RePlay interview: The Puddle, the Splat, and the Droplets
Raph Koster is one of the most influential game designers of the online era. As lead designer on Ultima Online, creative director on Star Wars Galaxies, author of A Theory of Fun for Game Design, and now CEO of Playable Worlds, Raph has spent more than 30 years shaping the future of online games.
In this episode of re:Play, Greg Off and Raph explore the origins of persistent virtual worlds, emergent gameplay, player-driven communities, and the ideas that inspired everything from Ultima Online to today’s online worlds.
They discuss:
• The surprising path from creative writing to game design
• Building Ultima Online and inventing paid Early Access
• The legendary Trinsic siege and emergent gameplay
• Star Wars Galaxies, player cities, and virtual communities
• A Theory of Fun and its lasting impact on game design
• Why Stars Reach is the culmination of three decades of work
• The future of MMOs, simulation, and persistent worlds
Whether you’re a game developer, designer, student, or lifelong MMO fan, this conversation offers a rare look inside one of the minds that helped define modern online gaming.
Video
Transcript
Replay Interview: Raph Koster
RAPH KOSTER: This was literally covered by mainstream press as the worst video game patch ever made. Get ready.
HOST: Welcome, everyone. My guest today is, by any honest accounting, one of the most influential people who has ever worked in online games. He’s also one of the most generous thinkers this industry has ever produced.
He co-founded LegendMUD in 1993 while finishing an MFA in creative writing. He was the creative lead on Ultima Online, the game that introduced persistent worlds to the mass market. He was also the original creative director on Star Wars Galaxies before SOE got involved. He founded Metaplace. He wrote A Theory of Fun for Game Design, which is taught in design programs on six continents, has been translated into nine languages, and is still selling at first-edition pace.
He wrote the Declaration of the Rights of Avatars in 2000, which is now cited in legal scholarship on digital governance. And in 2018 he founded Playable Worlds to build the MMO he’s been planning since the mid-1990s. That game is Stars Reach, and it’s in active pre-alpha right now.
His name is Raph Koster. Raph, welcome to Replay.
HOST: Raph, before we jump in, I wanted to say congratulations. I saw that today you announced your early access window for Stars Reach this morning. That seems like a big day to me. We’re definitely going to spend real time on Stars Reach a little later in the show, but if it’s okay with you, I would love to actually start somewhere else.
The thing I think a lot of people don’t know about you, and you can correct me if I’m wrong, is that you didn’t come into games from tech or fandom. You came in from a creative writing MFA. I thought that would be a really interesting place to start from.
RAPH KOSTER: Yeah, you have the story right. I did come in from a creative writing MFA. It actually goes back a little further than that. I did a double major in creative writing in English and in Spanish literature in undergrad, and also was one class shy of a major in music theory and composition, and one class shy of a minor in studio art.
Ironically, even though I have technical patents and whatnot these days, I never studied programming. I got started doing that as a kid on my dad’s CP/M Osborne 1, and a little bit later on my own Atari 8-bit computer in the early 1980s.
HOST: So did your career path, for lack of a better term, start from just tinkering around as a hobby? Like, “I’ve got access to this computer. I’m going to just play with it and see where it goes.”
RAPH KOSTER: Pretty much. I grew up overseas. My mother worked for UNICEF, so I would fly back to the States over the summers to spend time with my dad. I would see the latest arcade games. Then I’d go back to Peru or Barbados, or wherever we were stationed at the time, and I would make board game versions of the arcade games. Tabletop Pengo, tabletop Q*bert. We would play them during recess in class in school.
That was my actual early game design training. A group of us even sold one copy of a game in a Ziploc baggie in the early ’80s to classmates. We got it up and running on the Atari and on the Commodore 64.
HOST: That’s an amazing early start. I think that’s going to date both of us, because I shared a very similar experience growing up, buying Ziploc-bag, five-and-a-quarter-inch floppy disks for the Apple II Plus or whatever.
You probably, like me, bought a game by Lord British, who did the Ultima series.
RAPH KOSTER: Absolutely. And then later, my first job in the industry, I’m working for him.
HOST: Yeah, we’re going to get there. I definitely want to talk about that.
I wanted to talk a little bit about LegendMUD, which you co-founded in 1993. Talk about that experience, the process of doing that. I was wondering if, because I believe it used historical periods as its setting, that’s sort of how your literary experience got embedded inside a game world.
RAPH KOSTER: Yeah. The theme of LegendMUD was history the way they thought it was.
HOST: Yeah.
RAPH KOSTER: And that meant mythology was fair game. You actually could time travel. Still can. The game is still running, by the way.
HOST: I know.
RAPH KOSTER: Text-based online games are an alien world to most gamers these days. I think most people don’t know there’s still quite a thriving scene there, or might not even be able to wrap their heads around how you play such a thing. But it’s still a thing. Think a cross between text adventures and MMOs. They are the primordial ancestors of today’s MMOs.
Those games were actually more diverse than MMOs. A theme like history was a bit unusual, but not that unusual for what was going on in the MUD world at the time.
LegendMUD was weird because it had that theme. We supported time travel. You could be geographically in London and time travel to Roman Britain, then wander offshore and get entangled with the Beowulf saga. Or you could be in London and wander off to Nottingham and get entangled with Robin Hood or Spenser’s Faerie Queene. Or you could be in the Industrial period and get entangled with vampires in the belfry, and maybe catch a freighter with the East India Company all the way off to the Bay of Bengal and Hong Kong.
So, yeah, it was the literary side. We prided ourselves on very literary descriptions. We were an extremely quest-heavy game for MUDs of the day. Beowulf and the entire Ring of the Nibelungs was in there as a single connected quest.
It was pretty formative in a lot of ways. It led to exploring gameplay beyond tank, nuker, healer; I am of this class; let’s party up, kill ten rats, level up, get gear.
That game was a standard MUD game, and I kind of burned out on it by 1996 personally. Of course, the rest of the world mostly didn’t discover it for another decade or more. But it was a very common mode of play back then. LegendMUD intentionally tried breaking out of that.
We had early forms of crafting, player housing. It was a classless MUD. You didn’t have to be stuck in a single class. We had out-of-character lounges with events and live storytelling and guest speakers. It really wanted to be more of a virtual world than just the game. Even though it was a pretty fun game, and is still a fun game.
HOST: Speaking of the live events and lecture series and live storytelling, what did that teach you about what an online space could be, versus what people just assumed it was?
RAPH KOSTER: First, even the terminology. Today we call these MMOs, which of course as an acronym just means massively multiplayer online blank, because nobody wants to say MMORPG. Very quickly we got MMOFPS and MMORTS, because once you have an online world, we shouldn’t think of it as one game type.
I think that was one of the big lessons for me: virtual worlds were not a genre of game. They were a place within which games could be put.
Very early, we ran into big, lofty questions like governance. We pioneered things like having codes of conduct for the admins, which was actually kind of weird. At the time, the idea that admins might have ethical responsibilities towards the players was considered outre and weird and strange on our part.
So yeah, from pretty early on, maybe it was the theme, maybe it was the other things that we’d put in the game, LegendMUD really pushed us towards thinking of virtual worlds as being much more than just the game.
This was the mid-’90s. We’re talking full-on Snow Crash techno-libertarian mania. We saw it as kind of a future for the internet. And we weren’t entirely wrong.
HOST: Yeah.
I wanted to take a moment. You mentioned Ultima. You mentioned Richard Garriott, a.k.a. Lord British. I’ve heard the story that he logged into LegendMUD to evaluate you before hiring you for Ultima Online. Is that true?
RAPH KOSTER: It’s true. And nobody believed it was him. They all made fun of him.
HOST: That’s hilarious.
RAPH KOSTER: Well, I mean, why would you believe some rando shows up and goes on global chat, “Greetings, all, for I am Lord British”? Why would anybody believe you? It’s like that in a random Discord server.
HOST: Yeah, exactly. Did you know you were being interviewed? Was that pre-arranged, or did he just—
RAPH KOSTER: It was not pre-arranged.
There’s a bit of a “it’s who you know” kind of situation that happened there. The programmer who co-founded LegendMUD with us, Rick Delashmit, actually landed the job at Origin to work on an early version of what was then called Multima. That was the code name. This was sometime in early 1995, possibly very late 1994, somewhere in that window.
He and a few others, like Starr Long, Ken Demarest, and Jim Greer, actually built that Multima prototype. Then, when time came to actually build a team around it, they asked Rick, “Who would you recommend?” He recommended my wife and I, who were two of the other four co-founders of LegendMUD. He recommended the two of us as designers.
So we did a road trip. We were living in Alabama at the time, attending grad school. We did a road trip out to Austin. We interviewed in person. But we had no idea that they were going to log into LegendMUD and start exploring and checking it out and talking to random people. I wasn’t online at the time that Richard did that, so I found out later when I logged in. Everyone was, “Did you hear about this? Was that real?”
HOST: Once you got the offer, did you know what you were walking into at Origin?
RAPH KOSTER: Not entirely.
They sent us a bunch of interview questions, and Kristen and I had already been working on designs for ways to integrate simulation much more into MUDs and virtual worlds. We had come up with this idea of using an abstracted property system that we called resources, because they were economic resources. Things like, you could be able to chop down a tree and get some wood, and then you could use the wood to craft something else.
We were saying you could drive all the AI off of that. That was quite a design break from the way MUDs were typically built. They were really handcrafted entries in a database. A crafting recipe was item A plus item B. The idea that, no, wood can turn into all these things was just strange.
When they asked us these interview questions, they sent us a coding sample for a quest, and it was “find the bug in the code sample.” It was a programming test, and we found the bug. It was a missing parenthesis. Then we said, “But this is a terrible quest design. You ought to do it this way, with that generic resource system.”
They sent us an ask for how we would write quests, and we sent them Beowulf. We said, “Well, here’s the entire narrative of Beowulf, but we wouldn’t do it that way anymore, because it’s very labor-intensive.”
We show up for the first day at work, and we meet Mike McShaffry, who was the lead programmer at Ultima 9, and we say, “So, what’s the game?” Because we were junior game designers, barely there. Our starting salary, I think, was $25,000. We assumed, this is Origin, we create worlds, they must know what they’re doing.
So we said, “Well, what’s the game?” And they said, “Oh, it’s going to be great. We’re going to have this abstracted property system so that quests are autogenerated.” And we looked at each other and realized we had just been thrown in the deep end of the pool. The game was going to be designed around our job application.
HOST: Crazy. That is so crazy.
RAPH KOSTER: We were 23, 24, something like that.
HOST: It being the first time, really, a persistent online world was marketed and sold to at least the mass market, when you were building it, did you have any model for what success would look like, or were you inventing that metric as you went?
RAPH KOSTER: Electronic Arts, I’ve heard multiple numbers for this, but supposedly estimated lifetime sales for the game between 30,000 and 50,000 people ever. That’s what they thought. Total copies that would sell. That was their forecast.
We were not the only project that got going in that same time window. EverQuest, Asheron’s Call, Tibia, over in Korea Dark Ages. This was an inflection moment, and actually most of us came to know each other very quickly during the mid-’90s, because there were so few of us working in the space at the time.
We had commercial models that were on the closed online services, CompuServe and GEnie and the like, that paid by the minute. But we were MUD people, all of us. We were children of the open internet. We were people who used IRC and Usenet and Gopher and other obsolete things, and Telnet and whatnot.
We weren’t able to draw tons of things from the closed online services. It’s a different business model, a different audience, a different approach. We were graphical in a way that some games were 2D graphical like we were. Meridian 59 was getting going; that one was Doom-style 3D graphics. EverQuest would be full 3D hardware-accelerated, but it was coming along later. It didn’t come along until ’99.
It was one of those moments. I always say we were a scene, in the artsy sense. Like the scene around Warhol, or the scene around Seattle grunge, or whatever. Only we were a virtual scene. We were all talking together on the MUD-Dev mailing list, swapping ideas quite outside the view of executives at any of these companies.
We recommended friends to the other companies. I recommended Damion Schubert, who went on to work on Star Wars: The Old Republic, and who was one of the guys we were building MUDs with. I recommended him into Meridian 59.
It was a moment where we all knew each other. We all helped each other. Honestly, we all knew that we were going to change the world. We very much had that sense at the time, because we were in our early 20s and we were working on the cutting edge.
But we had absolutely no sense of what our budgets were, what it meant to be a professional game designer, what the rest of the industry was like.
We literally worked on a different floor of the building, which was under construction. The exterior walls had been knocked down. You surfaced onto bare concrete out of an elevator and could walk off the building to your death.
HOST: Oh gosh.
RAPH KOSTER: And we were in the only part of the floor that still had drywall around it. During the winter it was so cold that we all came in wearing jackets and fingerless gloves so we could type, because our one test server sat directly under the thermostat and kept heating it. So the AC was blowing all winter and we were freezing.
HOST: Oh my God. Unbelievable.
RAPH KOSTER: And we built the alpha of the game. My first day was September 1, 1995, and we opened the alpha in time for E3 of 1996. So six months later, we ran the alpha test.
We had to resort to posting on a fledgling webpage—legend has it, the first webpage anybody in EA had ever launched—with a giant FAQ about what the game was going to be, all of our dreams. We said, “We can’t afford to send out CDs, so send us five bucks so that we can afford to print a CD and send it to you so that you can join this test.”
And 50,000 people sent us money right then. EA went, “Yeah, wait a minute. You guys just invented the preorder.” Yeah. We kind of— it wasn’t even a preorder. We kind of invented paid early access or something.
HOST: It was the first of many things you would invent on Ultima Online. I think the game today is over 25 years old, and its influence is so pervasive that people do not realize how much of what we play in games today comes from Ultima Online.
RAPH KOSTER: Yeah.
HOST: Clearly there was chaos during the development time, but it seems like even the early days after it was launched were chaotic in ways that have now become legendary, between player killers, the ecology collapse, the assassination of Lord British. Was there a moment where you thought, “This is exactly what we built for,” rather than, “This is going terribly wrong”?
RAPH KOSTER: It was both at the same time.
One of my favorite stories about it is actually none of the famous ones. The famous ones have been told in many videos. Your listeners can find lots of material about all of them. Richard forgot to set his invulnerability flag, so he got murdered by a random killer—a thief, actually. Or the player killing problem, which we fought endlessly, trying to solve the problem of how do we provide the freedom without essentially making it so that every single dispute comes to us to get resolved.
That’s a problem that I often say I wish we had actually kept working on. We gave up eventually. I wish we had kept working on it, because today it still feels like the biggest problem on the internet, and we didn’t manage to really crack it back then when it was more controlled. Oh well.
My favorite story is actually players banded together and started a war, and they besieged one of the cities in the game. The way they did it was by building furniture. They built loads and loads of furniture and barricaded all the exits from this walled city called Trinsic.
We’re watching this happen, and the players who load into that city are like, “I can’t get out. If I go out, they’ve left one gate open, and it’s nice and narrow. If I go out that way, I get walloped on the back of the head immediately and all my stuff is taken.”
We were facing this player-management problem. The answer we came up with was to make it so that axes could chop furniture. Rather than constraining what people could do, we expanded the simulation so that people could do more.
The result of that was that the citizens of the town, the people logging in there, banded together, chopped their way out, and won. They won the war.
HOST: That’s amazing.
RAPH KOSTER: To me, those kinds of lessons about emergent gameplay stick more than any static quest that you might do 50 times. But it’s also incredibly hard to balance. It’s a very challenging balancing act. That tension between the amazing power and freedom of emergence versus the danger of grief play and all of that—
HOST: Yeah.
RAPH KOSTER: “Griefer,” a term that came from UO, along with many others. That tension, to my mind, is still in many ways the defining challenge of online spaces in general.
HOST: It feels like, talking about giving that freedom to players, you’ve spent your career fighting against the leveling design pattern. Ultima Online absolutely had skills-based progression specifically to push against that. From your perspective, has the industry learned that lesson? Are we still doing levels?
RAPH KOSTER: Levels are easy from a business point of view. The math from the design point of view is well understood. From a consumer point of view, they’re a very clear carrot. It’s very accessible. You tell them, “Hey, go acquire XP and do this.”
But levels were pulled from Dungeons & Dragons. In your typical D&D session, you don’t actually play with people of different levels. You play with a group of people who are all clustered really tightly around the same level, and you all move through the levels together. Right off the bat, that makes it odd for a world where you’re going to have all the levels interacting together at once.
A lot of the history of MMO design has been putting in hacks to undo levels. You look at sidekicking, mentoring, world phasing. I forget; there are so many terms for this. I forget what Guild Wars calls it when they—level scaling, I think? Or is that ESO? I don’t know.
They’re all ways of squishing everybody back down into a narrow level band again so that they can play together, because otherwise you get outleveling one another. You have a huge problem around content consumption and moving through all the zones of the game, and then 90% of your content ends up obsolete. There are all these problems that are real challenges, not just for developers but for players too.
For me, the big thing is actually that with standard leveling, you level because you get experience points. In D&D, the game master can give you XP for anything, and a good game master does give you XP for anything.
But a computer needs an event. It can’t tell that you are a great guild leader. It can’t tell that you are a really fascinating roleplayer. It can’t tell all these things that are crucial to how players enjoy MMOs. Absolutely, it all gets reduced down to: you killed something, you popped a bubble of XP.
Think about great D&D sessions. Only a fraction of them is spent killing the monsters. So if you want that sense of being immersed in a world, that time where the bard is playing their instrument, that time where you’re roleplaying in the tavern, that time where you’re haggling with the merchant to buy stuff—all of those things are valuable, important parts of feeling immersed in this alternate world.
Just making the games about fighting always felt hugely reductionist to me. It’s like you’re cutting out most of human experience.
So the reason to go classless, or to look for other ways to progress in these games, is to allow more ways to play, more ways to experience the environment, more ways to interact with other people, rather than saying that the only way you can interact is at the point of a gun.
HOST: Before we move to the next topic, I’m curious: with 30 years’ distance from when you were working on Ultima Online, what does it look like to you now? Do you go back and play it? Do you see things that weren’t visible to you when you were making the game?
RAPH KOSTER: Sure. To me today, it feels janky and clunky and primitive and all of that, because it’s a game from that long ago.
I don’t tend to go back and play my worlds. I don’t go back and play LegendMUD either, or Galaxies. But what I see is that despite the eventual dominance of the WoW gameplay model, there are entire game genres built around decorating houses, which UO had years before The Sims. It popped up during the beta in 1996.
There’s an expectation that you can go fishing, which was pretty much not a thing until UO. And I apologize to everybody in gaming, but even Uncharted and Tomb Raider and God of War make you go craft things now. That’s probably my fault. Because there wasn’t crafting like that until Ultima Online came around.
I could go on: even customizing your avatar that way. I look around today and what I see is Minecraft and Animal Crossing and the entire survival crafting genre, but also things like DayZ and Rust, which took away the lessons of the PKing and made it the point.
Today, I look back and I go, those are all children of what we did early on with sandbox MMOs.
HOST: You planted a lot of seeds that certainly took root.
RAPH KOSTER: My preferred metaphor is that we made Ultima Online. It was a puddle, and then World of Warcraft came along, but the droplets went everywhere.
HOST: Crazy.
I know you mentioned Galaxies, and I think that might be a good segue if you’re willing. After Ultima Online, you moved to SOE and became the creative director on Star Wars Galaxies, which, and I don’t say this lightly, I feel is one of the most ambitious games ever attempted.
When you were designing it, what was the thing that you were most certain was right that the market might not be ready for at the time?
RAPH KOSTER: It’s hard to pick one thing. I think I would pick that we used procedurality a lot for building the maps and having dynamic environments, which was something that we wanted, and the design of UO wanted, but we didn’t really have the tech to be able to do in the ’90s. In the early 2000s, we could.
We shipped, I forget how many planets it was, like a dozen planets. It would have been, I think, four gigabytes’ worth of map space, which was a very large map for that time period at the level of detail that we did it, and was too big to actually ship on CDs at the time. DVD drives were not common enough that you could ship DVDs.
Today, of course, we have a No Man’s Sky or something, and that sounds like small potatoes. But that was one of the things that we saw coming and said, “Hey, if you do this, you can also make it modifiable. Players can actually change it.”
I think the dynamic economy that we built in Galaxies was ahead of its time, and I think EVE was very much a fellow traveler at that time. They were out there saying “Let’s make UO in space,” kind of a vibe.
The one that I think didn’t get picked up on nearly as much was the player cities. I think that’s still something underdeveloped in games out there today, that sense of players collaborating to work on even larger-scale things and taking on some of those admin problems themselves.
In a weird way, and this analogy might make more sense later when we talk about Stars Reach, when we think about how Discord or Reddit work today, where you’ve got this big connected social environment but you devolve control of pieces of it—“that’s their server, that’s their subreddit, and they have control within there”—Galaxies’ player cities were an attempt to do that, to help solve the problems that we ran into in Ultima Online.
It was a way to say, “No, players could be mods of an area, and then it’s their problem. They get to make decisions about what they want that environment to be like.” Much of social media is UO, and Discord and Reddit feel like Galaxies cities to me in a weird way.
HOST: I know this is a small example for me, but I remember at the time there was an entertainer class, which was a viable supported profession in a Star Wars game. You could be a musician or a dancer, and it really mattered to the ecosystem. Walk me through that design logic.
RAPH KOSTER: You have to go back to UO and the desire to capture many play styles and support them and reward them in the game.
On many MUDs you might have a bard class, but it was used for taunts and combat, not actually being a musician. It always felt off to me. Sure, you can be a musician, but the music is the least important part. It’s a skin put on top of taunting and aggro management.
I’m a musician. I play a bunch of instruments. I compose and write music. It never sat right with me. But that was true of many of the other things as well.
Doing music and allowing people to build music was challenging. I was thrilled when Lord of the Rings Online finally broke the barrier, which was a legal barrier that LucasArts and Sony wouldn’t let us cross: allowing players to just enter musical notes and play. I was disallowed from doing that. I tried to make it happen, and I wasn’t allowed to.
But in a weird way, if on UO we got made fun of by everybody for having baking bread be a profession, a way to play, in Galaxies we got made fun of for having dancing.
That isn’t one where I go, “Oh, the market didn’t get it,” because today Fortnite makes all of its money off of that. Dancing is the argument that I won.
HOST: Good. Good job, sir.
RAPH KOSTER: That one’s everywhere now, and we take it for granted that of course dancing is a thing. The social dimension of it is a viable play style in its own right.
In the more experimental time periods of the early 2000s and the early 2010s, there were even MMOs entirely about dancing. There was one called Audition that was quite popular that came out of Korea.
I think it goes to show that a lot of the sandbox gameplay features have a much broader appeal than slaying ten orcs does. But they ended up flowering when they were outside of MMOs more than when they were inside an MMO.
HOST: On the other side of that, you had Jedi as an alpha class that pretty much breaks any multiplayer system they touch, and you saw that from the start of Star Wars Galaxies. What was the conversation inside the team about that problem, and how did you guys try to solve it?
RAPH KOSTER: Some of this is tied up with the complicated development history of Galaxies.
When Rich Vogel and I were told by Sony Online, “Hey, we want you to take over this project that is already ongoing,” Galaxies had already had a year-plus of time. It wasn’t called Galaxies at the time. It was called Star Wars Online, and it was being developed out of San Diego.
When they brought in us and our team that had been working on Privateer Online at Origin, which had been canceled, we essentially were handed the team. Development suddenly became split between the two locations. A bunch of people were very upset, rightfully so in my opinion, and quit over it. It was a challenging moment.
It meant that we ended up with a bunch of decisions that had been negotiated with LucasArts in advance. Among them was the time period for the game. It was set between Star Wars—I’ll call it, for the sake of everybody younger than us, Episode IV, A New Hope—and Empire Strikes Back.
That is a time period in which there are no Jedi running around. All this hunt is going on to find Obi-Wan and to find Luke Skywalker. Nobody knows about Luke Skywalker and so on. Bounty hunters get sent after you. In the EU, there were Force-sensitives cropping up here and there, but it was not a moment where there were Jedi all over the place.
So from the beginning we had this challenge of, well, the canon says this shouldn’t be around. The easy game design solution would be: don’t have them. Make them NPCs. But when we floated that, the response was, “But it’s the ultimate fantasy. Everybody wants to be a Jedi.” So that got ruled out.
Then you have the next side of it, which was actually the choice that the prior dev team had chosen, which was: don’t make it an alpha class, just make it a starting class. We looked at that and went, “But that means that a level 50 Jedi, or whatever we do, is going to be the same as a level 50 whatever. They’re not going to be special in any way.”
The response to that was, “No, no, no, no. That can’t be right. A Jedi needs to be able to mow their way through enemies.” And so they had to be rare, super powerful, and anybody needed to be able to be one.
Impossible dilemma.
I floated the idea of, how about we make it a hardcore mode? Diablo was out, of course, by then, and it had a hardcore mode which had essentially permanent death on your character. Permadeath was something that had been tried in many MUDs but wasn’t all that popular. But having it as a mode was kind of interesting, and I hadn’t really seen that before. I’m sure it existed.
The idea was, if you managed to unlock Jedi, you could get a Force-sensitive slot, but that slot would be in hardcore mode, and a timer would start running. You would play, and if you used your powers publicly, we’d start sending Bossk after you, then IG-88, then Boba Fett, then Mara Jade, and eventually Darth Vader. Darth would just kill you. You can’t win.
So you would aim for a high score based on how long you lived. It was a mode. It was a survival mode, a hardcore survival mode. As a reward for your other characters, you could end up with a blue glowy of your deceased Jedi that you could summon at will, for bragging rights.
I still really like that idea. We did talk about it with players, but eventually there was “Oh no, permadeath, we can’t really do that,” and the idea got watered down and removed eventually. I believe there are Galaxies emulators that have actually implemented the system as it was originally intended, but I’ve never actually played one.
The tragedy of Jedi is that then marketing said, “Well, we need them in the game.” We’d come up with a way that wasn’t random, but relied on hidden information, which any cryptographer will tell you is the worst form of security. Marketing insisted that there be Jedi by Christmas and that we needed to drop hints. Once the hints started dropping, players figured out the algorithm to become a Jedi, and it broke the game completely.
HOST: Wow. Here’s a topic that we do not have to talk about. I’m going to leave it up to you. I know you left SOE in 2006 before the New Game Enhancements, and you’ve been clear about that timeline. When you look at—
RAPH KOSTER: It was actually after the NGE, the New Game Enhancements. But I had been moved off the project and told not to interact with the team because they needed to find their own way.
HOST: Interesting.
RAPH KOSTER: I thought the NGE was a really, really bad idea.
HOST: I just want your read on what went wrong at a design level. Not even the business side or anything. From a pure design perspective.
RAPH KOSTER: The motivations were business-driven. The motivations were: we should have a larger audience. We have this huge IP.
Much afterwards, I dug into that and did research, and I found that IPs that come from outside games—we actually should not set any expectations on how well they do. They have a very checkered history. Homegrown IP in games is proven to be a great accelerant. But even back then there was a history of Star Wars games that didn’t do anything special in the market.
So the expectations were probably out of whack for what having Star Wars should have meant. But World of Warcraft had hit. There was all this desire to casualize, to bring in a larger audience, and so on, which I understand. There had been a big investment.
The design decision was to make the game more action-based, and that was a challenge because the networking and the code and all of that had not been designed to do that. There was a retrofit to make that happen, and it definitely felt clunky for quite a while.
But I think more importantly is the community lesson. You had a passionate audience that had come in and adored the game the way it was. They were sticking with it through thick and thin because of what it offered. To them, it felt like a betrayal to change it, alienate it to such a degree.
It’s been called the most infamous patch in all of video game history. Absolutely. I don’t say that lightly. We’re talking about a video game patch that made The New York Times, The Christian Science Monitor. This was literally covered by mainstream press as the worst video game patch ever made.
So many of the design choices made to make it a broad-appeal game were actually trimming away the sandbox bits. You could no longer be the entertainer who didn’t fight. You could no longer keep all of those pets.
In one of the great optics mistakes ever, they didn’t delete everybody’s pets. They left them in the inventory. You could just never take them out. They were trapped in amber. People held funerals for them.
It stands today as an object lesson in: if you build a community that is there for a particular idea, a particular ethos, a particular experience, that can evolve over time in collaboration with that audience, but you should never just rug-pull and change the experience wholesale out from under people.
HOST: Don’t bite the hand that feeds you.
RAPH KOSTER: That’s right.
HOST: All right. We don’t need to dwell on that.
I want to talk a little bit before we jump into Stars Reach. I want to talk about your book, A Theory of Fun. It came out 20-plus years ago, right?
RAPH KOSTER: Yeah.
HOST: It’s taught in game design programs worldwide. It’s translated into nine languages. I think you’ve described it as literally half cartoons.
RAPH KOSTER: It is half cartoons.
HOST: Did you know you were writing that book when you were writing it, or did it become that later?
RAPH KOSTER: The origin of that book is as a talk, which is why it’s half cartoons. I was asked to keynote the very first edition of the Austin Games Conference, which became GDC Austin. It was me and, I think, Mark Jacobs was the other person.
I had been talking with some friends. I’d been talking with Noah Falstein, who worked on Sinistar and at LucasArts Adventures, and I’d been talking with MUD friends like Dave Rickey and others about the evolutionary things that led to playing games. Why do we play games?
So I did a talk about that, and I didn’t have a lot of time to put together the slides. I literally drew cartoons in MS Paint or something for my slides, and used, yes, Comic Sans as the font. That was the original keynote.
The keynote went a little bit viral, for 2003 or whatever it was. It was actually academics and a book editor—Ben Sawyer and a few others. Curt Squire, I think, was the one who pointed it out to Ben, and then Ben got the book idea and approached me at E3 and cornered me and said, “I think you have a book. I think we can make a cool little 50-page thing with a few statements and a cartoon, and it can be an inspirational little coffee-table book.”
He talked me into it, and by the time I was done it was more like a 250-page thing, cartoons and a lot of academic citations and all of that. We published it, and it was very controversial, because the core thesis is that games are important because fun is learning.
Fun is the way in which we learn when there is no pressure, and it’s key to how mammals and other animal types and humans learn. That thesis was controversial. A lot of people said, “No, that narrows fun down too much,” and so on.
Then the book promptly fell out of print because the publisher folded, but it started going on eBay for hundreds of dollars, and more and more people were reading it. O’Reilly ended up acquiring the folded publisher and resurrecting the book. The book sells as many copies last year as it did every year since it came out.
Literally, if you start a game design job in South Korea, there might be a copy waiting for you at your desk on your first day. So it’s gone on to have an outsized influence. It’s still controversial, but a lot of what I said in the book has been more and more validated by science. Theories like predictive processing and so on.
It was a book that drew heavily from cognitive psychology, and quite a lot of the things that were my theory of fun have gotten validated since then. It’s had this fascinating life where it went from game designers to educators and now to AI researchers. AI researchers use the ideas in it as ways to tune gameplay AIs and train AIs.
HOST: Would you make that same core argument today, that fun is about the mastery of problems, and boredom is what happens when there’s nothing left to learn? Twenty-five years later, is that still a claim you would make and adhere to?
I know a lot of people—this may be my wrong hypothesis—they love some things about games right now, maybe comfort and repetition, which can be the opposite of challenge. Not that I’m challenging you, but—
RAPH KOSTER: No, no, no. I have two different responses there.
One is there’s nothing wrong with playing for comfort, but we can call it comfort. We don’t have to call it fun.
The other is people usually define problems really narrowly. I love narrative games. Narrative games are presenting problems. They’re just a different kind of problem than estimating the trajectory of a bullet. They’re presenting psychological problems, social problems, the same kinds of things we study in literature class.
So I always meant “problem” pretty broadly. To my mind, playing a multiplayer game, you are solving the problem of how other people think. If you are playing a narrative game, you’re solving the problem of how the author thinks, and how you think about the story they’re presenting, and what lessons you draw from it. If you’re playing an action game, you’re solving reflex problems and math problems.
All of those are forms of learning, and they’re all super important. In fact, what the latest research on play says is that the biggest skills that kittens and dolphins and so on learn from play are not actually the math ones and so on. They’re about getting along. That is the core thing that games trigger and help develop.
When they take lion cubs and raise them away from others, what they end up being unable to do is thrive in a pack or pride. That carries through. Think about what we encourage kids to go learn to do: please go learn how to play on a team, learn how to share, learn how to take turns.
This is another reason I often say games at their heart are really multiplayer, not single-player, even though the prevailing current in digital has been single-player for a long time.
HOST: We clearly need more of that in this world.
You said it’s controversial, and I’m sure there are plenty of critiques. What is the most fair critique you’ve heard from someone who’s genuinely engaged with the book?
RAPH KOSTER: The one you raised is actually one of the best: what about all the ways to play that are not learning?
I dug into that a bunch when I did a second edition. There’s a really interesting overlap between playing games and playing a game that you’ve mastered, that you know really well.
I remember for me as a kid, the first experience of this was playing an Atari game that I got so good at that I could literally close my eyes and kill-screen the game.
HOST: Which one?
RAPH KOSTER: It was Laser Blast by Activision. It took four or five hours to kill-screen that game, and I could do it with my eyes closed. I don’t know if I would call that fun, but it was profoundly meditative.
So I spent time looking into that. If you look at meditation practices around the world, they are about paying focused attention to a repeated pattern that you do over and over again until it becomes second nature and your mind goes clear.
HOST: That state. Exactly.
RAPH KOSTER: The idea that you can go back to a game for comfort or to use it as meditation actually slots in perfectly.
But I think as a designer, I want to know: am I designing this moment for meditation, or am I designing this moment for learning? Because I’m going to shape the interaction differently.
Whenever I talk design theory, I’m always coming at it from a craft point of view. These are not hypothetical, academic, artsy-fartsy things. They are all about how do I make games better. What is a tool I can put on my workbench?
The idea that if I want to sustain engagement in the game in a fun way, I need to keep presenting novel problems and unveiling new strategies and tactics—that is a tool on the workbench.
At the same time, saying, “Oh, I really enjoy the creativity of rearranging furniture, and it’s very meditative.” I don’t know that I would necessarily call that the same kind of fun. It’s a different thing.
There might be problems in it, like “Oh, the couch doesn’t fit. Now it’s a math problem again,” and then I’ll feel classic fun when I solve that. But if it’s really about the aesthetics, it starts moving away from problem and into other realms sometimes, and I pick up different tools from the design workbench for that.
That to me is the impetus behind all the theory stuff I do: how do I add more tools, not just for myself but for everybody, for other game designers, that help us think about how we craft what we make and how we can better hit the targets we’re aiming at?
HOST: I’ve read or heard that you would like to have some follow-ups, some sequels to that book. Is that something actively you’re thinking about, or even doing in the background? Or is time such a precious commodity, of which I’m sure you have very little, that it’s not on your plate at the moment?
RAPH KOSTER: I have the outline for the sequel, and have been working on it off and on for 20 years.
HOST: Wow.
RAPH KOSTER: It’s called A Grammar for Games, or A Grammar of Gameplay. I’ve had variations there. It’s basically going the next step deeper: how do game systems work? How do they fit together? How do they elicit emotion? How do they elicit learning? Are there problem types that we can go to that reliably give us things?
I’ve given dozens of talks and done research projects and so on around that subject, and a lot of that stuff has filtered its way into game design curricula over the years that various people have picked up on. I’ve had fellow travelers like Dan Cook, the late Mike Sellers, Stéphane Bura, Joris Dormans, who have been working kind of in that same area. But I still haven’t finished the book.
There’s a ghost listing on Amazon for it, but don’t hit buy. I’ve actually even outlined the one that comes after that. But finding the time to do it while actively working on games is challenging.
HOST: Speaking of working on games, let’s talk about Stars Reach. You’ve called Stars Reach “the game I’ve wanted to make for nearly 30 years,” and that is definitely not a marketing line. That’s a 30-year through-line from LegendMUD to now.
What specifically couldn’t you do in UO or Star Wars Galaxies that you’re doing in Stars Reach?
RAPH KOSTER: Quite a lot of things. One of the biggest is that we now have the compute to do the kinds of simulation things we were gesturing at in those games, that we just didn’t have before. My Apple Watch is more powerful than the servers that ran Ultima Online, by a lot.
HOST: Yeah.
RAPH KOSTER: Like, a lot. You could run an MMO on your watch these days.
There’s a famous software quote: software is a gas. It expands to fill the computer. What’s happened is that, for a variety of reasons, many of them just business pressures, we’ve ended up mostly making the same game over and over again. It’s old, still classic. It’s not that it’s a bad game. But, hey, I am one of N classes and one of N player races, and I group up with the other classes, and we go slay monsters, and we get XP, and we get gear, and I collect gear sets, and then I go on a raid and I solve a puzzle with a group.
That game is old. That is DikuMUD from 1991, roughly. I’m personally, as a player, burned out on it. But it also feels like it’s time for MMOs to rediscover their potential.
In a post-Minecraft world, in a post-EVE world, in a post-Dwarf Fortress world, how is it that our biggest, most ambitious alternate realities boil down to that? It feels like not fulfilling the potential.
So the parts that I’ve dreamed about for 30 years are not just, hey, we can actually simulate a world where you chop the tree and the tree goes over and all of the rest, but where the musician has a role, and also where the grass grows. Where the seasons matter. Where water flows. Where you can irrigate the desert, tunnel through a mountain, burn down a forest and watch it grow back. Where the dragon might land in a clearing in the forest and the forest catches fire.
Worlds can be responsive. They don’t need to be static handcrafted sets that are gorgeous, but they’re like stage sets. They’re cardboard. The minute you go around behind them, you realize they’re flat and optical illusions.
We now have the compute power thanks to cloud compute. We now have the technology to have worlds that react to players, that are living their own life, that the world itself is alive. The players therefore can’t predict and wikify everything in them. That opens up gameplay for players that has just outright not been possible.
If we build games that are all about static stage sets, that’s not really meant for the audience to come up on the stage and rearrange and take control. But these are alternate worlds. That’s what they’re for: for you to log in and jump on the stage.
That is what Stars Reach fundamentally is about. We want to go back to some of the ideas of the sandbox, but also modernize, draw in all the lessons that went out to all those other games. All those lessons from whether it’s Animal Crossing or Dwarf Fortress or The Sims or wherever all those casual droplets went, let’s bring them back into MMOs, because they went on to become the foundation of entire genres of game, and really showed that a lot of people want to play doing more than just killing ten rats.
Stars Reach is about bringing back that sense of possibility, that sense of the magic of stepping into a world where you don’t quite know what’s going to happen.
HOST: For someone who has never played an MMO, what does it feel like to log into Stars Reach for the first time? What’s the first hour like?
RAPH KOSTER: In the first hour, you are told that long ago, humanity really started destroying all of their home worlds, whether it be global pandemics—who would have thunk it—or climate change, or peak oil, or you name it. The different human civilizations out there were all doing what we do, which is wreck things often, because we can’t work together or whatever.
It turns out that our whole arm of the galaxy was terraformed by an incredibly powerful ancient alien species. They’re long gone. We just call them the Old Ones. We don’t know that much about them. They terraformed the galaxy. They actually made us. They genetically uplifted us and all the animals and so on. So there is a whole arm of the galaxy that’s perfect for us.
But there we were in our coops dying off. They had left behind these robot servitors, robot servants that were tending the garden. And the robots say, “Well, we’re not allowed to let you die off.” So they have let us out, and now we’ve formed the Transplanetary League to go adventuring.
So you actually wake up as a brand-new immigrant into the garden. You show up at a Transplanetary League base on a moon called Haven, waking up after having been brought all these light-years from where you came. You were turned into goo along the way. So they say, “Well, we don’t quite know what you look like, but why don’t you redesign yourself? Nobody will know. So if you want to make yourself prettier, go for it.” That’s kind of our excuse for character customization.
It sets you on your path. You come out, and the moon of Haven is in danger because some of those experiments that the Old Ones left behind are quite dangerous. They left behind a planet-eating goo hive mind called the Cornucopia that drops spores as meteors onto planets and infects them to swallow them up into sort of a biological Borg.
You’re not quite sure if you can trust the servitors because they seem friendly, but they also say things like, “Please don’t disobey any of our rules or we might take you for molecular disassembly,” which does not sound very pleasant, honestly.
So you show up in that world, and the TPL gives you your starter blaster, your starter spaceship, trains you in any of a number of professions that you could get going in. They say, “You might want to map a planet. Try mapping Haven. Or you might want to go hunting for meat for the chef. Or you might want to gather up loot so you can sell it. Go mining. Go find us rare crystals at the bottom of a cave.”
You try those out and you decide which ones you want to do, because the game doesn’t force you to do any one of them. You can choose. You could do all of them if you want. It is a classless game system.
Then, once you learn the ropes on Haven and the deadly planet that it orbits—Haven is a moon, remember—there’s an unstable world called Crucible. Lots of lava. Don’t fall in. Even getting close to it, you will catch fire. Actually, it’s hot. It’s lava.
HOST: It’s hot. That’s right.
RAPH KOSTER: Then you can venture out into the larger galaxy. That galaxy is as many planets as we need for the players we’ve got, because we do generate them. We create wormholes. We open up new worlds on the far side. Players can go explore them, kill monsters and gather loot and tunnel and dig and burn down forests, all of that, and bring the stuff back home.
Or they can move in and build buildings and establish settlements and even elect a mayor and say, “No, this is our place now. Here are the people who are allowed to build here. Here are the laws. What sort of PvP is allowed here? Who has permission to build, to dig?” All of that.
That’s where I call back to the Discord server, or the subreddit, or the private server that you might run for Valheim or whatever, because all those planets still exist in one MMO galaxy. You can visit them all, fly through space, explore wherever you like, but you and your guild, you and your friends, can own one of them and make it yours.
You can trade and exchange with those on the other planets, but it is one connected galaxy, not a whole bunch of separate shards, or “I log into server X,” or whatever.
HOST: One connected, fully persistent galaxy.
RAPH KOSTER: You dig a hole, it’s there forever. You drive creatures extinct on a planet, they’re gone. You might revive them if you got enough genetic samples before driving them extinct. If you abandon a world, the wormhole to it might collapse. It might be lost. And that’s okay, because we can always discover new ones.
HOST: Sweet. Sounds fantastic, and super ambitious.
RAPH KOSTER: I didn’t even mention the craziest part, which is everything that you see, every tree grew there. Trees actually go from sapling to maturity and then die. Every rock, every cubic meter of the world is modifiable, minable, deformable. The rivers can be rerouted. Water flows. When it flows over the dirt, it washes it away. You end up with a gravel riverbed. Reroute it through the desert, the desert will bloom and plants will begin to grow.
Everything is simulated. Every cubic meter of the entire game, even in space. All of it is simulated, modifiable. You can come to a virgin planet, burn and chop all the trees down, kill all the monsters, pave it over, and build a shopping mall. You can do that.
HOST: The architecture story has gotten much sharper publicly. You’re talking about a single global server, no resets, partly cloud-native. For someone who has shipped on the architecture of the 1990s and 2000s, what does it actually mean to design content for a world that doesn’t reset?
RAPH KOSTER: It’s definitely different. You do rely more on procedurality. You have to shift your emphasis in the same way that all designers of simulation games do.
When the Adams brothers are working on Dwarf Fortress, they don’t think in terms of, what story can we tell? They have to think in terms of, what stories can emerge, and what stories can players tell?
When Brian Bucklew is working on Caves of Qud, same idea. It shifts how you think about the systems you design so that you aim for emergence. You can have set pieces that you’re salting in there, but it isn’t a game about the story you’re telling. It’s a game about the story that players are discovering and telling themselves.
So when you create the content, you think of creating content that is reusable, that is parametric. That means you define the bounds for it, but it can be very different depending on the situation in which you find it. It isn’t just the same little set piece over and over again. The set piece feels like it organically belongs there because it adapts to the environment in which it was spawned.
The big thing that changes everything: this would not work if we couldn’t generate zones on the fly and throw them away.
It sometimes feels like that’s one of the things that is hardest to wrap your head around. We’re used to MMOs having a map that grows, but not one that changes. So the idea that, no, even the map itself is an organic entity is odd.
You might come back to the game and go, “Oh, wow. I really miss this place.” It’s like going back to the city you moved away from 10 years ago. It might be completely different, or maybe it’s gone. You have that moment of, “Oh, wow. This is a continuous tapestry.”
If it weren’t for the confluence of cloud—we run the AI for doing this in the cloud. It’s cellular automata AI, not gen AI. We run that in the cloud. If it weren’t for the confluence of that, the ability to stream the data, video cards, and computers getting good enough for us to take that data and build the entire environment on the fly around you, because that’s what we end up doing. We construct the meshes for everything around you, the terrain and all of that, on the fly, because it can change on the fly.
The tech to do this just wasn’t there.
Clearly, we do borrow technical ideas. I did a project called Metaplace back in the late 2000s that ended up being the backend for Club Penguin. It ended up being part of Disney. This was the same time period as Second Life and all those things.
There are architectural similarities between what we’re doing and things like VRChat or even Roblox. They also make use of cloud and streaming and so on.
So yeah, I’m lucky. This is my fifth MMO engine from scratch. Not all of them were game; Metaplace was a UGC platform. All of those lessons really go into how we architect this, and then the game that it enables. The tech is not there to be cool tech. The tech is there to enable gameplay that you just can’t get any other way.
HOST: You named some big things coming after early access, including the Cornucopia, the planet-eating hive mind, factions and political structures, PvP. Walk me through the logic of what ships first versus what comes later in a game that you’re building to evolve.
RAPH KOSTER: What we just announced is early access this summer. We still have a fair amount of stuff to put in the game before that. We know that early access has to be at a certain level, so we still have lots of work to do. We’re working very hard to get all of that in before we open up EA.
EA will still very much be early access. All the core loop type things should be there: combat and loot and getting better gear and making your character more powerful. The loops around merchants and item decay and crafting and combat and exploring and mining and all of that kind of stuff need to be there so that the economy is fully functional and the different ecosystems and roles are in place.
Actually, Cornucopia will be there, because that’s one of the threats that you fight against. If you get the chance to come check it out soon, you might even get to fight off a spore or two.
There are things that we expect to continue to add during early access. Adding additional profession depth. There’s all kinds of nuances and content that we want to keep adding. More monsters, more art. One of the things about making a game with this much crafting is you have to make art for a zillion individual little components.
So at early access, I’m sure that there will be icons in the game that have the name and “art pending.” That’s some of what we will flesh out.
The game should be fully playable. We’re going to continue to add. Players will be building buildings. They can govern planets right now. That’s already been in the game for months and months.
A lot of the big ambitious things are already there. Space flight you can only do right now in a space suit, but we have given players brief tests of the spaceships that are coming. We probably won’t launch the full spaceships feature until one of the early updates in EA.
The full fruition of this is, EA leads to a 1.0. There is absolutely a gap between the EA version and the 1.0 version. But the game should be fully playable, and everybody who plays during early access will have access, as part of their purchase price, to owning land inside the game.
HOST: Oh, that’s cool.
RAPH KOSTER: At 1.0 later, that’s going to be something that people will have to pay for. Because yes, virtual land is infinite, but it turns out that infinite virtual land is very finite expensive servers that we have to pay for. There’s cost there. So that’s going to be a paid feature eventually, but during early access, if you buy in now, it’s in the bundle. You get it for the duration of EA.
HOST: Just for the listeners, at the time of this recording, the announcement was literally this morning that the window for early access was going to be this summer. Plus, there’s a new trailer, which I highly recommend everybody check out. Super cool.
On a personal level, what does it feel like to put a window on something you’ve been carrying, legitimately, for three decades? How are you feeling this morning?
RAPH KOSTER: I was thinking of the Latin. I’m going to mispronounce it: alea iacta est. The die is cast.
It feels like this is the moment when you step onto the waterslide, or the moment where you step off the platform on your zipline. From here, it just gets speedier and scarier and a little less in your control. But we still have that whole zipline mile to go, right? There’s a lot to do that we’ll be building until that moment when it’s actually out there in players’ hands.
The other side of it is, it is a lot like urban planning a city. You know that what you’re setting up is just the start point, the foundation for what you hope are years to come.
It’s a cliche now: MMOs are never done. But they have to never be done. That’s part of what keeps people interested in them, that sense of ongoing evolution.
It’s almost like, you’d never do this if you were running a race, but we’re currently sprinting to the start point of the marathon.
HOST: Just to get to the start.
If you had to describe Stars Reach in a sentence to someone who’s never played a sandbox MMO, what would that be?
RAPH KOSTER: Imagine a game where you can log into this alternate galaxy and be who you want to be. If you want to be that intrepid smuggler type with the beat-up cargo spaceship, okay. If you want to be the person who runs the tavern with a bunch of weird people in it, or maybe the one who steps up on stage and plays their instrument, that’s available to you also.
But crucially, and unlike any other game I think you’ve played, it’s a game where when you go and you touch the world, everything just does what you expect. Dump water on the ground, it gets muddy. Step in the mud and you might get stuck in it.
Even simple things like that, but also cool complicated things like, wow, during the winter is the only time that the deadly ice snakes tend to come down from the mountains.
I would ask them, “Did you ever watch any of those isekai anime? Did you ever see a movie where people got sucked into a virtual world?” There’s loads of that stuff. Now, they might not know sandbox MMOs, but all of that media, that is what those are portraying.
The promise, the dream of alternate reality is not “kill ten rats,” right? So I would say to them, it’s your chance to step into an alternate world where you can be whoever you want to be, where you can do things that you couldn’t do, and you can do it with other people.
HOST: Brilliant.
I’ve taken up a lot of your time. I have one last question I want to ask that is not related to Stars Reach, but something that I like to end every episode of the podcast the same way. The name of the show is Replay. So I want to ask you: if you could go back and replay one game, with the caveat of experiencing it completely fresh for the very first time, which game would that be and why?
RAPH KOSTER: That’s really hard.
HOST: I know. I’m sorry.
RAPH KOSTER: I’ll get in trouble if I rattle off four or five, right?
HOST: You can. You make your own rules.
RAPH KOSTER: Okay. The game that always comes to mind is M.U.L.E., which was a classic 8-bit game. It pioneered multiplayer co-op strategy. But the thing is, I still go back and replay that game. I almost wore my M.U.L.E. T-shirt to this interview. I replay it periodically. I make people on my team play it, actually, and it serves as an inspiration for Stars Reach.
I would go back and replay Journey. I would go back and re-experience Edith Finch. I’m moving through decades here.
And that first magical moment when I first logged into a MUD, where people were talking about, “Watch out, don’t go around that corner. There are mobs over there,” and I thought they meant a mob of people. I didn’t know the lingo yet.
I’m moving around this world, and other people are moving around, and then a rabbit crawled out of my backpack and turned into a person. It was just a mind-blowing experience that we take for granted today.
One of the things I always think is probably true of all of the games I just listed: they were all moments of stepping across a threshold into an alternate world that blew my mind in some way, and that felt magical, and that felt like a possibility space larger than what I expected given the size of the disk or the screen or the install.
To me, that is one of the cardinal virtues of immersive games. We talk a lot about how it’s all old hat these days. Everybody’s connected. But I try to always remind myself that we are playing with lightning built on top of machines made out of sand, and we’re getting magic out of it.
I try to always find ways to see the magic. That, to me, is what I crave when I come to these games.
HOST: Where can people find more information about Stars Reach and yourself?
RAPH KOSTER: Stars Reach is easy: starsreach.com. There’s also a YouTube and so on. We actually even have a Linktree. Search the Stars Reach Linktree. It’s off the website, and it’ll link you to all the others.
HOST: We’ll post all that in the show notes as well.
RAPH KOSTER: I’m also easy to find. My blog, raphkoster.com, has been running since the late ’90s. It’s accumulated millions of words. There’s all kinds of things on there about all of my projects, whether it’s the games I make, the emulators I help maintain because I do a lot of retro emulation work or have done, music I create, you name it. All of that stuff is there on the website. There’s loads of material to go dig through.
HOST: Amazing. Frankly, I say this to a lot of our guests, but I would love it if you would come back, because I really do feel like we only scratched the surface. We didn’t even get to talk about emulation. We barely talked about metaverse, or all that other stuff you’ve done in the background, even design theory. I would love to talk more about that. So when the time is right, and when you have that time, that’d be great.
Thank you for giving me your time today, given all you have going on. I really appreciate it, Raph. Thank you so much.
RAPH KOSTER: Yeah. Let’s book more time as soon as I’m through this announcement.
HOST: Let’s do it. For sure. It’s been a pleasure, though. Thank you.
RAPH KOSTER: Thank you. Thanks for the time.
HOST: And that is Raph Koster, CEO and creative director of Playable Worlds, author of A Theory of Fun, and one of the architects of every online world worth talking about. Links to Stars Reach, his writing, and everything we discussed are in the show notes below.
If you wouldn’t mind, please follow or subscribe to Replay. It really helps out wherever you listen. I’m Greg Off, and this is Replay.