How to Make Healthy Games

[This is part two of a two part series. It is highly recommended that you read them in order. Part one, The Physiology of Gaming, can be found here.]

When I started selling virtual goods and making games, I wanted to understand why people played and spent in these games. I interviewed thousands of players and what they wanted tended to all point back to two main needs that were not being fulfilled in their real life: connection and power.

It appears Johann Hari has been studying depression first hand in the trenches. He's gone to a lot of work to understand why people are depressed, and how perhaps we have not been treating them in a way that leads to them getting better. What got my attention most was that he mentions the same two things—connection and power—in his book:

 Professor John Cacioppo of Chicago University: “Being acutely lonely is as stressful as being punched in the face by a stranger—and massively increases your risk of depression.”  

Professor Michael Chandler in Vancouver: “If a community feels it has no control over the big decisions affecting it, the suicide rate will shoot up.”

These were the two common themes I had observed over the previous 33 years. People were feeling lonely and powerless. In some cases they also were under direct physical threat that would not go away. This third condition is one we have to be mindful of but because this might be difficult for us to prevent through a computer game, I am focusing on connection and power as the…

Two Core Consumer Needs

People in our society increasingly are not having these two needs met. They feel disconnected and powerless. When they seek relief, either in another person, a book, a movie, or a game, fulfilling these two core needs are what we can best do to help them.

But the evidence is mounting that this is not what we are selling consumers. Instead we are stimulating them by increasing their dopamine or serotonin production to cover up their symptoms temporarily. When you use stimulation to try to cover up an underlying disease's symptoms the body eventually adapts and the symptoms return. To continue to suppress the symptoms you need ever increasing stimulus. By using this approach and not treating the underlying problem, their disease state gets worse over time and they need higher and higher doses of stimulation.

This is the mechanism of addiction formation and we are selling it.

One might counter that we have been doing that for over a hundred years and nothing has changed. But something has changed. Our level of science has advanced to the point that we understand human physiology so well that we have identified the systems the human body uses to temporarily adapt to extreme stresses. We have then taken that knowledge to develop products to massively and continuously stimulate our population to keep them going even as we increase the stresses that are causing disease. The reports of this society-wide disease progression are so numerous, that I am not going to flood this page with sources. If you are not already seeing daily reports of it in your news feeds, you certainly will be in coming months. The situation is breaking down rapidly.

~20% of Americans are on at least one drug to treat a psychiatric symptom. In the UK the rate of anti-depressant medication has doubled in ten years. If that rate continues it will approach 100% in less than two generations. The situation is serious enough that the United Nations concluded last year:

“the dominant biomedical narrative of depression” is based on “biased and selective use of research outcomes” that “must be abandoned”. We need to move from “focusing on ‘chemical imbalances’”, they said, to focusing more on “power imbalances”

Simultaneously the combined interactive media industries (social media, gaming, dating, etc.) have decided that Dr. Skinner was really onto something in 1930. Using variable rewards to stimulate dopamine release (and in turn consumers) is a Gift from God given to us to make us rich. Saying there has been a stampede to use this knowledge in our social media and games is no exaggeration. The use of these technologies in major social media platforms has reached 100% and in “games as a service” we will achieve 100% in just a few years if we are not already there.

I've been part of the problem. In 2012 I wrote Game Dosing which I published here on Gamasutra in 2013 during a period where America's National Public Radio interviewed me three times on related subjects. In my pride and desire to feel needed I took a leadership role in promoting the use of these technologies, teaching developers how to optimize their use of dopamine systems. While smugly describing myself as a force for good in the industry I probably did more harm than good for consumers. I regret this and I am sorry.

Here I attempt to deliver my first efforts towards a solution. I am calling it Physiologically Driven Design (PDD). The basis of PDD is using our knowledge of human physiology to design games to meet Core Consumer Needs. I am not taking credit for PDD. Many of us have been using physiology in our designs for years. PDD, when used to solve core consumer needs, can make consumers healthier and happier. It can also be (mis)used to harm people. I am asking you nicely not to do that, and providing clear examples of the difference so that if you choose to follow a different path you have that option.

As I present the ways PDD can be used beneficially or harmfully, I am not attempting to shame those who are doing the latter. I admit I have done this quite a bit in the past and I apologize. I am not better than you. Examples of harmful uses are here solely for the purpose of educating the community on how these techniques are harmful so that they can be mindful of their actions and consequences. The misuse of PDD is already so pervasive that if we waste time pointing fingers at each other then we are going to crash this train before slowing it down.

I call for an Amnesty and a collective action on the part of game developers and social media leaders to evaluate where we are at and make efforts to change direction towards meeting core consumer needs. We need to act collectively to solve this emerging threat, and the process of acting collectively will create a solution to the problems (loneliness and powerlessness) driving core consumer needs.

What follows is a detailed analysis of two classes of games that can be created using PDD: Power games and Connection games. Power games are well developed already, but Connection games are in their infancy.

Power Games

This discussion is broken down into two sections, “More Bad than Good” and “More Good than Bad”. A third section, a forewarning, discusses Stim Lock.

Stim Lock

Stimulation locking (“stim lock”) occurs when a person is repeatedly chemically stimulated intensely enough that they will self-stimulate until they collapse from exhaustion or some other physical malfunction, or a third party intervenes. This may sound like an extreme result but in situations where both the core consumer needs are being met simultaneously in an extraordinary way, any adult can become stim locked.

This most often occurs when a player is getting repeatedly dopamine stimulated while under a strong peer encouragement, usually in the form of spectators. A detailed description of what this looks like can be found in my “I'm Dying to Play” paper.

Children appear to be susceptible to stim lock at relatively low stimulation levels and without the need for peer pressure. This makes stim locking a popular method of “babysitting” as you can stim lock your child and then go off and do other things for hours and they will still be there when you come back. Extended stim lock causes the release of cortisol and endorphins, which further allow the body to maintain the stim lock to nearly super human durations.

While still limited, the existing body of research regarding the effects of early childhood exposure to high and/or repeated cortisol events clearly paints a picture where epigenic activation can cause permanent harmful dysfunction and loss of lifespan. The research is difficult for a lay person to understand, and painful for experts to read.

The prevailing wisdom in game development is that the more a consumer plays your game, the more likely they are to spend on your game. Following this logic, without a consideration of the consequences, leads developers to attempt to create Stim Lock mechanisms in their games. I would strongly encourage all game devs to intentionally place immersion breaks in their game designs, to allow players to comfortably stop playing long enough to use a rest room, get some food, or generally check in with their bodies in a healthy way.

In the case of content accessible to small children (this is as wide a category as it sounds), I would implore both game developers and parents/guardians to set up a system of soft and hard immersion breaks to allow the child to “cool off” and let their cortisol and endorphin levels return to normal. There is not enough research yet to set a definitive limit, but I would suggest a minimum one hour break after a two hour period of Stim Lock. If the child is non-responsive when you talk to them, or can't turn to look at you without you threatening them, it's fair to assume they are Stim Locked. Telling them you are using these standards to evaluate them may cause them to adapt by training to mimic appropriate responses in order to return to Stim Lock as soon as possible.

More Bad Than Good

Games that sell power to consumers are well established and relatively successful as they target a core consumer need. Current popular methods are, however, both low efficacy and also detrimental to the consumer.

“Pay to win” involves the selling of power directly to the player. Initially this is very attractive as this is exactly what many of them are seeking. They interpret this as a remedy to their insufficient power in real space. But when power is sold to one person, powerlessness is distributed (for free) to all other users. This has the net effect of disempowering players and making them unhappy. Their mental and physical condition can degrade over time.

As players become familiar with the initial illusion of power turning to powerlessness, they develop aversion to pay to win over time. Thus it is becoming more difficult to sell it. Players rapidly and collectively identify pay to win mechanisms and call for boycotts of your product. Thus peer pressure works against you, driving down sales and installs.

The selling of power-enhancing goods within “loot boxes” or similar randomized reward mechanisms is an attempt to stack dopamine with power delivery to enhance conversion. Selling dopamine to cover up the stress of unmet core consumer needs, while failing to effectively assist with the core consumer need, promotes addiction as described earlier.

Similarly, many games will sell increased power in a game environment and then raise the difficulty to take away the perceived feeling of power. This is typically done using a progressively advancing league system or otherwise matchmaking the player with stronger opponents. This is basically selling power and then punching the player in the face to restore their need to buy power again.

This cycle of reward---->face punch---->reward---->face punch causes degradation of the condition of the player, and addiction. The illusion is created that the misfortune of the player is their fault because they did not spend enough. If they were to just spend enough, they would be so powerful that the face punches would stop. Of course the face punches never stop and the player will be caught in an addiction loop until they realize this. Feelings of withdrawal and depression accompany the exit of the player from the system. This is painful to the consumer.

More Good Than Bad

Power is a comparative value. It cannot exist without weakness. Without that contrast, it does not exist. So how do you provide power to one player without taking it away from another?

Prior to my joining Wargaming this is generally what the matchmaker looked like:

If you had a tier 7 tank, the matchmaker would put you in a match rated at tier 6, 7, or 8. It would then populate the match with tanks within one level of that tier, with the limitation that tier 10 was the highest available tier, and tier 1 the lowest. The result is that in some battles you would end up “powerful” with a tank a tier level above the level of the match, and sometimes you ended up “weak” with a tank a tier level below the level of the match.

For power seeking players, being “up tier” in a match gave them exactly what they wanted. Being “down tier” made them feel disempowered. They would often quit these matches early in order to “try their luck” again. Of course the randomized nature of this matchmaking also has dopamine implications. So this is sort of a loot box you can keep hitting over and over again for free. This makes World of Tanks, and quitting World of Tanks matches, very popular.

I call this system a Rotating Power system. Rotating Power allows for power delivery using a randomized schedule that does not advantage one player over another long term so it is not perceived as disempowering by the players. Premium ammo is perceived as disempowering by players as it can give one player a paid advantage over another that is not skill dependent. Players experience disempowerment even if no one used premium ammo in a match, because they imagine that it was. The randomized nature of Rotating Power also contributes to dopamine delivery.

In 2013 when I was hired by Wargaming to direct the meta game design for all their incomplete non-Ukrainian (my safety could not be guaranteed in Kiev during the war) products, this was a very elegant design with a number of limitations that were reducing its efficacy.

I would like to go into more detail as to how to improve this system for the benefit of both consumers and developers, but will choose not to in order to avoid jeopardizing the entire contents of this paper by compromising any IP owned by Wargaming.

Revenue was generated in the Rotating Power system by charging for accelerated progression which gave the perception to the consumer of increasing power, even if their relative power never changed. There were of course other monetization channels but it will not benefit this discussion to disclose them.

Connection Games

More Bad Than Good

As I detailed in my 2016 talk in St. Petersburg, Russia on The Current State of Mobile Games, I described Tinder as the best mobile game in the market because it was aimed at building Connection with users that were demanding that, and because of its simplicity and good use of user generated content.

But the problem with Tinder is that while it could be used to facilitate Connection between users, this only happens if the users know how to Connect already. The research so far suggests they do not, even if previous generations might have known how to in the age range demographics that most use the product. Again this is suggestive that Disconnection is a growing problem in our society that is being exacerbated by our deployment of stimulation technologies.

While people are using pornography at ever younger ages (13 years old on average for boys), they are dating one year later than the previous generation. I would credit a substitution effect, similar to how people are substituting pets for human contact but ending up less happy and healthy. That delay in dating translates to an even larger deficit in dating experience and motivation. Thus the skills required to qualitatively connect are increasingly absent despite technology allowing it to be quantitatively easier.

As Tinder is sold to users seeking to fill the core Connection need, but does nothing to educate them on how to actually do that, it ends up treating the symptom not the cause like other addictive systems. As in all addictive systems, users tend to have a hard time not using them once they start, but disease rates increase in the using population.

In one study all measured metrics related to self esteem dropped for both genders using Tinder. Various States are reporting sexually transmitted disease rate increases of 30 to 70% per year since the wide adoption of Tinder in 2013. This is causing some STD rates to increase as much as 700% over that time, especially for women.

Tinder still could be “more good than bad” if it shifted its focus from unsupported communication to a focus more on guided communication and Connection education.

More Good Than Bad

During the same St. Petersburg talk I suggested that the first mobile game to break through the Connection Barrier was going to make a whole lot of money. Three months later Pokemon Go was released and the streets went crazy. Everywhere people were suddenly going outside and engaging in simultaneous entertainment. This was especially true in parks where the number of people walking more than doubled because of Pokemon Go.

People who were chasing the same Pokemon or who kept lapping each other over and over started to Connect in a way that was not focused on appearance or sexual objectification. People who may have engaged in social aggression in different circumstances were treated as “In-Tribe” and treated with respect. Overall exercise levels increased significantly, especially for those previously considered “sedentary”. I had really never seen anything quite like it other than perhaps when the Olympics were hosted in my city.

It must be said that there appears to have been little or no attempt to optimize the game for promoting Connection, and the social and progression systems were limited. There is a lot of room for growth in this genre of game.

My informal surveys and observations of players indicate that the #2 most demanded systems in games are those that promote social interaction. The #1 most demanded systems are those that promote social progression/rank/value. As I refer to in my Tinder example, social progression involves more guidance from developers compared to just allowing social interaction. For a detailed description of what a social progression system looks like in an AR game, I would recommend going all the way back to "The Driving Game" in my 2010 AR prediction paper.

It is understood that game dev studios are mostly populated by people who do not make social interaction their professional specialty, but in order for us to make large strides in creating Connection games we need to begin adding these sorts of people to our studios.

By meeting core consumer needs directly, we get cheaper installs, more organic installs, and much higher LTV's. Consumers are happy with us, as are regulators and the media. We in the industry get to live happier and healthier lives instead of living like we are on the ropes constantly (which we are).