Raph Koster on the Past, Present, and Future
The gaming industry rivals Hollywood in both creativity and financial clout, and yet aside from one or two digital mascots and a few cursory, sweary Call of Duty teenagers, there aren’t many physical, human icons or so-called “videogame celebrities” with which to speak of. In our latest series, we’re looking at celebrated MMORPG developers, asking where their ideas for design come from, and why they choose to make online worlds. I may be no James Lipton, but our subjects are as fascinating as any silver screen mainstay.
Raph Koster started his career with LegendMUD before moving onto Ultima Online. His work has helped characterise the MMORPG genre, with Star Wars Galaxies refining the much desired “sandbox” style of play, and his frequent talks and writings on game and world design keeping even the most ardent player weak at the knees. Over his vast career, now spanning over two decades, Raph has greatly influenced the online genre: today we get to speak to him.
MMORPG: What does the MMORPG genre mean to you?
Raph Koster: I have complicated feelings about it at this point. When I started out on MUDs, it was still fresh and exciting. It felt like we were in on a secret, a secret that was going to change the world. It was all the dreams of cyberspace. MUDs were a place where we were learning a lot of lessons about the way people acted in virtual spaces. I remember the first person I ever banned was another admin who was getting his kicks from invisibly spying on players – shades of the NSA topic today, right?
Don’t get me wrong – it was a game, too. But it was a game that was still very fungible and flexible. Even in the relatively rigid Diku format, there was room to experiment, to try new things. We put in crafting and a dreams system and an auction house and a moods system and all sorts of other things that were not the norm. Then we worked on Ultima Online. And again, we were inventing. I think it’s been long enough that people don’t understand how many crazy things were tried there. Even the jump between MUDs and UO was big, in terms of wacky design ideas we put in there.
But then the budgets got really big. The business of it. And with that comes a lot of, well, just inertia. It is harder to try things that are different. It gets harder to push boundaries. It gets harder to even just talk to players. With big money all the challenging questions about things like digital surveillance or whatnot fade away in favor of questions like “can we open up the market of MMORTS gameplay?” And it’s not that those questions are bad ones, they’re ambitious ones, even. Just not necessarily as big as the questions that were floating around when MUDs were going.
I ended up going to smaller worlds, hanging out in non-game worlds like There.com and doing Metaplace and all that, because it was a way to ride the wave of what was happening in the wider web: all the stuff that was going on with blogs and forums and web 2.0 and community participation and collaborative design and agile development. That felt like a wave that MMORPGs were not riding. They’re coming to it now – SOE seems to have a lot of programs on the table that fit, like the stuff where you can submit art, for example. But it’s been 10 years since SWG came out, and seven since I left SOE to do Metaplace. So it took a while.
I think that idea that there was a wave that was missed turned out to be true. Most of the social lessons about virtual spaces are translated to us today in the forms of Facebook and Instagram and Twitter. There’s a lot less game to them, but a lot of the gamey stuff has leaked out: friends lists and guilds and levels and star ratings are all over all the pieces of the web we use every day, and few realize the degree to which it all feels sort of like a MUD now. Issues of privacy. The questions of what is fair governance with a virtual community. And those questions have a lot sharper edges to them now that we are having discussions about things like whether employers can demand to see your Facebook profile. So we weren’t wrong, back in the MUD days, about how things were going to develop. It’s just the MUD part got left behind.
So MMORPGs feel in some ways like a dead end, like a branch that had a lot of early potential but then didn’t fully grow up. I haven’t been hooked on a virtual world, as a player, since Metaplace closed down. I’m excited to see the trend back towards simulation happening now with ArcheAge and EQNext, some of the things that Guild Wars 2 tried... It all feels sort of prodded by Minecraft, in a way. I wish that the genre had kept building on what we did with UO and SWG instead.
So… I owe MMORPGs my career. But I haven’t actually been playing one or working on one for several years now.
MMORPG: Before Ultima Online you helped work on LegendMUD, how important was that as a touchstone of experience in regards to future projects?
Raph Koster: Hugely, as the preceding answer should make clear.
It was on LegendMUD that the issue of player rights first came up, and where I took my first stab at writing a code of conduct that covered both players and admins. LegendMUD was way way more quest driven than any of the MMORPGs I have worked on, in some ways more quest-driven than most of the MMORPGs out today – oh, WoW certainly has more of them and the game pulls you from quest to quest pretty linearly, which Legend doesn’t do. But I would still hold up some of the quests there to anything out there in the MMORPGs.
Legend was also a fairly classless game. We did have the notion of “hometown” or birthplace, that affected what your character could eventually do, but there was a huge amount of flex within that. In Legend you can time travel, and I think that idea, that a virtual world can be made up of really different areas, with different rules even, was something that in retrospect informed my picture of what a virtual world should be pretty powerfully.
Crafting was highly unusual on MUDs at the time. We also did things like furniture and vehicles and player housing… all those ended up in the mix for later stuff.
LegendMUD is still running, so I suppose your readers can simply go log in and check it out.
A lot of the stuff that my wife and I put into the UO design came from notes that we were taking towards a new MUD that we were talking about for after LegendMUD. I still have the paper notes for it, and I put them on my site back when UO turned 15. It is actually a little depressing to see all the stuff in those notes that still doesn’t exist anywhere!
MMORPG: Where do you draw your inspiration and influence from when designing?
Raph Koster: Non-fiction is a huge part of it. I’ll read something about mathematical tiling problems, or about the social structure of early hominids, or about monks in medieval scriptoria and why they recopied manuscripts over and over, and it will give me ideas.
I rarely get ideas from playing games. I will sometimes see mechanics that I just take wholesale, and often I will feel challenged or intimidated by a game design, and want to see if I could pull off something similar. I find that I am more likely to get inspired trading ideas with other designers than from games I play, though. Sit down with smart designers, and start batting ideas or challenges around, and ideas just start to pop.
I do see videogames as part of the overall current of games and sports throughout history, so I do study them a lot. I’m the sort who will buy a handcrafted Nine Men’s Morris gameboard just because I’m a game designer and I should have one in the house.
I read a lot of fiction, and that also goes into the mix.
MMORPG: You talked about world design in presentations and on your blog, how important, within an MMO, is it to make the world feel "real" to the community of players that inhabit it?
Raph Koster: Well, there’s different needs for different players. But for my personal tastes, it is actually the most important thing. I personally, as a player, am willing to forgive worse mechanics if I can get that sense of immersion. It’s not that the mechanics are unimportant, it’s that they matter less to my personal enjoyment than the feeling of being somewhere else does. I think I once described it as “being somewhere you can’t be, doing things you can’t do, as someone you aren’t.”
Other people, of course, value the fun game mechanics above all else. I would say “more power to them,” except that I think they actually have had all the power! Certainly that is the direction in which things have been shaped over time. I think a lot of the evolution of MMOs has been towards less of a sense of immersion, in the name of cleaner interfaces and easier-to-use systems. Stuff like color coding monsters and items, soulbinding, all that, was all in the name of a more gamelike experience.
MMORPG: Do you believe in structuring a players experience, or prefer giving them tools to create a more emergent adventure?
Raph Koster: Both, really. But I strongly believe that you can’t build the emergent tools on top of a static world. As soon as you decide to make storytelling or quests or whatever the basis of your experience, you sacrifice having dynamic and emergent things in the game, because you can’t break or upset all the static content. Whereas if you start with a foundation of simulation or UGC, and layer static stuff on top, that works fine, because the static content is built to assume shifting foundations.
A lot of people think that my answer to this will only be on the emergent side, but they never played the quests on LegendMUD. :)
MMORPG: Was it easier working on Ultima Online simply because, in essence, you were making the rules as you went along?
Raph Koster: No. It was easier working on UO because we were a small team, and we were all pulling in pretty much the same direction. The only other time that has happened for me like that was when we architected Metaplace. SWG had a much larger team and many pressures on it. The social games also had pressures and lots of conflicting goals. That skunkworks moment, when you have a team that is small enough to move fast, big enough to make something substantial, and all in sync on what we’re making, is something magical.
With SWG, there were tensions almost from the get-go, on things like whether to have Jedi in the game, whether to use procedural terrain, internal politics around the San Diego team versus the Austin team, and of course, the stuff associated with being a licensed title. With social games, you’re always engaged in the cultural conflict of creativity versus metrics. And we were in a run-up to selling the company too.
It isn’t really about making up rules. I mean, rules can chafe, no doubt about it. It’s more about working in synchrony.
MMORPG: What different challenges in design did you face with Star Wars Galaxies in contrast to Ultima Online?
Raph Koster: Well, both were based on beloved IPs. Star Wars, though, did offer more constraints on us than Ultima did. The Jedi thing was a huge issue, and eventually, IMHO, it killed the game. We shouldn’t have had Jedi at all, I think. The fact that we had to have an alpha class in an MMO is right off the bat deeply problematic. Then we basically buried the class, made it nigh impossible to get. When the holocron hints on how to be a Jedi started dropping that first Christmas season, that was really the beginning of the end. Because of the way in which the Jedi system worked, for which I take full blame, you had to master all these different skills. When it was secret, that was one thing. But when it was public, it meant everyone started playing in ways they disliked. Botting skyrocketed. People started to grind. And they started to quit.
I know some disagree with me, and say that instead we should have made Jedi much easier to get. That was what was tried both by the NGE and SWTOR… it just doesn’t feel honest to the time period of the game setting to me.
SWG wouldn’t have been possible without the procedural terrain system. Every planet was generated on the fly as you walked around, using formulas that had consistent outputs, so that it was generated the same every time. It was the only way to have such insanely large maps at the time, with dynamic terrain modifiability. Someday I will have to write up how that worked. But it was a big struggle. It was hard to get good graphics performance. It played havoc with things like pathfinding, spawning, collision, and a bunch of other things. We enabled housing anywhere, but then we got, well, housing everywhere.
It doesn’t seem like much today, but SWG had eight planets that were each 16384 meters on a side, with 1m resolution on the heightfield. That’s a lot of data for shipping on CDs in 2003. One texture, one color, and a height value, per tile. Call it 5 bytes. That’s 10 gigabytes for 8 planets. I want to say we ended up reducing down to 2m heightpoles to save memory? And that doesn’t count flora, objects, all that – that’s just the ground!
The artists didn’t quite know what to do with this for a long while. It meant learning a whole new way to create maps. Many of them never quite managed the trick of it.
We were also supposed to have a lot more static quest content in the game than we managed to get. The challenge there was that we never got a real quest system you could fill out in template form. Instead, every quest was hand-coded in Java. I had expected our scripters to be able to do this, but Java turned out to be a challenging choice for the scripting, and we ran way short on time to implement any good quest tools.
MMORPG: Do you feel that the audience of MMORPG games has changed since your last online adventure?
Raph Koster: I’m sure it has, but I wouldn’t know exactly how, because I haven’t been in the trenches on MMOs lately.
I do get a lot of nostalgia attention, I guess you would call it, from people who want the sort of MMOs that I worked on. But I also think that one way in which things have changed is that there is a presumption of a fairly sizable team and budget behind an MMO these days. I am unsure that if I showed up with a new MMO idea, but done in less-than-AAA graphics, that many would bite. Not for the “main” MMO audience, that is – I am sure that there are indie fringes, the kids’ market, etc, which are still open to that sort of thing. But I worry that the barrier of entry to make something is just too high. Like, I got asked a lot if I would Kickstarter a sandbox MMO. I worry I’d have to have the most successful Kickstarter in history to get enough funds for what is needed.
MMORPG: What do you think about modern MMOs?
Raph Koster: This is going to sound terrible, but I mostly don’t. I watch YouTube videos of them, but few of them sound interesting to play.
One reason is that, even as far as we got with Metaplace, we managed to literally (I am not exaggerating) make basically the backend for Snow Crash or Ready Player One. Oh, the rendering of that world was still tied to crappy little Flash windows in a browser, and so no one saw everything that it could do. But it was really something pretty crazy and special. Still too hard to use, and really, too hard to “get a hold of” conceptually. But once you have worked on something that broad, it’s hard to squeeze back down into “kill some orcs.”
MMORPG: Where do you think the genre should go in the future?
Raph Koster: I took a lot of flak years ago for saying that MMO stuff was swallowing everything. But Minecraft is also a personal MMO, and Xbox Live is an MMO, and Twitter, and so on, it’s everywhere now. I think that the traditional virtual world is in a tough spot. A lot of the things that made it appealing are available in other formats now. So it really has to retreat to doing what it does best, doing the things that Xbox Live or a MOBA and so on can’t do. And that probably means losing audience, because the things that MMORPGs do best are more time consuming, slower paced, more immersive, than what we have gotten used to.
MMORPG: Frequent debate occurs between those for sandbox and those for theme park, being a proponent of the former do you feel that kind of design philosophy is most important to the genre?
Raph Koster: I think we get enough theme parks in single player games. Now that you can play those single player games with other people thanks to voice chat, you really don’t need an MMO attached to it. I don’t think it is an accident that Minecraft is as popular as it is.
MMORPG: Given an unlimited budget and resources, what MMORPG would you now design?
Raph Koster: I’d probably go back to simulating a world from scratch. One with a new fictional setting altogether. One where players can add to and contribute to everything. If you’ve read READY PLAYER ONE, that book basically describes what Metaplace’s technology was intended to be… but I think I would focus way more on making a fascinating world first, before going after all the rest.
Adam Tingle / Adam Tingle is a columnist and general man-about-town for MMORPG.com, RTSGuru.com, and FPSGuru.com. He enjoys toilet humor, EverQuest-themed nostalgia, and pointing out he's British: bother him at @adamtingle