Why We Play MMOs

Learning Is Fun …

Have you read Raph Koster’s book, Theory of Fun for Game Design? It’s one of those books that people are always telling game designers to read. But I hate it.

I hate Raph Koster’s book because it teaches such a shallow and non-universal lesson: that the core of fun is learning, and that a good game is one that teaches you everything it has to teach you before you get bored. This is a convenient definition for a lot of game designers, because it fits them. They like games that they constantly learn from. It’s almost like some sort of pre-requisite for being a game designer.

There are also apparently studies that show that learning new things releases endorphins in the brain, and I have no doubt this is true. So does eating chocolate, and acting out violence, and reacting to outside stimuli that you’ve come to associate with rewards, and sex, and… on and on.

To suggest that learning alone is the core of fun is a really … well, arrogant … thing to do. It reeks of the forced-grouping hypothesis that held sway over the industry for half a decade — made popular by designers who got into the industry after playing lots of EverQuest. Maybe you remember it: “MMO’s are all about interacting with other people. That’s what makes them different than single-player games! So we need to force people to group, even if they say they don’t want to, because they just don’t know what they’re missing.”

It’s crap. These bogus theories of motivation are a dime a dozen because:

  1. Designers tend to create hypotheses that allow them to create games they like to play. (Duh!)
  2. Designers aren’t typically scientifically minded. It’s not our strong suit. Like stone-age philosophers, we simply correlate what we see with the most obvious possible explanation. This is how superstitions get started, too: the mind loves to correlate things at the drop of a pin, even if they shouldn’t be correlated.

You can probably tell that this pushes my buttons. It’s fine for Raph to hypothesize, but the real problem is that his thin little treatise is used by a lot of other designers, many of whom should really know better, as a way to rationalize their personal style of game development. Entire systems of game design are now based around the idea that “learning is why games are fun”.

It’s not a bad model for making certain types of games. Miyamoto’s incredibly fun action games fit this model well … up to a point. But learning is hardly the only way to have fun with games.

… But Other Stuff is Fun Too

The thing is, for most professional game designers, learning really is the most fun part of a game. We tend to instantly strip away the context, the art, the story, and indulge in the mechanics. “Ooh, here’s something new! This is great!” We subconsciously assume that other people are like us. (“I’m sooooo bored of traditional MMO mechanics, and everybody else is too!” Not so. You are projecting, wishing it to be true.)

But this interest in mechanics and learning is hardly universal — even for designers. What struck me when reading the “learning = fun” theory is how bad it is at explaining my MMO behavior. For months after I stopped regularly playing EverQuest 1, I kept logging in occasionally just to fish. I wasn’t learning crap, and I didn’t even care what I caught or what my skill was. I was immersed in a fantasy world and I enjoyed the simple escapism provided by watching the virtual sun set.

Escapism is fun for me sometimes, but so is the feeling of being a part of a well-oiled team. And sometimes I like to just beat the snot out of things weaker than me. And so is … huh, I guess I have a lot of reasons for playing MMOs. So do you.

So how do we unify all these motivations into a theory of fun? We don’t. That’s stupid. As H. L. Mencken once wrote, “Complex problems have simple, easy to understand, wrong answers.” Just because you can unify every source of fun into one simple theory doesn’t mean it has any hope of being right. Our brains are chaotic, confusing, complex creations. A lot of different things trigger pleasure in our heads.

So how do we tell why people play games? Well, we can start by doing some empirical research.

…Wait, really? Empirical research? In the gaming industry?

Yes. It is true. Nick Yee’s Daedalus Project is based on surveys of real MMO players instead of empty theorizing. This is why it’s more valuable than your pet theory or my pet theory. It’s probably fairly flawed, but it makes theories like “learning is the core of fun” look like sun-god-worshiping prehistoric nonsense in comparison.

Here, let’s look at the Model of Player Motivations Nick Yee uncovered. Because I know you didn’t bother to click that link above, I’m just going to steal some of the important charts from his website. (Sorry, Mr. Yee. Hope ya don’t mind.)

Nick Yee's Motivations

Nick Yee's Motivations

Another common source of “motivations” is Bartle’s famous player types. After many studies, Nick found that several of Bartle’s types don’t really map up very well. For instance, Bartle hypothesized that Explorers love finding new places in the world, as well as unraveling game mechanics. In reality, these two motivations don’t tend to appear in the same players too often.

Most players do have multiple different motivations from this chart, though — and that’s one of the most important takeaways from Nick’s work: most of us have many reasons for playing games, which vary from day to day and game to game. You probably have motivations in many of these categories. (You can even take a little quiz to see which ones.)

It’s true that many of these motivations can be mapped to “learning” — especially if you’re willing to make really tenuous connections like “competition is just a form of learning about other people!” and silly stuff like that. But many of them are just simple human needs, like being dominant, being liked, being part of a group, or escaping from reality for a while.

Pick What Your Game Is Good At

Every MMO in the “virtual world” vein supports a large number of gamer motivations; this is part of the secret to MMOs’ success. But no MMO can be great at all of them. I suggest you pick several motivations from Nick Yee’s chart (three or four at most) and be really good at them, then let the others happen as best they can happen.

It’s tempting to conflate motivations with demographics, but that’s a mistake. For instance, it’s true that slightly more females than males prefer “socialization” motivations in MMOs, but the gender difference is pretty small. Similarly, you may be tempted to assume that any game aimed at 18-24 males should be all about achievement. But that demographic’s motivations are all over the place, and older demographics are strongly motivated by achievement as well.

One method of getting a handle on your target audience is to create personalities to represent them. This is an old marketer’s technique: create a handful of personal stories and backgrounds, and then tailor a game to what those people would like. If you have a well-established IP, you can pull from the demographic data of your IP to help you figure out what personalities to use; otherwise, you’re going to have to do some research on the sort of people you envision playing your game. If you’re planning to reach out to existing MMO players, maybe you can start by picking some bios of traditional MMO gamers.

It boils down to the First Question: who is this game for, and why? In theory, you should know this before picking motivations. But on the other hand, if you don’t have a clue who your game is for, choosing several key motivations will at least narrow down your choices!

If I had to guess, I’d say that WoW’s three most powerful motivations are advancement, competition, and teamwork. EQ2’s strengths, on the other hand, might be mechanics, teamwork, and discovery. Asheron’s Call could be mechanics, discovery, and competition. All of these games are considered “hard core” MMOs. They just don’t focus on quite the same motivations to quite the same degrees.

Unfortunately you can’t safely pick, say, your three personal favorite motivations and run with it, because there may not be a large enough audience to support that mix. So when you have your polling contractor research your target demographic, make sure they include some questions that can highlight their motivations.

And please don’t tell me that your $20,000,000 game can’t afford to use a professional pollster, because if you think that, you obviously haven’t even looked into the prices. Hey, wait, you were just going to make a game without even deciding who it’s for, weren’t you? Shame on you! You’re not a novelist who can just write anything you want and hope somebody likes it; you’re spending a ton of money to create a service. You need to know who your service is for.

Thanks, Daedalus Project

I’d like to thank Nick Yee for his many years of researching MMO players. He recently put the Daedalus Project into hibernation, which makes me sad, because this is impossibly valuable data for any MMO developer. If you haven’t perused his data collection, you should go do that now.

His stuff isn’t perfect: there are some motivational categories that aren’t as strongly tied together as others, and his polling subjects were largely self-selected, which no doubt influenced the results. It’s also highly biased towards Western culture in general and EQ1 (early on) and WoW (later) players specifically.

But it’s a lot better than what you had before, because before you were just making stuff up and pretending it was universal. I know you were. Yes, I saw you doing it. Cut it out!

You still haven’t clicked the link above, have you? Fine, I’ll leave you with another Daedalus Project chart, stolen for your amusement. This one breaks down how various stereotypical statements relate to player motivations.

Nick Yee's Motivations - Breakdown

Nick Yee's Motivations - Breakdown

This entry was posted in Design. Bookmark the permalink.

17 Responses to Why We Play MMOs

  1. Mike Darga says:

    Great post. I’m very guilty of being annoyed with players and designers who don’t care about Good Gameplay.

    The tendency to correlate and generalize has been driving me crazy lately too; I just wrote about it on Friday. Knowing your audience and being more objective have also been on my mind.

    This is my favorite game design blog, largely because it’s one of the few game design blogs that’s actually about designing games. Or maybe, I too just like to read things that reinforce my own opinions heh. Looking forward to reading more.


  2. Raph says:

    Sorry you didn’t like the book. But “though I wish he had provided at least some shred of evidence to back up his hypothesis” — did you read the endnotes? There’s quite a lot of evidence out there.

    FWIW, the four types of enjoyment cited in the book are from empirical research done by Nicole Lazzaro, and the book is almost entirely about one of the four, since it’s the only one of the four central to games as such.

    Your logging in to fish is also covered in the book. :)

  3. Eric says:

    I actually removed the line about evidence just a few minutes ago, after re-looking at your end notes. Sorry about that.

    I basically didn’t buy the book’s core premise, as it feels too contrived to me. But the content isn’t really why I hate the book — to be frank, mostly it is that your industry clout turned your hypothesis into an automatic “fact” in many MMO developer circles, when really it should be just a discussion point. Developers are always searching for “secrets to game design”, and your book feels like one at first glance, which makes it the dangerous sort of drug that designers like to consume without question. Mostly, I want them to question.

  4. Longasc says:

    Hi. I did not read the book. I also do not develop games. But I play them. ;)

    If the book made you think, it served a purpose. Not really a good consolation for poor Raph, but at least you came up with constructive and interesting ideas after criticizing the core premise of the book.

    As this is my first entry here, I just wanted to drop a line and say that I am pretty impressed by your blog, Eric, and I really hope that people read it and come up with a new and exciting MMORPG. I was a bit sceptical about your ideas, as I remember Asheron’s Call 2 as one of the worst MMO launches in history, I never ever quit a MMO as quickly as AC2. Closely followed by Age of Conan, which I decided not to buy a few hours after I left the starter island Tortage… and it was a friend’s account anyways, so I saved some money…^^

    You have some very good ideas, I really hope your or someone else turns them into a great game. I am a bit tired of getting served EverQuest derivates for ages by now. I prefer “Ultima Online style”, but MMORPGs are a bit stuck in what EQ and WoW coined to be a “MMORPG” nowadays. Probably a lot of designers, too.

  5. Raph says:


    There’s a reason the book is titled *A* theory, not THE theory. It bugs me too when people use the latter, actually.

    That said, I think you have a problem here when you say both “it’s fine to hypothesize” and “it’s arrogant.” If you want people to question, they have to hypothesize, offer up opinions, have them critiqued, think some more, and so on. We’re not going to get very far in our own understanding of games — even from an empirical point of view — unless people are willing to put their thoughts out there. Calling it arrogant is the same as saying it’s not OK to do so.

    Fortunately, I have a thick skin, I don’t care very much what you think about the book (though I am crying inside). But I do think the tone is counterproductive, especially when you end up saying “the content isn’t really why I hate the book.”

    PS, the stuff in Dan Cook’s chemistry article partakes as much from the highly empirical game grammar movement as it does from the theory of fun stuff.

  6. I have to agree with Raph here, Eric. I think you’re participating in your own sin by thinking that any single bit of theory is universal. Most of these theories and studies that collect empirical evidence explain one aspect of a game; Raph’s main theory in his book, for example, does a good job of explaining why people do or do not “burn out” on a particular type of gameplay. I wouldn’t necessarily use his theories, however, to plan out a PvP system.

    And, while reality is that people have widely different motivations for playing our games, the flip side is that we can’t consider everyone’s mind. We have to take some shortcuts and make some assumptions in our games to actually get the project done. But, the assumptions we make are important; as you point out, designers can’t just assume that their tastes are universal and use them as the base of their assumptions.

    It’s also dangerous to think that studies like Nick Yee’s are universal, too. While I appreciate the data Nick has collected, as you point out yourself it’s heavily biased toward specific games. It’s important to realize that it’s also biased towards games at specific stages in their lifetimes. He doesn’t have data collected from around a game’s launch, for example, and you know as well as I do that the audience attracted by a game’s launch isn’t necessarily the same audience that is there a year or two later. Trying to attract that later audience at launch may not work very well.

    As has been said before in many places, it’s important for game designers and developers to have the proper tools at their disposals. Hypotheses, data, and summaries are all tools that we should use. Our experience shows us when we should use those tools, and when we need to use something different. Less experienced people will use the tools at the wrong time, as you’ve noted. As the saying goes, “When all you have is a hammer, every problem starts looking like a nail.”

    The ultimate lesson here is that game designers need to look for more tools that they can use at the proper time, not ignore certain tools because they tend to be overused, or used incorrectly. An awl is still a useful tool even if most people use the blunt end as a replacement hammer.

    My thoughts.

  7. Teki says:

    Great write up, thanks.

  8. Scott M. says:

    I think you’ve made some good points and it’s always helpful to challenge current convention and theories of design, especially when they’re being thought of as “fact” to the point of not considering anything else. I’m primarily a Web designer and this happens all too often as the community is influenced and pulled along by the theories of various experts and pundits. Unfortunately the tone of your post, while understandable, undermines your insights (to an extent) because of your invectives and obvious anger at the state of things.

  9. Bart Stewart says:

    Woof. Where to start?

    1. “It boils down to the First Question: who is this game for, and why?”

    Yes. I agree completely.

    If you want other people’s money, you have to give them what they want. But how do you know what they want unless you know who your intended customers are? And even if you figure that out, how do you translate it into specific game features if you don’t know why they want what they want?

    “Know your customer” doesn’t guarantee success. You still have to deliver. But it has to improve the odds.

    This is why I’ve spent the past few years yapping about “player-centric design” as a general goal, and discussing in detail multiple “theories of fun” — i.e., playstyles and player motivations. Having workable theories of fun is highly desirable because they boost the odds of making a game that a commercially viable number of people will want to play.

    So I’m with you completely on this.

    After this… not quite as much. :)

    2. Obviously Raph can speak for himself, but I think you might be taking that word “learning” more literally than it was intended.

    I need to go back and re-read my copy of ATOF, but as I recall it, my impression was that Raph wasn’t using the word in a strict formal sense meaning deliberate pedagogy — ATOF seemed to be more about the process of becoming more sophisticated about games generally. I thought Raph was talking about learning in the same sense of someone learning about art by experiencing art, all kinds of art, in all kinds of ways. With more experience comes greater understanding of what makes one thing work while another doesn’t… and that’s half the foundation of game design. (Namely, the critical half, the other necessary half being the creative gift.)

    If that reading of the term “learning” is close to what Raph intended in ATOF, then maybe it’s not such a bad thing that lots of game designers are checking it out. Is the process of becoming more knowledgeable about games, of developing a critical language, really something that deserves scorn?

    Or is this more a complaint about laziness, and a concern that mindlessly racing to implement a naive understanding of Raph’s book will lead to a bunch of games with some cheesy form of “learning” as their core gameplay? If so, I’m with you on that… but then who’s to blame? Raph for not spelling things out even more carefully than he did? Or game designers looking for easy gimmicks to avoid hard creative thought?

    3. On the specific question of player motivations (i.e., what are they, and which ones are least bogus), the dueling models of Nick Yee and Richard Bartle have been a point of contention for years. Even the two principals have publicly discussed their differences of opinion regarding their respective theories of player motivation. (See http://terranova.blogs.com/terra_nova/2007/10/the-hidden-bart.html and http://www.nickyee.com/daedalus/archives/001298.php?page=2 .)

    Bottom line summary from my perspective: Richard holds that Nick’s multi-motivation model can have value but needs to show utility, while Nick asserts that Richard’s “player types” model is broken because he — Nick — is unwilling to accept the possibility that individuals can consistently prefer and exercise one motivation much more strongly than others… in other words, that people can have a “type.”

    The good news is that, unlike lame theories of behavior such as astrology and enneagrams that start with a theory and go trolling for only the data that appear to support that theory, both Nick’s and Richard’s models of gamer behavior start with behavioral data and then propose theoretical models that try to show patterns in that data. That means both Nick’s and Richard’s models potentially have explanatory and predictive utility… and I believe both do.

    Neither is flawless. Nick’s data come from self-selected responses to questions whose phrasing biases the results. Richard’s data come from close observation over many years of gamers and what they actually do in games, which opens up his data to charges of subjectivity.

    But because both models at least start with data, and then try honorably to find a model that adequately encompasses that data, both (I believe) can have value. Both ways of understanding gamers can be useful to game designers. (Please note that this is not an “all models have equal validity” navel-gazing statement. All models purporting to explain and predict human behavior don’t have equal validity; some — as noted above — are complete crap. Nick’s and Richard’s models happen IMO not to be crap, and that starts with their being derived from data, not the other way around.)

    Personally, while I’ve found Nick’s model useful, I’ve gotten more value from Richard’s original four-type model. Bearing in mind that nobody is ever “just one thing” forever, most people do seem to prefer one motivation over another at least a good chunk of the time. (I suspect there’s a deeper isomorphism here between Richard’s model, Nicole Lazzaro’s “keys,” Caillois’s forms of play, and David Keirsey’s general theory of temperament. I explore that notion at http://flatfingers-theory.blogspot.com/2005/01/styles-of-play-full-chart.html .)

    At any rate, to the extent that individual players do reliably demonstrate particular playstyle preferences in games, I think Richard’s model is a reasonable basis for designing the content of those games.

    It’s maybe not THE answer to the First Question, but, like Raph’s book, it’s AN answer we can use while we grope toward something even better.

  10. Interesting discussion here. Thought I’d pipe in on how the enjoyment of games comes from the emotions players feel around their choices, and how players move between playstyles.

    My 17 years of research indicates that there are 4 primary playstyles that cross all games. We analyzed player’s favorite moments in games, everything from Tetris to Halo to WOW and measured the emotions in players’ faces and bodies. We simplified Paul Ekman’s FACs coding to collect emotions directly from the player so they did not have to remember them in a survey post-game.

    Grouping these observations we discovered two striking relationships between emotions (where enjoyment and motivation come from) and player choices offered by the game mechanic:

    1. Player emotions connect with each other and have prerequisites (Frustration and Fiero, or Curiosity and Surprise)
    2. Game mechanics associated with each group of player emotions created a playstyle

    We call these the 4 Keys and best selling games tend to have at least 3 out of the 4 playstyles. By watching actual players at home/school/work we also found that players actually rotate between 3 out of the 4 styles within a 20 minute session. They don’t just choose one play style as Bartle and Yee’s research implies.

    There are many ways to enjoy games
    1. Hard Fun of challenge and mastery uses goals, obsticals, and strategies to balance player frustration with Fiero (Italian for personal triumph over adversity)
    2. Easy Fun of experimentation and role play where the core fantasy and balance between novelty and the expected inspires the imagination creating curiosity, surprise, and wonder.
    3. Serious Fun from visceral stimuli and value based play where players game to change how they think, feel, and behave.
    4. People Fun from spending time with friends where social emotions such as schadenfreude, amusement, naches, and gratitude.

    The similarity between Bartle’s taxonomy, Nick Yee’s Deadalus Project, and the 4 Keys indicates that we are getting closer to understanding what makes games fun.

    For those interested:
    Free white papers and presentations slides available here: http://xeodesign.com/whyweplaygames

    I’ll be talking about how to use these to use emotion to create viral distribution of games this year at GDC


    Game On!

    Nicole Lazzaro

  11. Tadhg Kelly says:

    So the question behind all of this for me is this: Are the attacks on Raph warranted on the basis that his theory is deeply flawed, or are they on the basis that the poster just doesn’t like the reductionism of a good theory. I suspect the latter.

    The reason I suspect so is that this is something that I’ve come across repeatedly in the last while, which is the unwillingness of developers (and designers in particular) to really break down their work into functioning fundamental parts. There’s a real urge among many to stick to an idea of inapproachable complexity rather than elegance, and it’s an emotional rather than rational argument. The designer doesn’t want to brooch the idea that the things that he works on are not in fact unique snowflakes, but are instead part of a set.

    This is, I think, self-destructive behaviour. It’s an attempt to cast creativity as a complex but solvable problem rather than a simple foundational one because in doing so the designer gets to not face up to an awful truth: Good games, like good everything else, are based on creative talent first and foremost.

    Making MMO’s is not complicated in the way of analysing thousands of player emotions and motivations through a blizzard of spreadsheets and studies. MMOs are generally pretty simple to understand, as are their motivations. What’s hard is coming up with a good idea, a fresh take, a new genre, an exciting spin, an elegant mechanic, all that sort of stuff. The stuff that doesn’t fall out of a model or a study much which captures the essential magic of the thing.

    Raph’s theory is so terrifying to game designers because what it actually says (and a few other theories like it) is that games are actually pretty simple, and that leaves them nowhere to hide. If the model of games is understood at least as a reasonable sarting point (and no model of any artform is ever 100% applicable – see Syd Field) then the focus is not the designer: He knows what the ground rules are. Now he has to actually have an idea. An original idea. And make it work.

  12. Richard says:

    Wow Eric, I’m impressed. It seems every person mentioned in the post came by to read your blog.

  13. Bryant says:

    Wizards of the Coast did a survey of pen and paper roleplayers in 1999; the results are here. I think it matches Nick Yee’s work more closely than Bartle’s, and like Yee’s work, it has the advantage of being based on real survey work rather than on collected impressions.

  14. Eric says:

    Most of these are great points; it’s true that I succumbed to the typical blogger sin of being too aggressive when discussing things I dislike, but I can’t honestly say I’ll stop doing it … it’s the nature of the medium to get a LITTLE carried away.

    Tadhg Kelly’s comment makes me go “whut?”

    Bryant’s link makes me sad because it doesn’t seem to go to the results he talks about.

    Nicole’s link is really interesting stuff and I’m very appreciative that she dropped in to add it.

  15. Wolfshead says:

    I really enjoyed Raph’s book. I found it inspirational and unique. The diagrams and charts were also very helpful. When I first got started in the video game industry there were very few books on game design and most of them didn’t dare try to define “fun”; they would skirt around the subject.

    Just as you don’t see many books on how to write a great song — as a musician myself I find that notion absurd – you also don’t see many books that try to deconstruct the underpinnings of what “fun” is. Even if you disagree with him, we should give Raph top marks for at least having the courage to tackle this subject which has had the end result in provoking thought and discussion.

    I highly recommend his book to any aspiring game designers trying to break into the industry. The knowledge and theories inside the book helped to give me confidence when I discussed concepts like fun and challenge during the interview that landed my first job. I am better game designer because of his book.

  16. Bryant says:

    No, I was apparently a bad copy and paster. Sorry! I did not mean to lure people to my site. :(

    Try this one: http://www.seankreynolds.com/rpgfiles/gaming/BreakdownOfRPGPlayers.html (and feel free to edit the old link).

  17. Pingback: New site to me « Mental Meanderings