From Slot Machines to the Gaming Industry: Is Addiction the Only Way Forward

The original article was published in 2022 in Huxiu, a Chinese tech media outlet. The translation was done by ChatGPT, with my personal editing.

“The best way to predict the future is to invent it.”

Las Vegas casinos boast substantial profits, but counterintuitively, the majority of their revenue doesn’t come from the dealer-manned tables but from seemingly simple slot machines.

Slot machines contributed over 45% of the revenue for the U.S. gambling industry in 1980, and this share increased to 75% and 90% after 20 and 35 years, respectively. Slot machine revenue comes from gamblers who sit at the machines for long periods, continuously betting. Drawing gamblers to the slots, keeping them playing longer, and spending more money is no accident but the result of continuous optimization and improvement of the slot machines. Gamblers also enter a state of addictive fascination with the machines.

Slot machines seem far removed from our daily lives, but we may also fall into another kind of virtual maze: Internet products with similar optimization goals, such as social platforms relying on monetizing attention, encouraging longer use; or free online games based on in-app purchases, coaxing players to spend more.

So how did slot machines and the Internet build addictive mazes around gamblers and users, step by step? And, is this maze the only path for the Internet?

1. Building the Gambling Maze

New York University Professor Natasha Dow Schull spent over two decades in Las Vegas studying slot machines. She pointed out that the focus on the slot machine maze issue lies not only in investigating individuals prone to addiction but also in “how commercial gambling activities and environments create and even encourage conditions for consumer addiction.”

In “Addiction by Design,” Natasha discusses the construction process of the slot machine gambling maze in detail, from the casino environment, the appearance of slot machines, to their internal reward mechanisms.

(a) Casino Design

Upon entering a casino, gamblers encounter intricately designed decor. The curves of the ceilings and even the angles of the lights are carefully adjusted and optimized. The two most typical design styles in casinos are Friedman’s maze design and Thomas’s playground design.

Bill Friedman, once addicted to gambling and later spent 20 years studying casino environment design, proposed the Friedman Casino Design Principles in the 90s. These principles guide gamblers’ attention step by step to the machines through maze structures, carpet patterns, lighting brightness, music rhythm, and other environmental designs.

Casino designer Roger Thomas proposed The Playground Approach. He believed the focus of casino design should be to create a relaxing environment for gamblers. Thus, Thomas’s playgrounds are not meant to keep gamblers’ attention on the machines for long periods but to alleviate mental fatigue and restore focus after concentrated gambling, enabling longer gambling sessions.

Thomas also introduced French Baroque aesthetic design into casinos, believing that luxurious environments can “anesthetize” gamblers who have lost money, encouraging them to spend more. Under the guidance of the playground design rules, the Bellagio casino set the record for the largest profit from a single property in Las Vegas.

(b) Machine Design

When gamblers enter the casino and comfortably sit at a slot machine, the optimization focus shifts to the machine’s appearance. In “Addiction by Design,” casino management consultants mentioned that gamblers’ “productivity” could be increased by raising frequency, extending duration, and increasing expenditure.

Increasing the frequency of gambling means reducing unnecessary elements, allowing gamblers to spend more time on gambling. From mechanically operated levers limited by rotation speed to electronically responsive buttons, technological advancements have increased the frequency of bets per hour from 300 to 600-900.

The playground design mentioned earlier aims to make gamblers feel relaxed, which is also central to extending the duration of slot machine use: Gamblers should feel comfortable both physically and psychologically to play longer. This is reflected in every aspect of slot machine design: ergonomically designed seats, control panels easy for gamblers to operate, and rich sound from BOSE speaker systems.

To increase spending amounts, it’s crucial to make spending easier. From the early heavy coins and full coin slots that needed changing to technology that allows direct bill-to-game ticket conversion usable across machines, gamblers can spend more conveniently without waiting. Also, as money turns into digital numbers in the machine, gamblers’ perception of money weakens, making it easier for them to keep betting.

TITO Scams > Slot Machine Ticket in Ticket out Scams Slot machine game tickets, also known as Ticket-in/ticket-out (TITO) technology

(c) Reward Mechanism Design

The win-lose mechanism of slot machines seems simple: match three identical symbols to hit the jackpot. But why does this seemingly straightforward mechanism keep gamblers hooked and continuously betting?

We can start by looking at the Skinner Box theory proposed by psychologist B. F. Skinner. Skinner put pigeons in a cage where they could press a lever to get food. He observed that if food appeared randomly instead of consistently, the pigeons pressed the lever significantly more. In other words, intermittent rewards can reinforce specific behaviors. Slot machines also use a similar intermittent reward mechanism, reinforcing the betting behavior of gamblers.

Additionally, if you’ve ever participated in an e-commerce event inspired by slot machines, you might have noticed that the winning patterns often hover near the jackpot, making you feel just one step away from the prize. This is based on the “near miss” experience introduced to the gambling industry as early as 1905: The winning patterns fluctuate around the winning line, giving gamblers the feeling of “near miss,” increasing the likelihood of them continuing to play.

In the 1980s, casinos further increased the probability of near misses through virtual reel technology, and slot machines with virtual reels soon brought twice the profit of traditional slot machines to casinos. Skinner noted on near misses: “Almost hitting the jackpot increases the probability of continued play, even though this reinforcement mechanism costs the machine owner nothing.”

(d) Tracking Gamblers

Ultimately, gambling comes from gamblers’ behaviors, so the gambler tracking systems that learn from behavioral data and reinforce gambling behaviors also emerged.

The premise of tracking is to identify individual gamblers, which is why casinos promote membership schemes. Packaged as a convenient way to collect reward points, membership cards allow casinos to easily record gamblers’ betting frequencies, win-loss situations, rest times, food preferences, etc., gradually building a behavioral profile. Similarly, the information collection terms we “agree” to in apps are also painting and restoring our digital portraits.

With individual data, casinos can offer targeted marketing plans. For example, the “expected lifetime value” algorithm predicts a gambler’s total expenditure, with higher-value gamblers receiving priority service responses. Casinos can also calculate the loss threshold that would upset gamblers. When a slot machine detects a gambler approaching this “pain point,” the machine might present a “Lucky Ambassador” to offer the gambler dining coupons and other rewards to reverse the negative experience and encourage continued play.

In the above discussion, we can see how gambling practitioners, aiming to increase gambling duration and expenditure, have built a slot machine system encompassing the casino environment, machines, reward mechanisms, and gambler tracking, bringing huge profits to casinos.

2. Building the Virtual Maze

The highly profitable gambling industry is undoubtedly a typical maze, but under relatively strict regulation, most people can keep their distance from slot machines. Yet outside the gambling maze, many of us have experiences of inadvertently falling into virtual mazes in life, like office workers who can’t stop binge-watching Netflix late into the night even when they have early meetings the next day. Situations overriding rationality are not entirely due to users’ lack of self-control. Like the slot machine issue, we should consider how commercial activities create and even encourage conditions for user addiction.

In a podcast conversation with former Google design ethicist and Center for Humane Technology founder Tristan Harris, Natasha mentioned that casinos do not employ neuroscientists to analyze human psychological weaknesses to increase gamblers’ productivity. Instead, practitioners, from the perspective of increasing casino revenue, constantly test and iterate, gradually discovering the most effective designs to immerse gamblers.

Similarly, Internet products start from targeted data goals, continuously conducting A/B testing and product optimization. Through “reverse engineering,” they find designs that keep users staying longer and spending more. For example, in 2011, Google set “user watch time” as the core metric for YouTube. Various aspects of its ecosystem, including recommendation algorithms and content creators, were optimized around this goal, leading YouTube to reach a milestone of 1 billion hours of daily watch time in 2016.

Internet products reliant on user attention have indeed made our lives better to some extent. However, in their optimization process, the more significant consideration is economic efficiency (longer use, more spending). Investors and shareholders expect companies to deliver impressive numbers every financial quarter, bringing more profits. To achieve these goals, companies need more data experiments and faster product iterations to capture more of users’ entertainment time.

In the process of competing for consumer entertainment time and expenditure, Internet companies have launched various ingenious designs.

Social media platforms have features like like notifications and continuous interaction records with friends, attracting users to keep opening the app. Mechanisms such as infinite scrolling and auto-play continuously push new content, making users spend more time unwittingly.

Beyond social media, another kind of virtual maze in the Internet world is “free online games.” How do they manage to keep players play longer and spend more?

(a) Play Longer

To keep players engaged for longer periods, games need to sustain players’ attention over time. Compared to traditional media like books and television, games have a unique advantage in this regard.

Richard Lemarchand, who worked on the “Uncharted” series, divides our interaction with the external world into two phases: capturing attention and maintaining attention.

Games borrow aesthetics (beauty) and storytelling from movies to capture attention, and use a variety of goals and interactive feedback in gameplay to maintain it. Thus, while movies might capture people’s attention for three hours, games can do so for dozens or even hundreds of hours.

However, this doesn’t mean that games should continuously grip players’ attention. To ensure longer play sessions, designers mitigate “vigilance fatigue,” which is the innate human tendency to be constantly alert to changes in the environment. After periods of intense combat, designers usually allow players to rest or switch to non-combat game experiences. For instance, “World of Warcraft” offers double experience points for players resting in towns, along with other experiences like inventory management and sightseeing. If a game’s mechanics can alleviate vigilance fatigue, it encourages longer play sessions.

This approach is similar to Roger Thomas’s use of playground design principles in casinos: creating a relaxing environment for gamblers, allowing them to relieve mental fatigue after intense gambling sessions, and thereby prolonging their gambling time.

In addition to designs that maintain attention, some games deliberately create “time black holes.”

Harry Brignull describes deceptive user interface designs on websites and apps as “Dark Patterns.” These designs manipulate user behavior to meet business goals, such as sneaking insurance fees into shopping carts during flight bookings or covertly downloading apps when clicking “read more.” In gaming, “Dark Game Design” refers to designs that occur without player consent, serving the interests of game companies, potentially leading to negative player experiences. This includes time-based dark designs that make tasks take longer than expected or difficult to estimate.

(b) Spending More

The primary distinction between free online games and the previous generation of buy-to-play games lies in the Free to Play (F2P) model. Players can enjoy these games for free, with the option to pay for specific content such as characters, outfits, experience points, and items. Game developers maintain these games as a service (GaaS), continually providing new content and creating new revenue opportunities.

Within the player base that supports game revenue, players are categorized based on their spending: free, light spenders, medium spenders, and heavy spenders. The latter, often referred to as “Whales” internationally and as “Super R/Big R” domestically, are of particular interest to developers. As the founder of Kongregate shared, a mere 2% of these whale players contribute to 75% of their revenue, underscoring their critical importance to the free online game industry.

In the F2P model, developers skilled at retaining whales and encouraging them to spend more can reap substantial financial rewards. Currently, popular revenue-generating designs that lead to increased player spending include loot boxes, player tracking, and dark spending designs.

1. Loot Boxes

Loot boxes, or Gacha Systems, allow players to pay a certain amount to open a box and receive random rewards from a prize pool, which often includes rare items or characters with low probability of acquisition. Legally, loot boxes are not considered gambling, but they share similar psychological mechanisms with the Skinner box mentioned earlier.

Loot boxes reinforce the behavior of repeatedly opening boxes through intermittent rewards based on probability. The low probability of obtaining rare rewards means players may incur high costs to acquire desired items. As the developers of War Robots, which significantly boosted its revenue through loot boxes, stated: “Gacha is a system that hides the real price of things.”

2. Player Tracking

War Robots, which previously earned substantial revenue through gacha systems, shifted its main revenue model in 2017 to the more stable Wild Offer personalized marketing model.

The Wild Offer model tailors item and discount offerings to players based on their payment habits, play style, and past behavior. It provides a variety of items and prices, iteratively determining the types of items and maximum prices players are willing to accept, while introducing randomness to prevent players from predicting the lowest price the game is willing to offer.

Additionally, the game employs detailed operations that serve the entire lifecycle of a player. For instance, it analyzes payment pathways to identify the most profitable operational strategies or tailors different versions of the game for players with varying spending capacities. Alex Champandard, a former Rockstar employee, remarked that unregulated data collection in games, combined with targeted capabilities, is akin to “placing a miniature cigarette right at your fingertips when you’re most vulnerable.”

Player tracking and data experiments are not only applicable to refined marketing but also to the iteration and optimization of the overall game model.

Independent game developer Jonathan Blow analyzed common game tactics: storytelling, high-quality audiovisual effects, attainable next goals, and a sense of continuous progress. Using these tactics, developers can transform monotonous gameplay, like clicking items every six hours, into addictive games like FarmVille. Blow believes these tactics partly derive from existing games’ strengths and partly from detailed tracking of player behavior, using continual A/B testing to find designs that maximize company profits.

3. Dark Spending Designs

These include designs that cause players to spend unintentionally or more than expected. For example, the “Currency Confusion Dark Pattern” obscures the real value of money, with in-game items priced in a way that makes it difficult for players to calculate their real cost.

Dark designs in spending involve players unintentionally spending more money than expected or being unaware of the cost to achieve a goal. For instance, the Currency Confusion Dark Pattern obscures the real value of money, where $5 buys 400 game tokens, but an item costs 480 tokens, making it difficult for players to calculate the item’s real price; constantly updated recharge activity leaderboards; and the Pay to Win model, where players need to pay to remain competitive.

III. Is the Virtual Maze the Only Way Out for the Internet?

From slot machines to internet products, we see how products are designed around specific optimization goals to trap gamblers and users.

However, gambling and free online games fundamentally differ. The interaction mode of slot machines leaves little room for player play, firmly controlling game probabilities and directing player actions toward a predetermined outcome. In contrast, games as art can create “exciting, expansive, and empowering experiences” for players.

Amid the tension between profitability and user value, is there an alternative path for free online games beyond constructing addictive mazes?

(1) Updating Business Models

Tech companies’ revenues depend on users’ continuous engagement, making it difficult for them to internally promote addiction-limiting initiatives.

In the internet product space, platforms like Discord are experimenting with membership models, attempting to foster a healthier relationship between the platform and users without relying on ad revenue. Regarding free online games, is there a better business model that allows developers to profit while considering user welfare?

Consider the dilemma of past buy-to-play games, where development costs have increased more than tenfold compared to 20 years ago, but game prices haven’t risen accordingly. To maintain profits, developers need to sell more game copies to cover substantial development costs. Chen Xinghan mentioned that “Sky: Children of the Light” was initially planned as a buy-to-play game, but Apple pointed out that even at the most successful single-player game prices and sales volumes, the development costs couldn’t be recouped, leading Chen to switch to the free online game model. The advantage of free online games is that continuous content and payment point updates create more profit space, allowing developers to fund the development of better games.

But do free online games have to earn money solely through virtual mazes? Chen Xinghan suggests that free online games can profit not by exploiting human weaknesses like “light doses, high frequency, randomness” to create addiction, but by capturing the brighter aspects of human nature. This business model innovation requires developers’ vision and long-term commitment, as opposed to simpler “borrowing” strategies.

Former Leyou CEO Xu Yiran discussed another approach: reducing game production costs. For example, differentiating from the current company studio OGC (Organization Generated Content) production model, adopting the more flexible PUGC (Professional User Generated Content) model based on the film industry, using AI-assisted creation, removing intermediaries, and establishing a more direct subscription-based relationship with users.

The mentioned subscription model, popular in the film and music industries, allows companies to generate revenue through subscription fees and create high-quality content to attract audience subscriptions. From the user’s perspective, a small subscription fee provides access to as many works as possible, creating a positive content creation cycle. Industry insiders believe that under the current model, publishers providing funds might refuse game development if it doesn’t meet commercial calculations. The income and new distribution models brought by subscription models could revolutionize the content industry like Netflix, giving game developers the chance to explore new game genres that were previously unfeasible.

(2) The Responsibility of Creators

In “Addiction by Design,” Natasha analyzed the blurred line between the work of slot machine creators and inciting addictive behavior. They “cognitively disconnect their profit-driven designs from the potential harm to gamblers,” believing they’re merely extending the time and intensity of gambling.

In the internet context, some creators of virtual mazes may also overlook the impact of their products on society and users. Former Reddit CEO Yishan Wong shared his experience operating social platforms: “Social platforms hope users won’t endlessly argue over trivial matters. They don’t want to arbitrate user disputes, preferring to spend time launching more features.” Viewed from another angle, “Yishan may not have realized that the lines of code written by engineers behind social platforms and each new feature, to some extent, led to the deterioration of the online environment he criticized.”

Thus, the creators step by step construct the mazes.

As creators, perhaps they should initially recognize their power. Wang Shu, a Pritzker Architecture Prize-winning architect, has spoken about the significant social responsibility of architects, as demolition can easily destroy time and communities, and rebuilding can change people’s lives. In the early 1990s, amid a construction boom, massive demolitions and reconstructions occurred continuously, with some experts even proposing to flatten West Lake to build houses. Wang Shu, who advocates cautious use of architectural power, chose to retreat in this context, only emerging ten years later to create “modern architecture with a Chinese feel.”

Returning to the gaming industry, many developers aware of their power are attempting to answer and address deeper questions: What kind of games should be made?

For instance, David Cage, the developer of “Detroit: Become Human,” has tried to answer why we don’t share games with parents and grandparents like we do with movies, TV shows, and books. He believes it’s because previous games had limited themes and overly relied on operational skills. If games can offer roles that ordinary people can relate to, storylines that evoke complex emotions, and themes that provoke thought, they can leave an imprint in people’s minds like the best movies and books they’ve experienced.

Independent game developer Jonathan Blow has questioned why people say sculptures and paintings have changed their lives, but rarely describe games in the same way. Blow has criticized social games designed to be addictive and treat players as cash machines. He designs games considering players as intelligent individuals he’d like to know and communicate with, creating games that don’t waste players’ time. Further, he aims to create games that can stand alongside great art and truly inspire awe.

Conclusion

Paul Graham, in 2010, mentioned that we are in an era of accelerated addiction, where the rapid advancement of technology has created things that are more addictive for people: from television to Facebook, from checkers/card games to World of Warcraft/FarmVille. Now, 12 years later, the internet industry has birthed social products and free online games that are even more addictive, just as Paul Graham predicted. The trend of accelerated addiction is expected to intensify with the continued development of technology.

In light of this scenario, do internet products have better options than just pursuing addiction?

Certainly, there are emerging new business models beyond the attention economy, and game developers who are conscious of their responsibilities are making new attempts. Although their explorations might seem feeble, as Zhang Ning, the Vice President of Strategy at Zhihu, pointed out during a discussion on the dilemma of optimization:

“Of course, this gives the impression that the entire industry is in decline, but this sensation of deceleration is a natural law of the industry. We are reluctant to fully believe this and to give up completely because the new wave still seems too weak to withstand the wind, lacking enough energy to end this cycle. We know that day will come, after all, no one is immortal.”

What will that day look like? A mindset suitable for practitioners might be derived from a quote by technology pioneer Alan Kay: “The best way to predict the future is to invent it.”

The challenge for users is to coexist with existing virtual mazes, a topic worth exploring in depth in future discussions.