A Theory of Fun

[HRose]

Player justice and player-created content are both pervasive myths. There's never been a game where either was worth a darn.

There is, fortunately, evidence that you are wrong, because there are such games out there.

Isn't every form of PvP already player-made content? A guild event isn't player-made content? A raid against an enemy town in WoW isn't another form of player driven experience?

I still believe that you cannot expect the players to become writer and artist to replace the work that the real devs are supposed to deliver.

But ALWAYS, when you put players within a structure that is well planned to offer various possibilities of interaction (forced interaction in a way or another) the possibilities are endless.

I believe Lum wrote the same on his blog recently (but around December, so I cannot link) with the example of the miner who goes out of the town at risk and can ask friends to secure the area.

[Murgos]

He probably knows that, since he posted in the thread.

Bruce

Sorry, I had to drop all the way down to +3 to see anything he wrote.  I can't read /. at that level it gets too full of noise.

[Margalis]

Not to re-rail the topic or anything, but if you read the book, it does discuss this. Basically, muds to mmorpgs *right now* (not forever, I hope) is mostly like comparing chess on a board to a nice 3d chess on a computer. The pieces might grunt and swing, but the action is still mostly representable on the traditional board.

OK, but as long as you keep saying "MMORPGS are just MUDS" that will always be the case. As long as the nomenclature is MUD nomenclature and the "thoughtful" papers are all MUD-centric, that will always be the case.

Most of the writings that focus on MMORPG theory immediately boil down to talking about MUDs, so it's no surprise that many of todays games are basically MUDs + graphics.

Then again, it's also no surprise that people will say things like "I will never play a game without a Z-axis again" or "man, movement in COH is cool!"

You aren't doing the world any favors by interchanging "MUD" and "MMORPG", and I don't see any attempt to actually differentiate them in the games you make.

If you think of a MMORPG as just a MUD, and you go about creating it as "hey I'll make a MUD with graphics" that's what you'll end up with, and that seems to be your approach.

[HRose]

OK, but as long as you keep saying "MMORPGS are just MUDS" that will always be the case. As long as the nomenclature is MUD nomenclature and the "thoughtful" papers are all MUD-centric, that will always be the case.

Most of the writings that focus on MMORPG theory immediately boil down to talking about MUDs, so it's no surprise that many of todays games are basically MUDs + graphics.

Then again, it's also no surprise that people will say things like "I will never play a game without a Z-axis again" or "man, movement in COH is cool!"

You aren't doing the world any favors by interchanging "MUD" and "MMORPG", and I don't see any attempt to actually differentiate them in the games you make.

If you think of a MMORPG as just a MUD, and you go about creating it as "hey I'll make a MUD with graphics" that's what you'll end up with, and that seems to be your approach.

Wonderful, precise and concise comment :)

[Raph]

The converse is just as true, that as long as you focus on the graphics, you can lose sight of what makes MMORPGs and MUDs different and special.

Look, as you can tell, I AGREE with you that MMORPGs need to actually make use of the distinctive elements they bring to the table.

However, both MMORPGs and MUDs are clearly varying instances of the same thing, which I tend to prefer to call "virtual world." And seen on that level, a truly massive percentage of the design work is identical, whether you are approaching things from a user-experience point of view or not.

To say that you don't see any effort to differentiate the two variants in the games I've made seems silly... is the way housing and object placement worked in UO something that would have worked in text? My preoccupation with allowing players visual customization--I coded colored metals into UO, remember? Dancing? Would the debate with HRose over freeform housing placement even APPLY in a text setting?

Of course there are massive distinctions in the experience between MUDs and MMORPGs. The graphics are far far more accessible (though there are losses as well in going to a graphical representation). If you read the book, you know that I spend a fair amount of time emphasizing how important the experience design is.

And yet, in both cases, the experience design is the layer on top of quite a lot of stuff. I'm all for starting with the experience design, don't get me wrong. But ignoring the huge huge commonalities behind the scenes is folly. Trust me, I've actually done both, and there is far far more held in common than held separately--far MORE than in designing a pen and paper RPG and designing a CRPG (I've also done both).

The battle isn't that. The real battle is between MMORPGs as single-player games and MMORPGs as multiplayer games. In that battle, MMORPGs and MUDs are brother and sister. I claim as cousins games like Planetside and WW2Online as well as games like ATITD and Second Life.

This is a far trickier discussion because frankly large-scale worlds haven't proven their fun. Much of the virtue of EQ2 and WoW and CoH cames from their intentional limiting of scale. As it happens, I remain convinced that large scale has a lot to offer, and that lies at the root of my dislike of stuff like excessive use of instancing. But I'd be stupid to deny the advantages of small scale for craftng user experiences.

[HRose]

All you write in undoubtedly true and correct. Again the point isn't wipe off what's behind (the MUDs and what they still have to teach), the point is to discard those observations used as models.

The genre needs new models that focus on new aspects (the z-axis is a valid example) more than everything else. The point isn't to borrow a few elements of the graphic layer, the point is to build *directly* a product focused and ready for the new medium.

It's about a switch of balance.

Adding 'physics' to a game world is probably the next, slow, step. I believe this is the only element that Garriott 'got'.

[SirBruce]

I admit I was disappointed that CoH didn't have any destructible world elements.  I reaalize that's a lot of overhead to do in a MMOG, and to prevent the world from becoming a wasteland you'd have to regenerate light poles and mailboxes and barrels and parked cars fairly quickly.  But such a big part of superhero fare is picking up stuff and tossing it around, and Freedom Force had trained me to come to expect this in a superhero game.

Bruce

[Raph]

All you write in undoubtedly true and correct. Again the point isn't wipe off what's behind (the MUDs and what they still have to teach), the point is to discard those observations used as models.

Say rather, don't use those observations as models in the cases where they are not applicable given the user experience you want to design.

There are plenty of things from text that aren't in the MMOs now AND are applicable to MMORPGs. Ignoring those seems... dumb, basically.

The genre needs new models that focus on new aspects (the z-axis is a valid example) more than everything else. The point isn't to borrow a few elements of the graphic layer, the point is to build *directly* a product focused and ready for the new medium.

It's about a switch of balance.

Adding 'physics' to a game world is probably the next, slow, step. I believe this is the only element that Garriott 'got'.

I'd say that for example, Second Life is both aware of the lessons of MOOs and clearly building something that is graphics-centered.

[MahrinSkel]

This is a far trickier discussion because frankly large-scale worlds haven't proven their fun. Much of the virtue of EQ2 and WoW and CoH cames from their intentional limiting of scale. As it happens, I remain convinced that large scale has a lot to offer, and that lies at the root of my dislike of stuff like excessive use of instancing. But I'd be stupid to deny the advantages of small scale for craftng user experiences.

I think that large scale worlds *have* proven their fun.  They just aren't quite large enough.  By this, I mean that what we have is a boundaries problem, our hand-crafted static content, our dynamic generated content, and our player sandboxes are stepping all over each other.

Think of it in terms of "footprints".  The footprint of an avatar is the maximum impact it can make upon the world.  In an EQ-like game, the active footprint is limited to the range of its longest-range attack, and its passive footprint is limited to the distance it (or any corpses or inventory objects it leaves behind) can be seen.

In UO or SWG, your footprint (both active and passive) is much larger, you can have pets, vendors, houses, and so on.  The balancing act is between the scale of the world, the size of the footprint, and the ways that these footprints can interfere with each other and with the functioning of the world.

In a small-scale world of an EQ-like game, where certain pieces of real estate contain highly desirable static content, then even the minimal footprint of an avatar and its abilities can be too large, and the players bump up against each other in various unfun ways.  Hence the attraction (to the players) of instanced content, as this effectively shrinks the footprint of the player by sealing it away from others (note that this is both strength and weakness).  Its attraction to the designer is that it removes the first M from MMOG for design purposes, and allows them to more directly apply single-player game design experience to the problem.  But it's making lemonade when circumstances hand you lemons, not really a solution.

UO wasn't that large in scale, and the various elements stepped all over each other.  Players found nasty things spawning inside their houses, when that was fixed the variety of spawns and hunting grounds was reduced, that was an example of the sandbox stepping on dynamic content.  When PK's planted houses near the entrances to dungeons, the sandbox was stepping on the static content.  And when NPC's tried to wander and emulate social activities, the players couldn't find them to buy things, which was an example of dynamic content stepping on static content.

SWG had a larger scale world, and a much better scheme for separating the dynamic and sandbox content from the static (and not really much static content), but it didn't do so well at separating the sandbox from dynamic.  

Probably the sandbox and dynamic content will always be in opposition to some degree, and the design problem there is to allow them to coexist peacefully.  A big part of that is almost going to have to be just plain *bigger* worlds, where the avatar footprint in the sandbox can easily be absorbed by the sheer scale without choking out the dynamic content.  And not just larger in scale, but larger in detail as well, or else we have either Horizons (lots of world and nothing happening) or ATITD/Second Life (lots of world and lots you can do with it, but nothing there to draw in players *except* the possibilities of what they can do there).

Which brings us smack up against two problems: Emergent content, and system architecture.  Emergent content generation methods offer the possibility of getting out a lot more meaningful detail than we put in, which is neccessary if we are to get such worlds without our budgets hitting *9* figures.  And the current server-side architectures cannot scale to running worlds of such complexity of such sizes with any cost-effective amount of hardware.  

We can't count on Moore's Law to bail us out for two reasons: Moore's Law doesn't seem to hold very well these days (the best PC I can buy today is only about 25% more powerful than what I could have gotten two years ago, and most of the increase is in faster memory, not CPU); And the current architectures haven't been tracking well with Moore's Law anyway, they are actually running into the 4GB memory limit (which theoretically we can bust with 64bit systems, but then we'll hit memory *bandwidth* limitations).

So, any way you slice it, those of us who want to create worlds rather than theme park rides have a big job to do first.  We've got to go back to basics and solve some fundamental problems:

1) What is fun, where does it come from, and why does it go away? (You have started that process, but your Theory of Fun is definitely not comprehensive)

2) How can we tame emergent systems so that they'll do what we *want* (create fun, see #1), while still doing more than we expect (so we can get get more productivity from our word-building budget without creating monotony)?

3) How can we run worlds large enough for the player footprints to play nicely with #2?

--Dave

[HRose]

Say rather, don't use those observations as models in the cases where they are not applicable given the user experience you want to design.

There are plenty of things from text that aren't in the MMOs now AND are applicable to MMORPGs. Ignoring those seems... dumb, basically.

But I'm not saying that, nor I'm trying to criticize your work.

I just say that if you take an "x" game and look at it from this point of view you'll probably finish to conclude that it's really a MUD dressed with graphic.

On this same forum you wrote a concept that can be extended to this topic:

Does your game NEED it? No. But given that it is one of the axes of gameplay that makes use of persistence, and persistence is one of the key things these games offer that other games cannot , well, leaving it out may be considered to be at least underutilizing the genre.

In the same way the graphic, the sound, the perception of the body, the environment etc... These are all elements that need to be used or you are just again underutilizing the medium.

The type of language is still too eradicated in just another wrap of a MUD model. What is needed is to fully utilize what's available, develop a new language that doesn't forget what was behind but that is also able to advance on its own qualities.

And this without fancy projects like Second-life. The use and 'feel' of the graphic is still way behind on the standard gameplay, how you move the avatar (another glaring weak spot of SWG and not a case that it was overlooked), how you interact with the environment, the practice of story-telling etc...

All this just begs to be fully utilized and I'm sure that the players will reward this approach because this type of evolution is based directly on the needs of the body. Something that lies in the human nature and so is one of those elements that can be shared directly between *everyone*, no matter of playstyles in this case. The vastest audience possible and the direct opposite of the niche of a MUD.

The fact of the very difference of success between EverQuest and the average Diku is the direct demonstration of how the market itself ASKS for a different approach. And I find again limited to read Richard Bartle melting these genres without even remotely considering the difference of MEDIUM.

This is another type of trivialization that does no good.

[HRose]

1) What is fun, where does it come from, and why does it go away? (You have started that process, but your Theory of Fun is definitely not comprehensive)

I wouldn't go further. The type of the argument defines the fact you'll never 'get it' completely.

It's like an hadful of sand, the more you tighten your first the more there's nothing within. The best way to 'get it' is about defining a wrap. The idea of patterns and the implication of the learning process at the base of the fun is already enough to define it without losing everywhere the various implications.

[MahrinSkel]

It's like an hadful of sand, the more you tighten your first the more there's nothing within. The best way to 'get it' is about defining a wrap. The idea of patterns and the implication of the learning process at the base of the fun is already enough to define it without losing everywhere the various implications.

No offense, but no it *isn't*.  If "pattern identification" and learning brain rewards were enough to to be a complete theory of fun, then there would be no "fun" activities that did not involve pattern identification and learning.  For example, many people find MMOG fishing fun, yet there is no pattern, no puzzle, no learning.  Many people find performing an activity they have *completely* mastered (bike riding, playing the same FPS map for the 300th time) to be fun, but they aren't learning anything and they've mastered all the patterns.

Raph's theory is a landmark, because it's the first solid place to stand, a fixed point to use as a reference in our search for "fun".  But it is not a "theory of everything fun".

--Dave

[HRose]

No offense, but no it *isn't*.  If "pattern identification" and learning brain rewards were enough to to be a complete theory of fun, then there would be no "fun" activities that did not involve pattern identification and learning.  For example, many people find MMOG fishing fun, yet there is no pattern, no puzzle, no learning.  Many people find performing an activity they have *completely* mastered (bike riding, playing the same FPS map for the 300th time) to be fun, but they aren't learning anything and they've mastered all the patterns.

Raph's theory is a landmark, because it's the first solid place to stand, a fixed point to use as a reference in our search for "fun".  But it is not a "theory of everything fun".

I'm sure Raph can defend himself on this :)

My opinion remains the same, I criticized his point of view, more than his considerations. I always considered his theories as half the medal, the other half is about a cultural aspect that cannot be converted into a formal system.

So the heroic sense, the archetypes, the goals etc... These are all about a cultural substance that cannot be tied directly by a theory of fun. It's all about the myth and its weight on our lifes.

[MahrinSkel]

I'm sure Raph can defend himself on this :)

I don't think he needs defending, because I'm not attacking.  For the first time we have a definition of "fun" that is useful in an engineering sense.  We can use it to direct our planning and evaluate our results, we can agree on methodologies for achieving and measuring it, and so on.  I'm just saying it can't be complete, because there are forms of fun that it can't account for.

But it may point the way towards others, for example what he describes is ultimately an argument that the feeling we describe as "fun" is our brain's way of rewarding itself for figuring something out.  So maybe other forms of fun are other brain rewards for other behaviours that evolution has found useful in helping shape our minds.

Is a physicist who points out that Einstein's theories of relativity are not a "Theory of Everything", but only of specific portions of the universe that we didn't understand the relations between before, criticizing Einstein for not being smart enough or knowing enough to go the whole way?  Game design theory is not religion, we don't stone the heretics.

--Dave

[Lum]

Game design theory is not religion, we don't stone the heretics.

Has it been that long since your last design meeting?

[AOFanboi]

Clearly people wanted SWG to be a better game.  I don't think it can be fixed now, though; a game's reputation becomes fixed.  SOE would be better advised to spend its resources on a completely different Star Wars game.

People want Massive KOTOR Online, and they aren't going to get it. That series is now the yardstick for Star Wars games to measure up to.

[HRose]

People want Massive KOTOR Online, and they aren't going to get it. That series is now the yardstick for Star Wars games to measure up to.

Peoples want what you offer. They didn't want KOTOR before the game was out.

Standards are set by quality, not by expectations. It's just a matter of order, first comes the quality, then come the expectations.

[SirBruce]

What I wanted was X-Wing vs. Tie Fighter Online.  And I still think they should make that game, and no amount of "But you can play Jump To Lightspeed and get all that and more!" is going to get people who want X-Wing vs. Tie Fighter Online to play Star Wars Galaxies again.

Bruce

[HRose]

What I wanted was X-Wing vs. Tie Fighter Online.  And I still think they should make that game, and no amount of "But you can play Jump To Lightspeed and get all that and more!" is going to get people who want X-Wing vs. Tie Fighter Online to play Star Wars Galaxies again.

Bruce

I agree, if I was leading the project back at that time I would have developed the space part to a great extent, then add planet activity as an expansion.

[HaemishM]

People want Massive KOTOR Online, and they aren't going to get it. That series is now the yardstick for Star Wars games to measure up to.

Peoples want what you offer. They didn't want KOTOR before the game was out.

Bullshit. They DID want KOTOR, they just didn't know to call it that. But many generations of Star Wars fans wanted some kind of role-playing game set in the Star Wars universe. They might not have known that they'd like an Old Republic game, but that's what designers are for, to dream up shit they think will hit the joy spots of their particular audience.

Standards are set by quality, not by expectations. It's just a matter of order, first comes the quality, then come the expectations.

No, if you want to make a profit, first you decide who your audience is. Then you figure out what that audience expects. Then you build a product that is supposed to satisfy those expectations, or at least some of those expectations. Quality is a part of the production process, not the design process.

No one starts out to create a shitty game. They just don't bother to dot the I's and cross the T's enough, and they end up with a shitty game. Quality is a part of the production phase, not the design phase.

[Fargull]

Think of it in terms of "footprints".  The footprint of an avatar is the maximum impact it can make upon the world.  In an EQ-like game, the active footprint is limited to the range of its longest-range attack, and its passive footprint is limited to the distance it (or any corpses or inventory objects it leaves behind) can be seen.

Dave,

I like the way your thinking.  UO presented such huge potential, mainly because it was so open.  EQ closed a lot of the openness, but it also provided a more streamlined path.

I still think the worlds created are two small, both in physical space, but in closed streamlined power focus by the avatar.  Vanilla flavoring is becoming way to common.  I am talking beyond the tank/mage/healer triad, but the power selections are more and more the fireball, fireball 2, fireball 3 generisism...  SWG did a few things right, one of them was the variety of classes that did not have a focus even near the triad.

It is nice to hear that people in the seats of power are at least entertaining or directly looking over the thoughts you have expressed.

[HaemishM]

I believe that not only are the worlds too small, but more importantly the populations are TOO BIG. The latter can make the former even more obvious.

Of course, being a complete dick, I'm not going to elaborate because my explanation is enough to make an article out of that I've been meaning to write for weeks now.

[Raph]

Giant necro!

At GDC Online in Austin, i did a "ten years later" on the original talk.

Video is posted here: http://www.gamasutra.com/view/news/179928/Video_Raph_Koster_revisits_his_Theory_of_Fun_10_years_later.php

[Slayerik]

Legendary necro.

[Sky]

I think the highlight was when Raph pointed out the typo in the title of the book 'A Theory of Funk'. Then he put on star shades and rocked out some Bootsy basslines.

[Ingmar]

So how's that article coming along, Haemish?

[HaemishM]

I don't even remember what I was trying to say then. 

[Nebu]

Giant necro!

At GDC Online in Austin, i did a "ten years later" on the original talk.

Video is posted here: http://www.gamasutra.com/view/news/179928/Video_Raph_Koster_revisits_his_Theory_of_Fun_10_years_later.php

Really enjoyed the talk.  Thanks for posting the link.  You are an excellent speaker (and teacher).

[UnSub]

Clearly people wanted SWG to be a better game.  I don't think it can be fixed now, though; a game's reputation becomes fixed.  SOE would be better advised to spend its resources on a completely different Star Wars game.

People want Massive KOTOR Online, and they aren't going to get it. That series is now the yardstick for Star Wars games to measure up to.

HAHAHAHAHAHAHAHAHAHAHAHAHAHAHA etc.

[lamaros]

I thought the talk was really great Raph. Engaging and stimulating. Even entertaining, in parts!

[Samprimary]