge exists in psychology.”

  That’s putting it mildly. Whether by trial and error or by design, game developers have become frighteningly good at making games compelling. Something as simple as a list of high scores can do it, Lantz said, beckoning us to join and thereby quench our deep thirst for status by playing and playing and playing until we crack the top 100 (or our thumbs fall off). So can “nesting” goals. In the 1991 video game Civilization, for instance, players take turns making moves to “build an empire to stand the test of time,” as the blurb says. Starting in 4000 B.C., they assume the role of ruler of a particular empire-to-be, starting with a single warrior and people they can dispatch to form settlements. Through exploration, diplomacy, and warfare players advance their civilization by building cities, amassing knowledge (should you invent pottery or an alphabet first?), and developing the
surrounding land.

  “Why is it so compelling?” Lantz throws my question back to me. “Because it overlaps immediate goals, like dispatching settlers or getting your character successfully through a quest, with medium goals that you might achieve over the next three or four turns, like developing a city, with longer goals over the next ten to fifteen turns,” such as achieving the pinnacle of civilization. “There is a rhythm to the play, so as your mind comes to a resting place on an immediate goal you are also thinking about the next few turns ahead. Overlapping, or nesting, short-, medium-, and long-term goals is very compelling. In the real world, often we don’t even know what’s connected,” such as which short-term achievement would lead to something bigger later. The digital world of a video game offers certainty; A really does lead to B.

  Another way designers make games sticky is by offering instant gratification. “You do something and a character jumps,” said Lantz. “That can be very appealing when, in the real world, a lot of the buttons are broken.” That is, you press the button marked “study hard in high school” but it doesn’t get you into a selective college, as promised, or you press “go to college” but it doesn’t lead to a good job. In video games, the button performs as advertised, “which makes them deeply compelling,” he said.

  World of Warcraft, which induces compulsive play through its variable/intermittent reward structure, has an additional psychological hook, much like a good novel, detective story, or thriller. “It’s why you go back to War and Peace every night to read another chapter,” said Lantz. “You want to find out what happens next,” and not knowing makes you anxious in a way that drives the compulsion to play. “It’s a skill to go into it and come out the other end and be able to walk away,” to overcome the compulsion.