Theory of Fun for Game Design

by Raph Koster

When something in a chunk does not behave as we expect it to, we have problems.* It can even get us killed. If cars careen sideways on the road instead of moving forward as we expect them to, we no longer have a rapid response routine unless we have trained a “chunk” for it. And sadly, conscious thought is really inefficient. If you have to think about what you’re doing, you’re more liable to screw up. Your reaction times are orders of magnitude slower, and odds are good you’ll get in a wreck. That we live in a world of chunking is fascinating. Maybe you’re reading this and feeling uncomfortable about whether you’re really reading this. But what I really want to talk about is how chunks and routines are built in the first place.

Excess chaos just doesn’t have pop appeal. We call it “noise,” “ugly,” or “formless.” My music teacher in college said, “Music is ordered sound and silence.” “Ordered” is a pretty important word in that sentence. There’s some highly ordered music that doesn’t appeal to most of us, though. A lot of folks say that the strain of jazz known as bebop is just noise. But I’m going to offer up an alternate definition of noise: Noise is any pattern we don’t understand.

You’ll be at sea for a bit, but you may experience a little thrill of delight once you get it, and experience a moment of discovery, of joy.

“Grok” is a really useful word. Robert Heinlein coined it in his novel Stranger in a Strange Land.* It means that you understand something so thoroughly that you have become one with it and even love it. It’s a profound understanding beyond intuition or empathy (though those are required steps on the way). “Grokking” has a lot in common with what we call “muscle memory.”

When our brain is really into practicing something, we’ll dream about it. This is the intuitive part of the brain burning neural pathways into our brain, working on turning newly grasped patterns into something that fits within the context of everything else we know. The ultimate goal is to turn it into a routine.

The distinctions between toys and games, or between play and sport, start to seem a bit picky and irrelevant when you look at them in this light. There’s been a lot of hay made over how play is non-goal-oriented and games tend to have goals; over how toys are aimed at pointless play rather than being games; about how make-believe is a form of play and not a game. A game designer might find those distinctions useful because they provide helpful guideposts.

But what a book will never be able to do is accelerate the grokking process to the degree that games do, because you cannot practice a pattern and run permutations on it with a book, and have the book respond with feedback.*

Fun comes from “richly interpretable” situations.* Games that rigidly define rules and situations are more susceptible to mathematical analysis, which is a limitation in itself.

This is an important insight for game designers: the more rigidly constructed your game is, the more limited it will be.* To make games more long-lasting, they need to integrate either math problems we don’t know the solutions to, or more variables (and less predictable ones) such as human psychology, physics, and so on. These are elements that arise from outside the game’s rules and from outside the “magic circle.”

Fun from games arises out of mastery. It arises out of comprehension. It is the act of solving puzzles that makes games fun. In other words, with games, learning is the drug.*

This doesn’t mean it necessarily craves new experiences — mostly, it just craves new data. New data is all it needs to flesh out a pattern. A new experience might force a whole new system on the brain, and often the brain doesn’t like that. It’s disruptive. The brain doesn’t like to do more work than it has to.

Any of these will result in the player stating that she is bored. In reality, some of these are boredom+frustration, and some are boredom+triumph, and so on. If your goal is to keep things fun (read as “keep the player learning”), boredom is always the signal to let you know you have failed. The definition of a good game is therefore “one that teaches everything it has to offer before the player stops playing.” That’s what games are, in the end. Teachers. Fun is just another word for learning.*

Exploring conceptual spaces is critical to our success in life. Merely understanding a space and how the rules make it work isn’t enough, though. We also need to understand how it will react to change to exercise power over it. This is why games progress over time. There are almost no games that take just one turn.*

Music excels at conveying a few things — emotion being paramount among them — but as a medium, is not very good at conveying things outside of its “sweet spot.” Games also seem to have a sweet spot. They do very well at active verbs: controlling, projecting, surrounding, matching, remembering, counting, and so on. Games are also very good at quantification.

It’s worth asking ourselves what skills are more commonly needed today. Games should be evolving towards teaching us those skills.

Our world grows ever more interdependent; if a currency collapse occurs on the other side of the world, the price of milk at our local grocery could be affected. A lack of empathy and understanding of different tribes and xenophobic hatred can really work against us. Most games encourage “othering” the opponent, treating him as “not like us,” teaching a sort of ruthlessness that is a proven survival trait. But historically, we’re not likely to need or want the scorched-earth victory, despite legends of “sowing salt” over conquered cities.* Can we create games that instead offer us greater insight into how the modern world works?

If I were to identify other basic human traits that game designs currently tend to reinforce, and that may be obsolete legacies of our heritage, I might call out traits like: Blind obedience to leaders and cultism: We’re willing to do things in games simply because “those are the rules.”* Rigid hierarchies or binary thinking: Games, because they are simplified, quantized models, usually reinforce notions about class, jobs, identity, and other fluid concepts. The use of force to resolve problems: We don’t tend to see a way to form a treaty with our opponent in chess. Like seeking like, and its converse, xenophobia: Seen in countless role-playing games where we slaughter endless orcs.

There are whole genres of games that are about husbandry, resource management, logistics, and negotiation. If anything, the question to ask might be why the most popular games are the ones that teach obsolete skills, while the more sophisticated ones that teach subtler skills tend to reach smaller markets.

A survey of games featuring jumping found that the games with the “best controls” all shared an important characteristic: when you hit the jump button, the character on screen spent almost exactly the same amount of time in the air.* Games with “bad controls” violated this unspoken assumption. I’m pretty sure that if we went looking, we’d find that good jumping games have been unscientifically adhering to this unspoken rule for a couple of decades, without ever noticing its existence.

This offers a possible algorithm for innovation: find a new dimension to add to the gameplay.

Games are not stories (though players can create stories from them).* It is interesting to make the comparison, though: Games tend to be experiential teaching. Stories teach vicariously. Games are good at objectification. Stories are good at empathy. Games tend to quantize, reduce, and classify. Stories tend to blur, deepen, and make subtle distinctions. Games are external — they are about people’s actions. Stories (good ones, anyway) are internal — they are about people’s emotions and thoughts. Games are generators of player narratives. Stories provide a narrative.

Nicole Lazzaro* did some studies watching people play games, and she arrived at four clusters of emotion represented by the facial expressions of the players: hard fun, easy fun, altered states, and the people factor.

My personal breakdown would look a lot like Lazzaro’s: Fun is the act of mastering a problem mentally. Aesthetic appreciation isn’t always fun, but it’s certainly enjoyable. Visceral reactions are generally physical in nature and relate to physical mastery of a problem. Social status signals of various sorts are intrinsic to our self-image and our standing in a community.

There is a constellation of positive emotions surrounding interpersonal interactions. Almost all of them are signals of either pushing someone else down, or pushing yourself up, on the social ladder. Some of the most notable include: Schadenfreude,* the gloating feeling you get when a rival fails at something. This is, in essence, a put down. Fiero, the expression of triumph when you have achieved a significant task (pumping your fist, for example). This is a signal to others that you are valuable. Naches, the feeling you get when someone you mentor succeeds. This is a clear feedback mechanism for tribal continuance. Kvell, the emotion you feel when bragging about someone you mentor. This is also a signal that you are valuable. Social behaviors, such as signals of intimacy, often representing relative social status.* One example is feeding other people, which is a very important social signal in human societies.

fun tends to hit when bumping up at the top edge of flow.

Fun is about learning in a context where there is no pressure from consequence, and that is why games matter.

Successful games tend to incorporate the following elements:* Preparation. Before taking on a given challenge, the player gets to make some choices that affect her odds of success. This might be healing up before a battle, handicapping the opponent, or practicing in advance. She might set up a strategic landscape, such as building a particular hand of cards in a card game. Prior moves in a game are automatically part of the preparation stage, because virtually all games consist of multiple challenges in sequence. A sense of space. The space might be the landscape of a war game, a chessboard, or the network of relationships between the players during the bridge game. A solid core mechanic. This is a puzzle to solve, an intrinsically interesting rule set into which content can be poured. An example might be “moving a piece in chess.” The core mechanic is usually a fairly small rule; the intricacies of games come from either having a lot of mechanics or having a few, very elegantly chosen ones. Most of these mechanics turn out to be from a fairly small set of problem types: estimating curves, optimizing, matching, balancing, or classifying. A range of challenges. This is basically content. It does not change the rules, it operates within the rules, and brings slightly different parameters to the table. Each enemy you encounter in a game is one of these. A range of abilities required to solve the encounter. If all you have is a hammer, and you can only do one thing with it, th...

There are also some features that should be present to make the experience a learning experience: A variable feedback system. The result of the encounter should not be completely predictable. Ideally, greater skill in completing the challenge should lead to better rewards. In a game like chess, the variable feedback is your opponent’s response to your move. The Mastery Problem* must be dealt with. High-level players can’t get big benefits from easy encounters or they will bottom-feed. Inexpert players will be unable to get the most out of the game. Failure must have a cost. At the very least there is an opportunity cost,* and there may be more. Next time you attempt the challenge, you are assumed to come into it from scratch — there are no “do-overs.” Next time you try, you may be prepared differently.

Not requiring skill from a player should be considered a cardinal sin in game design. At the same time, designers of games need to be careful not to make the game demand too much skill. They must keep in mind that players are always trying to reduce the difficulty of a task. The easiest way to do that is to not play.

Simply check each system against this list: Do you have to prepare before taking on the challenge? Can you prepare in different ways and still succeed? Does the environment in which the challenge takes place affect the challenge? Are there solid rules defined for the challenge you undertake? Can the core mechanic support multiple types of challenges? Can the player bring multiple abilities to bear on the challenge? At high levels of difficulty, does the player have to bring multiple abilities to bear on the challenge? Is there skill involved in using an ability? (If not, is this a fundamental “move” in the game, like moving one checker piece?) Are there multiple success states to overcoming the challenge? (In other words, success should not have a single guaranteed result.) Do advanced players get no benefit from tackling easy challenges? Does failing at the challenge at the very least make you have to try again? If your answer to any of the above questions is “no,” then the game system is probably worth readdressing.

The lesson for designers is simple: a game is destined to become boring, automated, cheated, and exploited. Your sole responsibility is to know what the game is about and to ensure that the game teaches that thing. That one thing, the theme, the core, the heart of the game, might require many systems or it might require few. But no system should be in the game that does not contribute towards that lesson. It is the cynosure of all the systems; it is the moral of the story; it is the point.

So what is art? My take on it is simple. Media provide information. Entertainment provides comforting, simplistic information. And art provides challenging information, stuff that you have to think about in order to absorb.

Mere entertainment becomes art when the communicative element in the work is either novel or exceptionally well done. It really is that simple. The work has the power to alter how people perceive the world around them.

What would a game like this be? It would be thought-provoking. It would be revelatory. It might contribute to the betterment of society. It would force us to reexamine assumptions. It would give us different experiences each time we tried it. It would allow each of us to approach it in our own ways. It would forgive misinterpretation — in fact, it might even encourage it. It would not dictate. It would immerse, and change a worldview.

All of these are organized around the same principles: negative space, embellishing the space around a central theme, and observing perturbations and reflections. There was a zeitgeist at that time;* these approaches were “in the air,” but there was also conscious borrowing from art form to art form. This occurred in large part because no art form stands alone; they bleed into one another. Can you make an Impressionist game? A game where the formal system conveys the following: The object you seek to understand is not visible or depicted. Negative space is more important than shape. Repetition with variation is central to understanding. The answer is: of course you can. It’s called Minesweeper.*

The best test of a game’s fun in the strict sense is playing the game with no graphics, no music, no sound, no story, nothing. If that is fun, then everything else will serve to focus, refine, empower, and magnify. But all the dressing in the world can’t change iceberg lettuce into roast turkey.

The ethical questions surrounding games as murder simulators, games as misogyny, games as undermining of traditional values, and so on are not aimed at games themselves. They are aimed at the dressing.

The problematic case is a game that contains both brilliant gameplay and offensive content. The commonest defense is to argue that games do not exert significant influence on their players. This is untrue. All media exert influence on their audiences. But it is almost always the core of the medium that exerts the most influence because the rest is, well, dressing.

Much of our view of the world is shaped by our perceptions and the way we filter information as it reaches us. Clarifying our understanding of that filter is reshaping our relationship to the world. In this light, it’s interesting to see how many of the most famous quotes of Jean-Paul Sartre seem eerily applicable to our relationship to the virtual worlds created by games. Students of philosophy would tell you that he was simply recognizing the artificiality of every world we perceive, since they are all mental constructs in the end. Games thus far have not really worked to extend our understanding of ourselves. Instead, games have primarily been an arena where human behavior — often in its crudest, most primitive form — is put on display. There is a crucial difference between games really illuminating and exploring the human condition, and seeing how the human condition manifests while we play games. The latter is interesting in an academic sense, but it is unsurprising. The human condition manifests anywhere and anytime we interact.

But I see glimmers of hope in many games.* I see the possibility of creating games where the rules are informed by our understanding of human beings themselves — counters that react according to the newly discovered rules of human minds.

Consider a game in which you gained power to act based on how many people you controlled, but you gained power to heal yourself from attacks based on how many friends you had. Then include a rule that friends tend to fall away as you gain power. This is expressible in mathematical terms. It fits within an abstract formal system. It is also an artistic statement, a choice made by the designer of the ludeme. Now, the tough part — the game’s victory condition must not be about being on top or being at the bottom. Instead, the goal must be something else, perhaps ensuring the overall survival of the tribe. Now, suddenly, we see that being at the top, and having no allies, is a choice. Being lower in the status hierarchy is also a choice, and it may be a more satisfying choice. The game is presenting a pattern and a lesson with a specific desired outcome. We need the right feedback in place as well, of course: we should reward all players for sacrificing themselves for the good of the tribe. Perhaps if they are captured in the course of the game, they may no longer act directly, but still score points based on the actions of the players they used to rule. This would represent their legacy — an important psychological driver that mere power fantasies tend not to tackle. There are many possible lessons to be extracted from such a game, and there’s no right answer to the question of choice of strategy. It is simply representing some aspects of the world as it is. It’s crude, and n...

But games do need to illuminate aspects of ourselves that we do not understand fully. Games do need to present us with problems and patterns that do not have one solution, because those are the problems that deepen our understanding of ourselves. Games do need to be created with formal systems that have authorial intent. Games do need to acknowledge their influence over our patterns of thought. Games do need to wrestle with issues of social responsibility. Games do need to attempt to apply our understanding of human nature to the formal aspects of game design. Games do need to develop a critical vocabulary so that understanding of our field can be shared. Games do need to push at the boundaries. Most importantly, games and their designers need to acknowledge that there is no distinction between art and entertainment. Viewed in context with human endeavor and what we know of how our inner core actually works, games are not to be denigrated. They are not trivial, childish things.

All of which is to say that I don’t think we get to ignore the sexism, the classism, the occasional racism, and the general crudity of the commercial game industry. The prostitute in Grand Theft Auto may be a power-up in mechanical terms.* But in experiencing the game, it takes a game critic to divorce her from the context in which she appears. And frankly, game critique isn’t even developed enough to give that particular game object and interaction a name. My answer here is, I’m content with accepting my responsibility on that front. We must improve.

The task I have to make my grandfather proud of what I do seems fairly simple, really. It’s not that dissimilar to the role he took up each time he picked up his carpentry tools in his workshop. Work hard on craft. Measure twice, cut once. Feel the grain; work with it, not against it. Create something unexpected, but faithful to the source from which it sprang.

There’s a science of happiness now (such a phrase!). Researchers tell us that happiness is driven by factors like gratitude, using one’s strengths, a sense of social connection, striving for goals, and optimism.* These sound a lot like what games do at their best, and that may be the most important closure of all.