Games as Learning: Raph Koster's Theory of Fun for Game Design

The credit for the inspiration for this blog has to go to Raph Koster’s Theory of Fun book. It’s a good, quick read, one that I think can help a lot of gamers and designers.

This isn’t a review of that book. Not exactly. But it’s a look at one of the core ideas of the book, then applying it to my own understanding of education, psychology and gaming.

Koster doesn’t take a stab at what a “game” is precisely, but rather chooses to look at overlapping elements of numerous respected academic attempts to codify the term.

From there, he posits a theory of what makes some games fun for us and others not. Further, why do some games stagnate and others continue to inspire us? He pulls from his vast knowledge of gaming and merges it with behavioral psychology to find insights.

These are not easy questions. Koster’s is an answer that I think is broadly true, though perhaps not quite sufficient to explain all modes of fun. Pleasantly, this is a fact he seems to accept himself, and I believe it’s a nuanced, interesting book overall.

What Is Fun?

Reviewers somewhat notoriously hate calling games “fun” because it’s such a nebulous term. How can we expect to be taken seriously when we’re resorting to such subjective ideas?

However, what are we actually talking about in game reviews if not how fun they are? There are ancillary purposes for some gaming, but “fun” is the intent of the vast majority of those engaging with gaming as a medium, digitally or analog. By and large, it’s why they exist at all.

Tic-Tac-Toe

As adults, Tic-Tac-Toe isn’t much fun because it’s too simple. If you don’t know the solution to tic-tac-toe to always have it result in a draw, you could almost certainly figure it out after studying the game for a couple minutes. After this, it no longer holds appeal.

Chess could undoubtedly be “solved” by a powerful enough computer, and indeed it is entirely solved with – I believe – seven or fewer pieces on the board (any pieces in any legal configuration).

But Chess occupies some for a lifetime. Why?

Play is Learning

I’m truncating the full theory so that I can add some of my own thoughts to it, but Koster’s idea is that play is related to learning and challenge.

Small children can enjoy tic-tac-toe because it’s still challenging. It only loses appeal once it’s solved.

Conversely, that same small child isn’t going to have fun playing Chess, because it’s beyond them. They lack the ability to meaningfully improve. It’s not a challenge they can define and work at. It’s merely confusion and frustration. Yes, children can pick up Chess at an early age; we’re talking about before this is possible though.

Chunking, Honing and Iterative Progression

Chess offers something else, though. We have the capacity to learn thousands of complex games, but most of us don’t want to. Why? Because it’s starting from scratch each time.

Chess allows us to “chunk” (Koster’s word, but also one from behavioral psychology) a lot of information. I’ve played thousands of Chess games. I no longer have to think about how the pieces move. I no longer have to remind myself what victory looks like. Numerous concepts are absorbed to the point where I don’t have to devote thought to them.

So the thought becomes on iterative tactical and strategic improvement. Improve enough, and you’ll start to understand both what you still need to improve at, and how you would go about improving those aspects of your game. All at lower levels of cognitive exhaustion.

This is learning. It’s progression. It’s satisfying. And it’s not dissimilar to a small child with tic-tac-toe. It’s just less solvable.

Chess itself won’t be everyone’s cup of tea, but it illustrates the overarching point well.

Granted, playing another combinatorial abstract game will likely allow us to transfer some skills and knowledge from one to another. So it’s not quite starting over from ground zero. But it’s still a skill regression.

What’s Shown vs. What’s Learned

There needs to be a point of reference in reality for a game to make sense and thus be playable. Even the most psychedelic, abstract game will have some basis of mathematical or logical coherence, or else it wouldn’t be playable and would frustrate anyone eventually.

What a game simulates isn’t necessarily what it’s teaching, though. Koster uses the example of first-person shooter video games. These games are actually worse than useless in learning to legitimately shoot firearms. But that’s not what you learn in them. You learn spatial awareness, decision-making in crisis situations, tactical positioning, and often cooperation between one’s teammates. It’s fertile ground for the type of punctuated improvement that triggers the “fun” reaction in us.

Some games model interpersonal situations; the mechanical, mathematical side of the game is secondary or nonexistent. There are passionate treatises on the value of games for children in learning to navigate the complexities of human interaction, which can be every bit as complex as the varied combinations on a Chess board. There’s a lot to be learned here for children.

Tangentially, I’d say the same for many adults. But that’s perhaps another topic.

Applying it to Observation

What about the gamer who always wants to buy new stuff? They play games 1-5 times then move on to the next one.

There are likely other things going on in such situations, related to the chemically addictive behaviors of consumerism. But, to throw a bone to such gamers, variance is a valid way to have “fun.”

These gamers have “chunked” a lot of cognitive processes in gaming though. Very little is actually “new” in gaming, so these people are able to meaningfully engage with a game much more quickly than someone might who lacks the heuristics that come from playing hundreds or even thousands of hobby board games.

Playing “new” all the time is more frustrating to certain family members, for instance, not because they lack the cognitive capacity to learn the games but because they’ve had fewer opportunities to “chunk” playing heuristics that will allow them to engage the game quickly to the point where it’s fun. They could get there, of course, but the barrier exists between them and that fun which makes a known entity more enticing.

I have experienced this myself. In 2014, I played a game called Terra Mystica. It was orders of magnitude more complicated than any board game I’d played previously. I hated it, and wondered why anyone would play something like it. Unsurprisingly, I made no strategic breakthroughs in my session of it. It was obtuse, opaque, and frustrating.

Fast forward to 2022 and I played again and had to relearn it from scratch, but this time with several hundred more games of experience (and thousands of sessions) under my belt. I had “chunked” a lot more experiential information, which allowed me to absorb Terra Mystica much more quickly.

These days, while Terra Mystica is not a favorite game of mine, but I do enjoy it. And the only big difference is that it is within my personal challenge threshold where it both presents interesting challenges to me but I can also understand it well enough to apply deliberate strategic thinking to its challenge.

I’m was no smarter in 2022 than I was in 2014, at least not to an extent that would explain this profound gap in my ability to understand the game. Experiential heuristics explain the entirety of this disparity.

So it’s not even about cognitive capacity, as it is with young children vs. adults in their gaming. It’s about where you are along certain skill or experience spectrums, relative to the challenge being presented to you.

Variance in Its Various Forms

So if Chess can retain interest for 10,000 plays (or far more), it has to be able to provide enough variance to reward investment that many times. And it does! In fact, you’ll almost certainly never play two identical games of Chess through even 1 million sessions. Meanwhile, a couple toddlers playing tic-tac-toe might repeat themselves after playing for 10 minutes.

Do all games do this? Likely not, but it’s also a tall order to ask a game to withstand 10,000 sessions. I’ll never reach that number for any game other than Chess in my life, and would be surprised if more than a few handfuls reach 100+. So for most, all they have to do is provide a challenging conceptual space for the duration of these 1-100 sessions.

It’s also worth looking at the styles of variance that exist. Many “simple” games owe their continued popularity to how the players at the table provide infinite variance within the confines of the game. Some of my favorite games have interesting mechanical puzzles, but more so interweave the mechanics into player actions and personalities that affect the proceedings. This is the adult version of interpersonal learning that we discussed earlier.

You may be trying to read others’ intentions, or body language, or trade and negotiate with them, or scheme and bargain, or obfuscate your own body language and intentions. These are tangible skills, ones that have corollaries in the so-called real world. They can also be developed like any other skill, and this, too, is fun for a lot of people.

Conversely, the mechanical puzzle may be too obtuse to quickly engage meaningfully with. Or the interpersonal nuance may seem too random to feel as though you can make meaningful insights. This may or may not be the case, but it’s true for that person. And so they’ll bounce off of these experiences as “not fun” because they either struggle to make progress or the barrier to similar progress is lower elsewhere.

Designing for Fun

This has design implications. There’s a term called “cognitive load” in games that’s usually a bogeyman to be avoided, wherein the amount of cognitive processing required simply to parse a game state is too high to engage easily on a tactical or strategic level. Usually this barrier can be pushed through to enjoy the game, but many people won’t, because they become frustrated.

And why wouldn’t they? The “learning” in your game – i.e. the Fun – is unlikely to be any better than their favorites, and you’re hiding the good stuff past too much cruft.

I see another flaw in games sometimes where the insights are too obvious, and the emergent considerations are too few. “Oh, so Pelicans make Storks worth more. So if I happen to draw Storks, I should trade for Pelicans.” Sometimes similar combinations are obvious and lacking in layered strategic considerations. There will be a ceiling to the enjoyment this game provides, even if any gamer – even the most hardcore – can likely enjoy the progression for at least a few sessions before they tire of it.

Notice that a lot of these things have nothing to do with winning? Sure, they do in an oblique sense. But winning in and of itself isn’t the “fun” activity for most. It’s the climb there. It’s learning to cooperate with your squadmates in Counterstrike. It’s instinctively knowing when your sibling is bluffing with 10% more accuracy because you can pick up on subtle body language on an almost unconscious level.

It’s that journey, and it’s the thousands of little touchpoints along the way where minor insights lead to breakthroughs in the experience. Even if “the journey” isn’t the destination of a lot of such efforts, particularly with those who push for mastery to the point where a game system is rote and effectively mastered. Koster addresses such instances, even though they don’t alter his main argument.

Application to Learning Models

I was originally an education major, and a lot of academic study on education models mirrors this. If a child is at Point B and you want to get them to Point Z, it’s best to focus them on Point C.

Maybe they could grasp Point D or E, but if you try to teach them Point M, they’re likely to become frustrated. Similarly, if you teach Point A, they’ll get bored.

This creates struggles for teachers with students of varying skill levels, and is why things like class size are vitally important. But it’s the same idea in play.

We might not call Point C in an Algebra course “fun” in a traditional sense, but there’s a satisfaction in it for the person learning it. And this isn’t so far removed from the types of iterative progress we see in the most lasting games.

It’s creating the appropriate challenge and then scaffolding the structure such that anyone can climb to mastery, or at least higher levels of mastery than they’ll start with.

This is backed by formal study of learning and education. Koster even cites it briefly. Look up “zone of proximal development” if you’d like to read more. It’s interesting stuff. Basically, it’s that B-to-C push, except instead of a tutor or teacher, the game itself provides the requisite guidance to progress to the next level, sometimes figuratively and sometimes literally!

Limits to the Theory

Some of this is me extrapolating on Koster’s ideas, and these are likely the weaker points in the blog. I encourage you to read the book in its entirety. It’s good!

I don’t think this theory is sufficient to explain “fun” in all its multifarious forms. But Koster wisely titled the book “A” theory of fun instead of “THE” theory of fun.

He also discusses other reasons for playing games, of which there are several. Not all of these map perfectly to our ideas of what fun is, but they can still be useful or satisfying in other ways. Regardless, gaming can be a tool for more than just “fun,” though I’d say it’s uncontroversial to suggest this is often our primary purpose for it.

I do think there’s a lot here to consider both about how we learn and how we experience satisfaction in a variety of activities. Whether or not we call it “fun” is not the point. It’s the underlying feeling – and the scaffolding in design and practice that leads to it – that’s the point.

Theories and Practice

I’m glossing over a lot from the book, but I hope I’ve whetted your appetite for reading more.

Koster talks about how a lot of learning in games still relates to survival skills that apply to our hunter/gatherer days as a species, muses on how games could evolve to help reward other types of skills, and discusses the differences – both strengths and limitations – of games compared to literature in storytelling and narrative or thematic structures.

He pens my favorite defense I’ve ever read of games as art, yet openly calls them an immature artform compared to both their potential and other forms of art and media that they’re often compared with. He challenges himself and his audience in the process, in ways that he hopes will inspire new ways of thinking about game creation.

It’s a dense book in terms of ideas and takeaways, but is also a surprisingly quick read. Insofar as I see the relation between how we learn and how we experience fun in games, I think this takeaway alone is quite revelatory and applicable to various social, competitive or entertainment endeavors. It bears keeping in mind for any designer of games but also players of games looking to enjoy themselves and be fulfilled in their efforts.