Addiction Machines: How Slots are Designed for Compulsive Play

Your player card, please.

The image of the compulsive gambler has traditionally been the male poker player, drink in hand, recklessly betting the night away. Slot machines? Those were for amateurs, the out-of-towners, the meek and the mild. But that irritating clang and buzz coming from over the card player’s shoulder is not just the sound of new money—it’s the sound of a new technology tuned to a ruthless edge.

Digital slots and poker machines have become the new games of choice for pathological gamblers. In 1999, Harvard addiction researcher Howard Shaffer predicted that, “as smoking crack cocaine changed the cocaine experience, I think electronics is going to change the way gambling is experienced.”

Modern gambling machines drive the casino gambling industry, and generate far more revenue than “table” gambling. Because of the manner in which they “facilitate the dissociative process,” as one psychologist puts it, excessive gambling is built into the design and structural characteristics of the technology itself. One physician has even suggested that machine gambling produces a trance state by closely matching human breathing patterns with its “basal slot play rate.” We don’t have to wait for the Singularity to observe the merging of man and machine.

By 2000, digital gambling machines were generating twice the revenues of “live” games. Today, the modern slot machine “drives the industry,” said the president of the American Gaming Association in Natasha Dow Schull’s book, Addiction by Design: Machine Gambling in Las Vegas. They are allowed, in one form or another, in at least 41 states. Journalist Marc Cooper, who covered Las Vegas in his book, The Last Honest Place in America, said in 2005: “The new generation of gambling machines has, predictably, produced a new generation of gambling addicts: not players who thrive on the adrenaline rush of a high-wager roll of the dice or turn of a card but, rather, zoned-out ‘escape’ players who yearn for the smooth numbness produced by the endlessly spinning reels.”

“A gaming machine is a very fast, money-eating device,” according to a spokesperson for Bally. “The play should take no longer than three and a half seconds per game.” Gambling engineers attempt to fine-tune the “capacitive logic of haptics,” by, for example, designing chairs that tingle and pulse in response to events in the game. The ideal is to achieve an “embodied relation,” in which a gambling machine becomes an extension of the gambler’s own cognitive capacities and spatial skills. Professor Schull of MIT’s Program in Science, Technology, and Society sees a digital gambling machine as “an interactive force that powerfully exerts its program for ‘player extinction’ and in so doing constrains the possible outcomes of play.”

At the simplest level, gambling machines function as Skinner boxes for human rats. Intermittent reinforcement, as psychologists showed long ago, is an effective way of shaping behavior. “If the number of responses required to receive a stimulus varies,” writes biologist Jason Goldman at his Scientific American blog, The Thoughtful Animal,“then you are using a variable ratio schedule. The best example for this is a slot machine, which has a fixed probability of delivering a reward over time, but a variable number of pulls between rewards. It is no wonder that variable ratio reinforcement schedules are the most effective for quickly establishing and maintaining a desired behavior.”

Casinos were early adopters of biometric surveillance methods, and now have the capability of offloading much of this work to distributed digital devices like player loyalty cards. Theoretically, machines could achieve and maintain an active feedback loop with each gambler. The machine could compile data on betting patterns, recent outcomes, time of day, and rhythm of play. The machine would have the ability to “automatically alter the volatility level for gaming events to match the general player preferences at specific times,” in the words of one patent application. The longer you play, the more the machine would understand your style, and offer more of what will keep your ass in the seat.

The advent of poker playing machines brought in more players aiming for time-on-device rather than supersized jackpots. Poker machines gave out some kind of reward on 45% of plays—the perfect intermittent reward, if you asked Pavlov. And there was a razor-thin component of skill to the gambling machines. But all of the trademark features of addictive play are present in Draw Poker machines as well.

Here’s what casinos currently depend on to keep compulsive gamblers at their machines:

—Faster play. The key introduction was the virtual reel, which allowed play to take place faster than mechanical reels could spin. The use of touch screens is on the upswing, to further increase play speed. And the “BET MAX” button is always nearby.

—Longer “time-on-device”. One industry expert said: “If the chase lights on the slot signs are running too fast, they make people nervous; if they run too slow, they put them to sleep. If the machine sound is too loud, it hurts the player’s ears; if it’s not loud enough, the energy level of the room suffers.”

—Upping the ante. So that players can spend their money more easily, designers have engineered bill acceptors, digital credit counters, loyalty program cards, and other ways to reduce the actual handling of coins and cash and eliminate physical payouts at the device site. But what about the continued popularity of the nickel slot? “A nickel game isn’t a nickel game,” said one game developer, “when you’re betting ninety nickels at a time."

—Disguising the odds. The wonders of the random number generator are perfectly disguised in digital machines. Virtual reel mapping, or “weighted reels,” is credited to mathematician Inge Telnaes. It describes a system in which there is no logical correlation between the actual number of choices seen by the player and the number of stops contained on the virtual reel. Blank reel spaces help increase the confusion, while a secondary mapping program translates the virtual stops selected by the RNG microchips into the actual stops visible to players onscreen.

Something has to give, since recent research seems to show that machine gambling pushes gamblers into an addictive relationship with gambling at a rate three times faster than gamblers who stick to live table games. Back in the skunk works, where the machines are designed and manufactured by companies like IGT, the nation’s leading maker of gambling machines, weakening the hold of the machines would mean limiting near-miss effects, coming clean about virtual reel mapping, and placing restrictions on building ATM access into upcoming models. But maybe none of that will matter. As a software designer who moved from slot machines to games for kids told Professor Schull, “it wasn’t that big of a leap, in fact it was very similar. That really struck me. I saw it as appealing to the same part of the mind, a really simplistic instinct for distraction. Similar types of customers—toddlers and gamblers.”

Photo credit: http://www.all-slot-machines.com/

That Pesky Gambling Question

The DSM-V is set to label problem gambling an addiction.

Nobody has ever bet enough on the winning horse.
— Unknown wise person

I used to gamble. Back when I did, I was also an active alcoholic and a chain smoker. Camel filters, if you’re wondering. And we had a running joke, my wife and I, although the humor leaked out of it for her pretty quickly. We would breach the doors of the gambling palace, and plunge into the dark, icy interior of a casino at Las Vegas or Tahoe, and stand on the edge of the gaming room, taking it all in for a moment. “Ah,” I would say, surveying the roomful of cigarette smokers with drinks in their hands, making bets or hitting buttons at one o’clock in the morning, “my kind of people.”

Gambling can be defined as an activity in which something of value is put at risk in a situation where the outcome is uncertain. That’s really all there is to it. Howard J. Shaffer and Ryan Martin, whose article in the Annual Review of Clinical Psychology, “Disordered Gambling: Etiology, Trajectory and Clinical Considerations,” takes on all the interesting questions about gambling as an addictive disease, have chosen to favor the term “disordered gambling.” Just as there are divisions between alcoholic drinking, heavy drinking, and social drinking, there are similar states we can call pathological gambling, excessive gambling, and social gambling. On the problematic end of the scale, pathological or problem gambling has proven to be “a more complex and unstable disorder than originally and traditionally thought.” No kidding. Once the neurophysiology of the gambling state of mind came under scrutiny, the parallels with addiction cropped up so rapidly that investigators have been hard pressed to come up with suitable explanations for it all.

The new DSM-V proposes to shift pathological gambling from “impulse control disorder” to the new category of “addiction and related disorders.” So it’s a good time to rethink the question along with the psychiatric community.

In the traditional view, pathological gambling was a matter of exposure to the proper stimuli—it could happen to anyone. But as more and more gambling outlets and opportunities bloomed in Nevada, on reservations and riverboats, and in convenience stores, that view began to fall out of favor, because a funny thing happened. According to Shaffer and Martin, the prevalence of pathological gambling has remained stable over the past 35 years, even as opportunities to gamble have exploded. The lifetime prevalence rate of pathological gambling in the U.S. in the mid-1970s was 0.7%, say the authors, and by 2005, U.S. lifetime rates had actually fallen slightly, to 0.6% or less. Where was the concomitant explosion in the number of pathological gamblers?

Next, researchers got technical, wondering whether certain types of gambling, or certain types of gambling machines, were more “addictive” than others. They quickly ran into the same kind of trouble substance abuse researchers got into when they first tried ranking drugs according to strict hierarchies of addictiveness. In so doing, the staggering metabolic diversity of the human animal got lost in the shuffle, as did the fact that my metabolism and my behavior when taking drugs, or knocking one back, or losing money in a casino, is going to be different from yours.

Then came Internet gambling. In 1996, the first online casino to accept real money began operation, and by 2001, there were more than a thousand. Previously, researchers had to rely mostly on the time-honored but not always accurate system of self-reporting. If you ask people why they gamble, they tend to answer that they do it for the fun, the excitement, the challenge, and the chance to win some money. But what gamblers can’t recall very well are specific patterns of play over time that might benefit researchers. For example, in a 2009 study in which observers actually watched gamblers gambling, one long-standing observation from the self-report literature—gamblers become more liberal risk takers as they approach the end of a gambling session in a behavior called “chasing”—didn’t prove out. When researchers watched actual gamblers in action on the Internet, or playing lotteries, they found that problem gamblers in fact began betting more conservatively as they approached the end of their gambling, the authors write.

Another approach is to consider risk factors of all kinds—neurobiological, psychological, and social—and look for similarities between those for substance addiction, and those for “activity-based expressions of addiction.” The “syndrome model,” or what I usually call the umbrella model, derives from neurobiological research suggesting that “addictive disorders might not be independent: each outwardly unique addiction disorder might be a distinctive expression of the same underlying addiction syndrome…. The specific objects of addiction play a less central role in the development of addiction than previously thought….”

 All of this opens the door to some informed speculation about a broader range of disorders that may lurk beneath the umbrella of the addictive disease concept. Among these are such conditions as body dysmorphic disorder, bulimia, depression, and extreme PMS, which are all found more often in addict populations. In addition, impulsivity and low “harm avoidance” are behavioral traits often found in association with addiction. Shaffer and Martin call these “shadow syndromes,” and they are found to be associated with BOTH substance and behavioral addictions.

But what, exactly, is the high in gambling? The researchers believe that, “similar to ingesting stimulants, there is evidence that gambling is associated with autonomic arousal including elevated blood pressure, heart rate, and mood.” That's not very specific, and could also describe a craving for teddy bears. But recently, fascinating evidence of neurobiological influences on gambling arose when Parkinson’s’ patients on strong dopamine agonist treatments, with no history of gambling whatsoever, began behaving for all the world like pathological gamblers. I cannot imagine a better suggestion of neurogenetic involvement than this unexpected finding. Previous research had shown that dopamine-active drugs were capable of increasing the incidence of other addictions, too. Shaffer and Martin list compulsive eating, compulsive sexual behaviors, and compulsive shopping as activities that can also be boosted with dopamine agonists or diminished by lowering dopamine activity. 

And it does not strike me as surprising to learn that, yes, gambling problems tend to run in families, or that twins studies show that pathological gambling is higher among twins born to pathological gamblers than twins born to non-gamblers. It is the same evidence anyone can bring forth to bolster the argument for neurobiological influences on alcoholism, heroin addiction, and the like. “In sum” Shaffer and Martin conclude, “genetic influences might not determine the development of specific expression of addiction; however, genetics does influence the risk of addiction in general.”

If all of this is true, we should expect to see a corresponding connection between pathological gambling and substance abuse disorders. Some degree of overlap would be good evidence. And we have it in spades. Pathological gamblers are five and a half times more likely to have suffered from a substance abuse disorder. “75% of PGs (pathological gamblers) have had an alcohol disorder, 38% have had a drug use disorder and 60% have had nicotine dependence.” Also, “PGs are 4 times more likely than non-PGs to experience a mood disorder in their lifetime….”

So when I used to stand on the edge of the casino floor, as an alcoholic and a nicotine addict, casually calling those gamblers my kind of people, I think I was more right than I ever could have guessed.

Does all this mean that playing games on the Internet is an addictive behavior if making bets with real money is involved? The authors crunched the studies on that question, and discovered that maybe 1% of the Internet population has used the Internet for gambling purposes, and that “the case of Internet gambling provides little evidence that exposure is the primary driving force behind the prevalence and intensity of gambling…. The relationship between the extent of gambling ‘involvement’ is a better predictor of disordered gambling than any particular game that people play.”

By gambling involvement, the authors mean the number of different kinds of games a gambler plays. The more he or she plays, the more likely they are, or are likely to become, problem gamblers. However, it’s not hard to see where the online notion came from. Gambling folklore has always held that addiction is more of a risk with electronic gambling devices like slot machines and 5-Card Draw machines, than with traditional table games like roulette and craps. But the authors don’t find any convincing clinical evidence for this assertion at all. Internet gambling isn’t more addictive, and doesn’t confer any extra risk on people participating in other forms of gambling.

Here is a list of potential gambling behaviors that Shaffer and Martin believe might be risk factors to look out for in the development of problem gambling, with my additions in parentheses showing the connections to other drug addictions.

--Betting Intensity: how many bets per day. (With alcohol, how many drinks.)

--Gambling Frequency: number of gambling days (Number of drinking days.)

--Gambling Trajectory: tendency to increase the amount of wagered money. (Tolerance, in the case of drugs and alcohol.)

--Gambling Variability: deviation from consistent gambling pattern. (Inability to predict duration or outcome of drinking event.)

But if we list gambling under Addiction and Related Disorders, must we list all the possible variations on the theme—shopping and sex and all the rest, even though the picture is still fuzzy? No, the authors argue, we don’t. Not right this minute. But gambling is ready to join the roster. Shaffer and Martin would like to see their syndrome model of addiction used to identify “core features of addiction and then illustrate these with substance-related and behavioral expressions of this diagnostic class. Conceptualizing addiction this way avoids the incorrect view that the object causes the addiction and shifts the diagnostic focus toward patient needs.”

Shaffer HJ, & Martin R (2011). Disordered gambling: etiology, trajectory, and clinical considerations. Annual review of clinical psychology, 7, 483-510 PMID: 21219194

Graphics Credit: http://gamblinghelp.org/pages/resources/toolkits.php