Games, Play, and Joy: Part 6
The promises and perils of simulations.
(Wow, um, hi! I published my last essay way back in 2021, but at least not much has changed in the political or digital landscape since then, right? Right???)
This is Part 6 of my series Games, Play, and Joy, which is based on a class I taught in 2019 at NYU’s ITP (as part of my residency there), aimed at games-curious artists. As ever, this essay series doesn’t represent the opinions of my employers or whatever — just me.
Hope you enjoy.
Theory
- Part 1: Creating joyful possibilities with games.
- Part 2: Using rules as excuses and catalysts.
- Part 3: Playing with systems of power.
Tactics
- Part 4: The art of finding the fun.
- Part 5: Adding gamefeel and juice.
- Part 6: The promises and perils of simulations. (you are here)
- Part 7: The joy of feeling seen.
Your Turn
- Part 8: Epilogue.
In Part 3, we talked about how artists use games as spaces to experiment with changing the rules and norms of everyday life. As Doug Wilson says in Designing for the Pleasures of Disputation, games can function as an “alibi — as a culturally recognized excuse to ‘act out’ and play the fool[,]” giving us a safe space to slip out of the clutches of the systems of power around us, experiment with new values, and inhabit new ways of being.
Another way to experiment with these structures is to turn those systems into games themselves — simulating them so that we can see what it’s like to change them. Indeed, designers use games to simulate all sorts of real-world phenomena, whether on a grand scale (designing cities) or a micro scale (navigating social interactions). A common hope expressed by folks who make those games is that giving players agency over these systems allows them to understand the root causes of systemic problems; empathize with those facing systemic problems; and naturally discover better solutions to the real-world problems they might face around them.
The idea of changing hearts and minds just by writing code is very seductive (and it’s pervasive — you’ll find this attitude in larger tech-land too, with plenty of capital to back it up). One reason I cover it is because I personally find that a lot of new game designers flock to it, especially around thesis-time. They find out about ‘serious games’ and immediately decide that they will change the world by simulating real-life systems, to show players how screwed up things are and what the real solution is. Noble as this intention often is, it can backfire badly, sometimes reinforcing the exact societal problem or misconception the designer was trying to solve in the first place.
Let’s cover this territory together, so that you can have a better critical eye for the perks and pitfalls of real-world simulations you might be tempted to make.
Since we take nothing for granted in this class, let’s step back and ask: what makes complex simulations run in the first place? What makes it possible to simulate a human’s life, a city’s functioning, a world’s warring empires?
We can think about simulations as being like stacks of rules — a big pile of intersecting if-then statements, triggering each other like falling dominoes. If we burn too many fossil fuels, then Co2 levels rise; if Co2 levels rise, then our simulated planet will heat up.
If we’re trying to represent a complex system accurately, we usually need a very detailed ruleset to represent it. While the if-then chain above represents one environmental phenomenon, it is certainly not a comprehensive model of what contributes to our climate.
The functions and parameters governing these systems can easily be more complex than the average person can execute by hand. Even in the oldest versions of The Sims, a Sim’s mood at any given moment is based on the output of a set of interconnected rules governing personality, recent actions, prior disposition, physical conditions, physical surroundings, social relationships, and more. Calculating a snapshot of a Sims game by hand would take even the cleverest person a lot of time (and paper).
Lucky us: computers are extremely good at rapidly executing these functions. The more processing power our computers have, the more detail we can add to our system while still being able to run it in real time. Computers allow us to simulate exponentially more complex systems and relationships than we could by hand.¹
Computers are so good at executing these instructions — sometimes called ‘processes’ — that some argue they unlock a fundamentally new form of expression. From Janet Murray’s blog (which is in turn from her 2011 book, Inventing the Medium):
The procedural affordance has created a new representational strategy, as powerful as spoken language or recorded moving images: the simulation of real and hypothetical worlds as complex systems of parameterized objects and behaviors.
With procedurality, we get something called procedural representation, defined by Ian Bogost in Persuasive Games as “explaining processes with other processes.” That is, we can layer systems on top of systems to describe the larger systems that emerge from their interaction.
Let’s think about what systems we can describe with procedural representation.
Material phenomena
You’re probably already familiar with games that simulate really tangible, material things: flight sims, rocket-building sims, truck-driving sims. These simulation games often have an explicit or implied goal that is quite concrete: deliver the goods, build the PC, get to the moon — one that mimics real-world success as closely as possible.
Not all of these simulations are trying to be 1:1 mirrors of Earth, of course. In Kerbal Space Program, your astronauts are aliens, not humans, and their solar system isn’t the same as ours. But Charlie Hall at Polygon points out that the physics system underlying the game is so sophisticated and realistic that actual NASA employees play it regularly. Hall quotes Doug Ellison at NASA’s JPL:
“I knew KSP was something special when I watched a young kid — probably less than 8 years old — playing KSP and using words like apogee, perigee, prograde, retrograde, delta-v; the lexicon of orbital mechanics. To the layperson orbital mechanics is a counter-intuitive world of energy, thrust, velocity, altitude that this kid — just by playing Kerbal — had managed to get his head around.”
Of course, players can goof off in even the most serious simulations — one player might work super hard to create the best space program possible, while another carelessly crams parts together to laugh at the resulting disaster. But I would argue that the latter play style also depends on the accuracy of the simulation: part of the fun of setting up a disaster is the surprise of seeing the unexpected ways your invention will fail. The system’s complexity results in surprising emergent behavior: even with the best (or worst) laid plans, you can never be 100% sure how something will really shake out.²
Social phenomena
Simulations can also attempt to model social phenomena, like how a city functions. Of course, these sorts of simulations are concerned with material phenomena as well (nearby pollution might lower property values), but there is a bigger focus on modeling fuzzier social phenomena (lowered property values might piss off your local citizens).
A classic example of this is the Sims series, especially Sim City. The implicit goal of this game is to balance various factors to create a healthy and thriving city: Ava Kofman in Les Simerables quotes the manual, which says the goal is “to design, manage, and maintain the city of your dreams.” (But you can’t really lose, either — you can make your own hell city just for fun, the same way you can condemn a Sim to death by ladder-less pool.) Ultimately, the game is about simulating — and responding to — emergent social phenomena relating to city design, citizen happiness, and government choice.
Like material simulations, social simulations don’t have to be set in the real world to represent systems seen in real life.
In Animal Crossing, the player lives in a village inhabited by anthropomorphic animals. You build out your little house, improve your town, and become friends with fellow townspeople through trade, giving gifts, social calls, and so on. It’s generally classified as a social simulation game.
But to some designers, what Animal Crossing really simulates is capitalism. In his paper The Rhetoric of Videogames, Bogost cites the character of Tom Nook — a business-owning tanuki who offers the player a house at the cost of a hefty mortgage — as “a condensation of the corporate bourgeoisie”:
[T]he player participates in a full consumer regimen: he pays off debt, buys and sells goods. Tom Nook buys the player’s goods, which he converts to wealth. As the player pays off debt and upgrades his home to store more goods, he sees Tom Nook convert that wealth into increased commercial leverage — one’s own debt makes the bank rich. Tom Nook then leverages that wealth to draw more capital out of the player, whose resources remain effectively constant. While the player spends more, Nook makes more. By condensing all of the environment’s financial transactions into one flow between the player and Tom Nook, the game models the redistribution of wealth in a way even young children like my five-year-old can understand.
[…]
Animal Crossing is a game about everyday life in a small town. It is a game about customizing and caring for an environment. It is a game about making friends and about collecting insects. But Animal Crossing is also a game about long-term debt. It is a game about the repetition of mundane work necessary to support contemporary material property ideals. It is a game about the bittersweet consequences of acquiring goods and keeping up with the Joneses. Animal Crossing accomplishes this feat not through moralistic regulation, but by creating a model of commerce and debt in which the player can experience and discover such consequences.
I will allow all of you to decide whether you believe Tom Nook is a rotten capitalist for getting you into debt, versus a working-class hero for not charging any interest. (You are also welcome to think the entire debate is stupid and just keep catching bugs.) What’s worth taking out of this example is how Animal Crossing’s fictional town simulates a simplified, but familiar, system of debt and consumption. And inside each function to evaluate whether you’re making friends, designing a ‘good’ home, and ‘improving’ your town is some encoded belief about what counts as friendship; about what counts as a good home; about what counts as a good town.
If games encode and convey beliefs, then can we use games to experiment with our own beliefs? This idea has long excited game designers: the belief that we can more deeply understand the world through playing with simulated systems. In his article “Persuasive Games: The Birth and Death of the Election Game,” Bogost says:
When we make video games, we construct simulated worlds in which different rules apply.
To play games involves taking on roles in those worlds, making decisions within the constraints they impose, and then forming judgments about living in them.
Video games can synthesize the raw materials of civic life and help us pose the fundamental political question, What should be the rules by which we live?
Imagine: if we can put the players in the driver’s seat of an experience and allow them to explore, maybe they’ll come to better conclusions than before. If they can really feel how things interact — how Action A leads all the way to Consequence Z — then maybe they will leave the game as more informed, better prepared, empathetic humans.
Furthermore, we can even use game systems to persuade people of ideas. This is the notion of ‘procedural rhetoric,’ which Bogost defines in Persuasive Games (the book) as “a general name for the practice of authoring arguments through processes.” He continues:
Following the classical model, procedural rhetoric entails persuasion — to change opinion or action. […] Procedural rhetorics afford a new and promising way to make claims about how things work.
The idea of using games to make political statements predates this term, with Chris Crawford’s 1985 game Balance of Power being an excellent example. Balance of Power is set during the Cold War, with the player taking on either the role of POTUS or the General Secretary of the Soviet Union. The goal is to improve the country’s standing relative to your opponent’s. The world is divided into 62 countries, all described in a massive database. During each yearly turn, various events happen that you can choose to respond to through various means: diplomacy, military action, economic consequences, or simply ignoring them. Overly-aggressive behavior may lead to escalating threats from your nuclear opponent, and it’s entirely possible to call too many bluffs and cause nuclear war.
In Balance of Power, inciting nuclear war leads to an automatic game over. There is no coming back from nuclear war; no possibility where the ends could justify the means. Crawford refuses to even render the outcome, predicting (probably correctly) that any attempt to horrify the player with gruesomeness visions of nuclear war would probably act as an incentive or reward instead. In a 1985 interview with the New York Times, Crawford said, “[m]y hope is that the game will bring about a more mature set of attitudes about the world for those who play it.” The procedural rhetoric of the game re: nuclear warfare is pretty straightforward.
Compare the auto-game-over in Balance of Power to a game like Fallout 76, in which players are explicitly incentivized to nuke each other. From Matthew Gault at Vice:
In what the developers call “the nuke loop,” Fallout 76 players can hunt down nuclear codes and launch ICBMs at the map. Fallout 76 gives players many incentives to do this — it’s a great way to wipe out a pack of annoying players and it also generates some end game content. Lands irradiated by a fresh nuclear missile will spawn enemies, gear, and plants that don’t otherwise appear in Fallout 76’s version of West Virginia.
Of course, representing something doesn’t necessarily mean trivializing it. The Fallout series is known for its satirical bent, and Gault mentions embracing such launches as “a way to deal with the existential horror of nuclear war.” But it seems fair to say that the procedural rhetoric of Balance of Power versus Fallout 76 is very different.
You can see the hope of ‘learning through playing’ in Will Wright’s SimCity, summarized by Kofman in Les Simerables:
Wright has stated that his goal in designing the game was to create a space of possibility open-ended enough for the player to experiment in, “a problem landscape” large enough to generate infinite solutions. According to Wright, the game encourages utopian thinking: “So when you start SimCity, one of the most interesting things that happens is you have to decide, ‘What do I want to make? Do I want to make the biggest possible city, or the city with the happiest residents, or most parks, or the lowest crime?’ Every time you have to idealize it in your head, ‘What does the ideal city mean to me?’”
There is an implicit belief here that, if the sandbox is open enough, a person could experiment with all sorts of city-running tactics and come out with an embodied sense of what is necessary to have a healthy city — which could then inform their civic behaviors in real life. If your city in SimCity does markedly better when you spring for more parks, how might that activate your real-life voting if a candidate wants to fund more green spaces?
This is a very nice, very inspiring idea. However, it is not without caveats and consequences.
The thing is (and as Bogost points out throughout his work), procedural rhetoric is political. It’s kind of in the name — after all, rhetoric is the art of persuasion. The notion of ‘persuasive games’ should make us ask: who we are trying to persuade, and of what are we trying to persuade them?
We should be paying attention to these messages and values even (perhaps especially) when a game claims to be apolitical. What developers choose to put in or leave out matters. Consider like how Nintendo explicitly refused to allow non-heterosexual relationships in the social simulator Tomodachi Life. Or consider a simulation game that gives the player agency, when in real life, a person in that situation would have very little agency at all.
You might say, “well, you’re being too picky. Of course a game has to leave things out — we can’t model everything. And of course we need to have at least a little bit of agency — otherwise it wouldn’t be a game at all!”³
And hey, you’re not wrong. After all, as Chris Crawford says, in the Balance of Power manual:
“Any game deliberately simplifies reality to emphasize some aspect of the world. […] The painter does not slavishly copy an image but instead emphasizes those things that are important about that image while eliminating its unimportant aspects.”
I actually very much agree with this. All I want to point out that choosing what to leave out is also meaningful. Who gets to say what’s important or unimportant? What are the consequences of putting something in, or leaving it out? What views and behaviors are hardcoded as default? Is the underlying system still seen or marketed as being totally unbiased?
These questions are not mean to invalidate the games we’re talking about: all games simplify reality in some way, and most objects reflect — to some degree — the values and assumptions of their creators. (This essay series certainly does for me!)
Rather, these questions are meant to spur your media analysis skills when it comes to simulations and procedural rhetoric. And part of the reason that is important is because, in your career, you may be offered good money to make simulation games, or interactive vehicles for procedural rhetoric. Even when those causes align with your own beliefs, you still need to think about them carefully, because those vehicles can blow up in your face in ways you’d never otherwise expect.
Let’s go back to the SimCity series.
Earlier in this essay, I cited a Will Wright quote about a problem space big enough to “generate infinite solutions” (emphasis mine). Infinite! It’s an exciting idea: you define your ideal city, and you can experiment with endless strategies to get there. The world is your oyster, and you have all the tactics in the world to experiment with!
Sort of.
As Kofman in Les Simerables points out, Sim City’s procedures, algorithms, and models take certain political ideas for granted. For example:
An increase in the number of police stations, for instance, always correlates to a decrease in criminal activity; the game’s code directly relates crime to land value, population density, and police stations. Adding police stations isn’t optional, it’s the law.
The notion that more cops always means less crime is complex and hotly debated in real life, and its encoding in the game cuts off one avenue of the “player’s exploration of utopian possibility.” After all, how much success could you have with a more abolitionist experiment if the game’s code directly links more cops to less crime?
The article points out that the same goes for taxes, quoting the manual:
Keep taxes too high for too long, and the residents may leave your town in droves. Additionally, high-wealth Sims are more averse to high taxes than low- and medium-wealth Sims.
Now, I’m not an international tax expert, but as an American, this feels like a very American point of view on taxes. Consider Denmark, a country with one of the highest tax rates in the world, but where US News points out that “almost nine out of 10 Danish people happily pay their taxes to some or a high degree.” The difference, says the author, is the cultural “awareness of the fact that the welfare model turns our collective wealth into well-being.” It continues:
We are not paying taxes. We are investing in our society. We are purchasing quality of life.
The key to understanding the high levels of happiness in Denmark is the welfare model’s ability to reduce risks, uncertainties and anxieties among its citizens and prevent extreme unhappiness. […]
There is free quality health care for everyone and the welfare model works as a risk-reducing mechanism. Danes simply have less to worry about in daily life than most other people and that forms a sound basis for high levels of happiness.
Whether or not you agree with the Danish tax system, I think it’s fair to say that this view of taxes is quite different from the view of taxes encoded in SimCity. But you can’t experiment with cultivating a citizenry that feels that way— that opinion on higher taxes is in the bones of the game’s systems.
Again, leaving something out of a simulation isn’t necessarily the problem, and it doesn’t make these games inherently invalid. What I am trying to point out is that when our sandboxes are circumscribed in this way, we are only able to learn certain types of messages from them, and those messages are shaped by the values of the creators.
“So, only work on projects with good values, right?” Sadly, it’s not that easy, either.
Although we can control what goes into our simulations, we can’t truly control what people take away from them. Even projects made with the best of intentions can lead to the exact opposite interpretation as you’d hoped.
“Immersive journalism” is a tactic in which documentarians use VR simulations of real-life events (like torture in Guantanamo Bay, or the Syrian refugee crisis) to try and bring people closer to stories, and to induce empathy. The thesis is that being able to see through the eyes of the victims of these stories would allow players to understand these tragedies from the inside. A noble goal — but consider this excerpt from Empathy is Not Enough, about One Dark Night, which was based on the murder of Trayvon Martin:
In many cases, these “empathy machines” may lead to outcomes that undermine their own stated purposes. In a revealing interview [she] says, “I’ve had a journalist say that they better understood Zimmerman’s position.”
Spent was a popular online game about the daily struggles of surviving poverty. Players take on the role of a single parent working a minimum wage job, and need to figure out a way to stretch their pay and a stash of $1,000 over a month. The month is, of course, packed with lose-lose decisions and punishing random events, making it easy to see how a person who slips into poverty can get trapped in that cycle for the rest of their life.
Or at least, that’s what a lot of people thought it would convey. Unfortunately, the paper “Playing below the poverty line: Investigating an online game as a way to reduce prejudice toward the poor” discovered that sometimes the exact opposite happened — that the game had the ability to actually increase prejudice against the poor:
We found that because playing a game about poverty (and thus having control over one’s outcomes) led participants to believe that poverty is personally controllable, it did not positively influence attitudes toward the poor. […] Furthermore, personal ideology influenced the effectiveness of playing an agency-promoting poverty game. People low in meritocratic beliefs were more strongly and more negatively influenced by the game.
[…]
This is problematic because the current paper suggests that interactive games have the potential to actually increase prejudice toward certain groups. Thus, if policymakers and members of Congress are encouraged to play one of these games that, due to some aspect of its design (e.g., focus on agency), promotes negative attitudes, it is possible that these games could lead to the passage of state or federal laws and policies that perpetuate or increase the inequalities faced by members of stigmatized groups.
To design Spent as a game meant giving the player some immediately-recognizable game-like agency in it — and yet it was this agency that led more people to see poverty as being a natural consequence of personal choices. Ouch.
Am I saying you should give up on trying to make simulation games, or that making one means you’re doomed to fail? Of course not! I only bring up these examples to make you think about what games are doing when they model things, either intentionally or unintentionally. Things like:
- What are we trying to model?
- Who is trying to model it?
- What assumptions do our simulations have?
- Who benefits from those assumptions?
- What do those assumptions leave out?
- What do users actually take from it?
And the reason I devote an entire two-hour session of this class to simulations is because, as a game designer, there is a nonzero chance that you will be asked to make rhetorical devices for all sorts of people and companies and causes — and it will be up to you to think critically about the systems you are designing, and the rhetoric you are putting into the world.
Whew! A lot of heavy stuff this time, I know. But I promise: these cautionary tales aren’t the final word on simulation games — just a necessary primer. Next time, we’ll talk about simulating things a little closer to home, in ways I think are successful, exciting, and — of course — joyful. See you in Part 7, which hopefully won’t take me another 2 years.⁴
¹ Now, of course, it’s not like computers are the only things that can execute and evaluate the outcome of complex rulesets. We have terribly complicated board and card games that function just fine without computers; heck, people can simulate dogfighting with books. But I think it’s fair to say that computers can handle the execution of a series of rules at a far greater scale and speed than the average human.
² For a more recent example, Tears of the Kingdom fails have a similar energy.
³ I am once again begging you to not try and litigate the definition of a game with me. Find someone on Twitter to argue with instead.
⁴ Update: it didn’t.
Hope you enjoyed! If you want to buy me a socially-distanced coffee, you can do so here. If you would rather donate somewhere, I’d be delighted if you’d send these folks a couple bucks.
