The Evolution of Online Worlds

This talk was originally created for a Casual Connect conference in 2018. When I got there, I discovered that my hour-long slot was suddenly only 30 minutes long including time for questions. Needless to say, the talk was hurried and mangled. I therefore never put it up here on the site.

In 2025, I was asked to do a session for the local IGDA chapter. I dusted it off, and this time had the full hour and was able to capture video as well!

Video

PDF of the original version

A couple of slides differ and the graphical treatment is slightly different, but this is a version you can read, because there is text on the slides.

PDF of the video version

This version does not have any text on the slides.

Transcript

A Brief History of Online Worlds

HOST: Welcome, welcome. Thanks so much for coming out, guys. Round of applause.

My name is [inaudible], and I’m part of the IGDA here in San Diego. Look at this community. Just absolutely wonderful. I do want to thank a couple of people. I want to thank Mesa College for letting us be here. I want to thank all the rest of the board members for organizing this event. I want to thank Playable Worlds and Raph for being here as well. And of course, I want to thank everyone who’s demoing a game today.

The industry is what it is, and just looking out at events like this, I’m seeing all the passion and I’m seeing all the knowledge that’s being shared. What this tells me is, here in San Diego, we are keeping it alive. So round of applause for that, guys.

[Applause.]

So without much further ado, here to give a talk about the history of online worlds: he has been making games for 35 years. You might know him from designing Ultima Online, Star Wars Galaxies, A Theory of Fun. Yeah, books. He’s been named by GDC as an Online Game Legend. Now he is with Playable Worlds making an MMO called Stars Reach. Please put your hands together for Raph.

[Applause.]

RAPH KOSTER: You ready for a really dense history lesson?

One of the minigames that we can play throughout this talk is that you can shout out the year you were born as we go, because we’re going to be hitting an awful lot of years that I think people don’t realize how far back this all goes.

I have been making online worlds for 35 years. I got my start when they were just text, and they were already several decades old then.

One of the things I think happens in the game industry is that it moves so fast that our history gets lost, buried, forgotten. You look at “greatest games of all time” lists and somehow they’re all in the last five years because the people writing the lists have only been playing for five years. That kind of thing goes on.

Me, I’m a bit of a student of all of this. I always want to go back through the history and ask: why do we make things the way that we make them? What are the pivot points that led to us choosing to do things one way versus another?

So when I think about online worlds, I actually have to start before computers.

Anybody a fan of this guy?

So writers have been making secondary worlds, the idea of alternate realities, for centuries. Centuries and centuries. However, the idea of worldbuilding one to the degree that it was internally consistent, we can pretty much blame this guy for it, because he was, among other things, a professor of linguistics.

So he started building the history of his world, the languages of his world, the mythology of his world, and tying it all together in a super-realistic way. Ever since, this has been our model.

We’re going to do this by chapters. This first one is secondary worlds.

The moment that people started making cool secondary worlds, people started writing fan fiction about them. That started super, super early. There were fandom groups among sci-fi fans forming as early as the late 1920s and early 1930s, in New York in particular. They kept in touch via zines that were typewritten and mimeographed, because there was no photocopying yet.

However, the first widely spread-around fanfiction zine was actually one called Spockanalia. This is where we get the term slash fiction from. Who here knows what that is? How about shipping? You know what shipping is?

Okay. Slash fiction comes from Kirk/Spock shipping. It is the origin of LGBTQ shipping in fan fiction. Pretty much it all comes from Star Trek.

Fan fiction is the first time the players got to really play in an alternate world. So we wouldn’t have online worlds without fan fiction. This is your first moment of participation.

Now, we have to remember, even with all the zillions of video games that we make today, for 99.9% of humanity, games have been multiplayer.

Single-player games prior to sometime in the 1970s consisted of solitaire, chess problems, stuff you made up. Games are multiplayer kind of definitionally, historically.

Computers are really the thing that allowed us to make true single-player games. Until a certain point, games ended. Most multiplayer we can think of today as session-based multiplayer: you go into a game, you play it, and then it ends.

Computers enabled single-player games to do new things. The early computer games were all multiplayer because there wasn’t enough CPU to write an AI to play as an opponent. So if you go look at the earliest computer games, your Pongs, your Tennis for Twos, your Spacewars, you couldn’t play against a computer. You were playing against another person.

The first time we saw networking, which is a crucial piece of online worlds, was 1969.

I’m still not born now.

This was done by a guy named Rick Blomme, and he is networking—this is on a PDP-11, I believe—a game called Spacewar, which is sort of a proto-Asteroids with a gravity well in the middle, and is really hard to play.

I highly encourage everybody who’s interested in the history of video games, or learning to be a game designer, to get yourself a MAME install and play the games chronologically, oldest to newest. It is a great way to learn. You will come across this in your first few days.

All right. So, alternate worlds. Session multiplayer. Our third key ingredient is going to be the role-playing game, invented not by Gary Gygax, but by Dave Arneson, the late Dave Arneson, who’s a much nicer guy than Gary.

He’s the one who really took what had existed with existing miniature tabletop rules—there’s a whole line of history there that I’m not going to trace, and it’s a fascinating story. Basically think turn-based strategy with miniatures, going back through Kriegsspiel, which was used to train actual militaries. H. G. Wells—yes, that writer guy—was a huge proponent. Science fiction fans of that turned out to be part of that sci-fi fandom in the ’20s and ’30s and founded the first government think tanks. It’s a really weird history.

But anyway, Blackmoor was the first real role-playing campaign, because they added fantasy rules first, and then people started role-playing in it, and they said, “Okay, you can keep doing that more or less.”

The big thing this adds isn’t just the idea of being an alternate person in the game. It’s also that now, for the first time, we have what we might call a persistent world. We have a campaign instead of just a session like in the previous column. So now players could actually live in this secondary world.

I’m born by now, because I was born in ’71. No? ’71? All right.

Few people realize just how old esports are. This is the first known esports tournament. It was a Spacewar tournament held on a campus, probably at MIT. So there you go. It’s really much older than people realize.

This is going to matter because we’re going to come back to this competitive line over and over and over again.

There was, believe it or not, a university-based system with plasma screens that could only draw orange, called PLATO, that about a dozen campuses were connected to in the ’70s. PLATO today is mostly forgotten, but it was the place where many, many folks first really got to play online games. A huge amount of player pioneers ended up in the video game industry. PLATO was where a lot of this happened.

PLATO is also where most of the tropes of the internet got invented, including the idea of real-time chat. There were proto-forums called notes files, and of course email was already a thing because the internet at this point was already a decade old. It’s just nobody but the military could get it.

So early versions of instant messaging, forums, multi-user chat rooms, all of these things were popular on PLATO even in the early 1970s.

Role-playing really started booming, of course, with this. And now you can say Gary. Dave and Gary worked together on this stuff. Dave was the RPG guy, and Gary was the “let’s add more tables into the DMG” guy.

D&D really popularized and set a template for what the RPG was. For better or for worse, we still use classes, and we still have levels. Even though systems like GURPS and others came along pretty quickly that said maybe those aren’t the best way to build an RPG, we’re stuck with classes and levels.

I’ll tell you right now, I think they aren’t the best way to build a persistent-world RPG, but that’s another day and another talk.

Now we get our next big genre. Empire came along on PLATO as early as 1974. This game had real-time networking, which meant that your keystroke lag was only ten seconds.

So, turn-based, real-time, what’s the deal, right? This was the birth—and there were a whole bunch of these, by the way. These games were usually hosting no more than a dozen or so people at a time. It was very, very laggy. You could watch the screen redraw in real time because the instructions for the graphics were being sent over the wire. Sort of the GeForce Now of its day.

By this time period, we also had the first online FPS. It’s called Maze War. This was also running on PLATO. It was a session-based online game where players wandered through this maze and shot each other. So take that, Quake.

The thing about all of these early things is that they were happening when almost nobody was seeing them. That doesn’t mean that forward-thinking folks weren’t seeing the way that technology trends were going and anticipating what might come.

So the first real science-fiction treatment of what we would think of as the internet was actually written clear back here in 1975, in this novel called The Shockwave Rider by John Brunner. It was the first fictional description of what would, within only a couple of years, be called cyberspace.

The key thing that it described and anticipated rather presciently was identity hacking, which has actually been a key part, of course, of everything online ever since, whether it’s avatars or catfishing. Brunner was pretty sharp to foresee that this would be, in many ways, the defining characteristic of the internet: that somebody’s identity would be fluid and not fixed.

In 1976, there was a fellow named Will Crowther who really enjoyed going down into caves and avoiding getting stuck. He was a spelunker, and he was also in one of the early D&D campaigns with other fellow spelunkers and whatnot on the East Coast. In particular, he had been exploring Mammoth Cave in Kentucky.

He decided to use a computer to use text-based rooms, which for those of you who haven’t played in text—how many of you young folks have played a text adventure? How many of you young folks have not played a text adventure? Probably good.

Imagine clicking through web pages and the links at the bottom are “go north,” “go south,” “go east,” “go west.” It’s kind of like that, except you actually typed words. You typed “go north.”

He made an explorable, map-accurate version of Mammoth Cave and called it Colossal Cave Adventure. He immortalized the phrase, “You are in a maze of twisty little passages, all alike.” He also stuck in there things out of his D&D game, like there was a nasty little dwarf with a knife, and there’s a dragon that you have to kill in one of the classic puzzles of gaming to this day.

“Kill dragon with sword.”

“Kill dragon.”

“Kill dragon with what?”

“Your bare hands.”

Yes. And that worked.

I hope the spoiler is okay for those who haven’t played it.

This was the birth of the adventure game genre, but also in many ways the birth of computers simulating places and spaces.

Now we have to rewind to that alternate-reality thing, because now we get the third major, call it mass-market secondary world after Tolkien and after Star Trek and the previous ones.

Star Trek had the main show. It had an animated series. You could definitely go get action figures for it. There was fan fiction. So we could say that it had gone transmedia.

But we had never seen anything go transmedia like this ever. I mean, I saw this thing thirteen times in the theater on release, right? I was this big.

Comics characters had moved between page and screen. We’d had Buck Rogers. We’d had Tarzan. We’d had the Lone Ranger. Those things had happened. But we had never seen anything quite as all-encompassing as what Star Wars managed to do.

It turned out to be a really powerful model for what a fictional world could be, in part because it had a lot more world to it. Buck Rogers, Flash Gordon, Lone Ranger, the world was not highly developed. Star Wars, people started immediately just blowing it out. There was Expanded Universe stuff and all of this, and it became the model for what an alternate world could be.

All of this goes into the soup when a couple of students, Roy Trubshaw and Richard Bartle, students at the University of Essex—which meant that they were not from the upper crust of British society. They had the wrong accent and all of that.

Roy had this idea first as an engineering challenge. He’s like, “I wonder if we could take—” They didn’t like Colossal Cave Adventure, which was called ADVENT because file names could only be six letters long on those computers. They liked a clone of it called Dungeon. D-U-N-G-E-N.

So they’re like, “If we use shared memory, we might be able to allow multiple people to wander around those text rooms at the same time. Then we could put ‘Fred is here’ and you could say ‘hello, Fred’ and Fred could answer you.”

They married the PLATO-style chat, which they had never seen, were completely unaware existed, with the idea of persistent worlds in an alternate reality and these text-based adventures, and invented a thing called MUD: multi-user dungeon. That is the invention of the virtual world.

Pretty much all the virtual worlds that you have played today descend in a direct line from MUD.

Still with us from those days: the concept of having a player character, the concept of having gods of the world, spells, inventories. All of these kinds of things were tropes that ended up developing out of MUD.

Richard took over pretty quickly from Roy and even started commercializing this stuff, but it was only really available in the British Isles, because they did it on the University of Essex computer system, and internet connectivity was still really rare.

One of the things that’s interesting is because he came from the wrong background to end up in a position of power, despite being a brilliant guy, he wanted to make sure that the worlds they made were more egalitarian and equitable than the real one.

So he chose to lift levels out of D&D because in the world in which he lived, in 1970s UK, somebody with his accent was not allowed to level past six. He was like, “No, no, no. Anybody can come into this game. Doesn’t matter if you’re a man or a woman. You can come into this game and level all the way up to 20 and become a wizard and gain godly powers.” As far as he was concerned, that was a political statement.

Further, he logged in one day. Bear in mind, these are on timeshare computers. Everybody could see who else was logged in. He logged in and he made a character named Polly. Then he started playing the game as Polly. He started getting these messages from people like, “Are you a girl?”

And he said, “This is another world. We can try out and be whatever the hell we please.” It was a political statement all the way back in 1978.

Blows my mind.

But internet connectivity is still rare, so MUD was mostly confined in the British Isles. He actually started trying to commercialize this stuff, and there are examples of commercial MUDs running from as early as 1981, 1982.

Now in 1982, the world was starting to get a little bit unsettled in some weird ways. Atari is selling games. Apple is selling computers. Corporations are getting really big, and people start getting nervous about what could end up happening. This new tangent form of science fiction starts to develop that questions all of this.

The epicenter of this was actually a writing workshop in Austin, Texas. A couple of writers, the most prominent of them William Gibson, workshopped a book called Neuromancer at that writing workshop. First there was a story called “Burning Chrome,” then Neuromancer.

We’re talking wraparound shades, and a classic opening line that alas does not work for anybody today. This will be a dividing line in this room: “The sky above the port was the color of television, tuned to a dead channel.”

Raise your hand if you think that’s blue.

Raise your hand if you think that’s black-and-white static.

Unfortunately, he meant black-and-white static. TV didn’t go blue back then.

More writers were soon to follow: Pat Cadigan, Bruce Sterling, a bunch of others. Turkey City, by the way, kept going. I was a member of the Turkey City writing workshop, actually, with Bruce Sterling and Cory Doctorow and those folks, clear in the ’90s while we were making UO.

All of this stuff that they predicted in cyberpunk has kind of come true in all the most unpleasant of ways. Constant surveillance by giant corporations, controlling what you think, even down to the wraparound stupid shades. All of that’s here.

All the things they warned about, the government coming in and tracking down the hackers: there was a famous incident in 1987 called the Hacker Crackdown. Bruce Sterling wrote a book about it. They actually raided Steve Jackson Games. Yeah, the guys who make Munchkin.

They raided them because, among other things, the hacker culture and the phone phreak culture—people could hack the beeps on phones. If you hacked the beeps, you could play back beeps on a cassette tape and dial long distance for free. Stuff like that was going on. So the FBI actually raided these guys with just egregious violations of civil rights.

It was one of the first battles between law enforcement and the hacker community.

All right. BBSes started to flourish. People couldn’t get to the internet, but they could dial up to a local number where somebody had left their Commodore 64 on an open phone line. That Commodore 64 could handle running a few very simple games.

Some of those included early online-ish RPGs that you could play. Legend of the Red Dragon, others, with TradeWars. Some of these evolved into other things. TradeWars was a game of space trading and economic strategy. There are elements of TradeWars even in EVE Online today, because these ideas kind of persist.

Now that we had some of these computers in people’s hands, it was time for somebody to do something crazy.

How many programmers in the room? Any of you ever used JavaScript?

One of the crazy people who built this, Habitat, was Doug Crockford, who, if you’ve worked with JavaScript or JSON—JSON was invented as an online-world technology. You can blame virtual worlds for JSON syntax, because we needed a way to send digital objects over the wire that were self-describing. Then it turned out to be useful for other purposes rather than games.

Randy Farmer, Chip Morningstar, the other few movers behind this project: Habitat was cramming an entire graphical virtual world onto a Commodore 64. It came out in 1985 and then promptly closed almost immediately.

This was the first time that somebody said, “Let’s make a virtual world that isn’t about fantasy adventuring.” It is also the birth of the graphical virtual world. This is the first one.

You unfortunately can’t really go play it online today, but it is server-emulated. You can go to museums and play it at the museums up in San Francisco.

By 1986, we were getting online flight sims, where people were blowing each other out of the sky in games like Air Warrior, which was done by a company called Kesmai that went on to also pioneer early graphical online worlds and then eventually get swallowed and digested by EA, as is the fate of so many others.

Cockpit view, multiplayer battles, full crossplay between Atari, Amiga, Mac, and PC. Already working clear back then.

These games were running on online services before the internet was widely available. We had essentially private internets. These were America Online, CompuServe, The Source, GEnie, services like this. You paid by the minute, and then you paid extra by the minute.

If you played games back then the way you play now, you were looking at four-digit bills a day. They made a lot of money.

Back then, one of my high school gigs was doing LexisNexis research for the local PBS station. Why? Because I could type on a computer fast enough that if they needed me to find the name of the lawyer who did some case that they needed to mention in a news item, I could find it in less than ten minutes, which meant that the research cost less than $100. So they would bring in teenage computer kids to do it because it was cheaper than having the adult do it.

The internet was not yet really internet. It was really not mass market. However, it was around. Campuses were getting hooked up and connected more and more and more. Apple was going out to college campuses and giving them thousands of computers to put all over the campus. Networking was really happening a lot.

This is the time period in which the kids were all growing up playing Hunt the Wumpus on ancient computers in their labs and high schools. Networks were starting to get going, and the internet was primarily available to students, primarily technical or scientific students on campuses, the military, research labs, things like that.

But it was around. If you wanted to send an email to somebody, you actually had to know the name of every major router in between you and them. Their email address was “send to Mesa College router, through NSF gate, to SDSU, through Bitnet gate, to Stanford,” like you had to type all the routing yourself in the email. But the email would go through, and you could even do live chat with them. IRC dates clear back to then, the same IRC that people are using now.

Now MUDs were not standing still. You connected to them using a protocol called Telnet, which is actually the protocol that still to this day underpins the entire internet. But they don’t let you use it because nobody trusts you anymore. I’ll explain why in a bit.

MUDs started to break into different sorts. These were essentially engines for MUDs. Don’t think the rendering-engine part; think the physics, the simulation, how they manage rooms, that kind of stuff.

LP MUDs was one sort. It was fairly combat-based. But some other folks took the idea of, “Why don’t we make stuff that’s mostly about chat and creativity and alternate worlds?” and they made a branch of it called TinyMUDs that was much more social.

Because LP MUDs were kind of super powerful but hard to use, they kind of begat this new thing called DikuMUDs, which was actually the most successful version of MUDs over time. It was the simplest, very template-based hack-and-slash. You had classes, races, tanks, nukers, healers, levels.

You have played this game. How many of you played WoW?

Strip the graphics off WoW. It is a Diku. Literally it is a Diku. Skill rotations, cooldowns, stun, all of that crap. It plays the same.

Because it was templated, it was super easy to just throw up a new one. So everyone did. There were thousands of Diku-clutters. People banned them like a plague because of that.

Then came the MOOs, which were creativity-based worlds. You notice these all got invented within a few years of one another. This was a techy name for a magical idea. It stands for MUD, Object-Oriented, which was this brand new programming fad at the time.

What this meant was, what about a world where you could type in code while the game is running as a player and conjure up a magic sword? A world that you could modify from inside itself, which had been in sci-fi as a vague idea, but here it was existing for real.

To this day, these are the three kinds of online worlds that you tend to see. This is it. They got templatized clear back then.

They also started attracting the attention of the mainstream press because there was one world in particular called LambdaMOO where a very unsavory, trigger-warning kind of incident happened, where a person used that ability to magically conjure code to make an exact duplicate of another person and then have them perform obscene acts on themselves in the middle of the first room that people entered.

It became a cause célèbre as the first virtual rape. “A Rape in Cyberspace” was the title of the article that ended up in the Village Voice and brought this to awareness.

LambdaMOO then went on to say, well, do we turn this person into a toad or not? Because after all, technically that was legal. And there happened to be a journalist embedded live, who wrote a book about it that you can still buy today. Highly encourage it if you want to see the origins of internet history.

This led to the first doxings, people getting fired for things that they did in an online world that only had a hundred players. That led to them declaring the first virtual democracy. The admins abdicated and said, “From now on, you can run the world yourselves.” It lasted less than six months before it all went to hell.

Ground zero for a ton of internet culture.

But that wasn’t all. People invented combat games that mandated roleplay. You were required to submit an application to be able to play this game, and if your character wasn’t described up to standard, they just wouldn’t let you in. So think of it as almost like online LARPing with full PvP.

This game, which almost nobody played, is one of the most important games ever made, because this was a MUD that experimented with the idea of simulating stuff. You could actually grow a crop in this MUD. You could buy a building. This is the very first—the term didn’t exist until later—this was the first sandbox.

Almost nobody played it.

We also got the first persistent PvP worlds, aptly titled Genocide.

Then this writer based up in Seattle, who was part of the cyberpunk crowd, wrote this book that became a massive bestseller. He claims he coined the word avatar, but actually Habitat had done that in 1985. And suddenly everybody was talking about virtual worlds, even though nobody could play in them because they weren’t around.

The metaverse first came operational in 1992, right alongside Snow Crash. A little earlier than you thought, right?

It was really simple. A bunch of LP MUDs said, “Why don’t we make it so that if you walk through this exit on my MUD, it automatically logs into that other room on your MUD.” And they hooked a bunch of them together. So it wasn’t just a metaverse. It was better than Roblox. It was a distributed, user-owned metaverse in 1992.

It didn’t click then either.

By 1993, these old protocols—the absolute base one is Telnet. Open a socket to another computer and type some text into it. If you knew how, you could open a socket and type in the header for an email and send an email by hand through Telnet. You could open a Telnet socket to port 80 and identify yourself as a browser by typing the user string for a browser at the top and get back a web page.

Telnet is what’s underneath SSL. Telnet is the base protocol of the internet. All these others are stacked on top of it. You might still see this one sometimes if you hover over a file you’re going to download: File Transfer Protocol. Gopher is what we used instead of the web before there was the web. Mailto is still used whenever you hover over an email link.

But it didn’t matter because somebody came up with one that ate all of the others: Hypertext Transfer Protocol. Boom, all the others died. It took about three years, but they all died.

Of course, Mosaic was the first browser to offer graphics. We had one called Lynx before that. And Gopher was better. So we kept using Gopher until Mosaic came out. Then we all switched. Now the internet had pictures.

And now we don’t even call it the internet. We call it the web. Even though the web is actually just this layer on top, and if you know where to look, there’s lots more internet to go find if you’re willing to get outside of your browsers. People don’t know it’s there. Some of that stuff today gets called the dark web when it’s actually just the basement.

As soon as there were pictures, people said, “Well, then can we send some pictures over with our text MUDs?” And people started coming up with proto-MMORPGs. Back then, we just called them graphical MUDs. There were dozens of them that got going post-Mosaic.

Some of these, like this one, this is Nexon’s Kingdom of the Winds. This was done out in Korea, where you had the advantage: small country, densely populated, easy to wire up, and as a result, their internet got better way faster than it did here. By 1996, they already had this kind of graphical MMO, but it wasn’t technically MMO, massively multiplayer.

The massive part: then MUDs capped out at 250, but in practice you almost never saw that many. It would be like 50.

This is in contrast to games that were just beginning to get popular, like Doom and Quake, mostly played as LAN games with a few other folks in a room, but now had full 3D graphics. Online sports are developing, but people were playing this using hacks to tunnel them over the internet using things like DWANGO and Kali and services basically dialed into locally that happened to be internet providers and hooked you up.

ISPs, or internet service providers, were springing up as a business gradually.

Now, if we go back to the RPG line, at this moment in game history RPGs were “dead.” They had become extremely expensive, very niche. Most of the new projects were failing. The Might and Magic brand was pivoting over to be a strategy game. The Ultima series had just had a massive flop. Ultima back then was what Final Fantasy or Elder Scrolls is today. It was the premier RPG line. So they’d been written off commercially.

Diablo, which was actually taking one of those old text-based mainframe games—Rogue, Hack, you’ve heard of those—Diablo started as, “Wonder what would happen if we stuck isometric graphics on Rogue?” That was the birth of Diablo. And it was a huge hit. It used lobby-based games, no more than four people at a time.

There were predecessors. There were other games such as Meridian 59, The Realm, and so on that were doing graphical MUDs at around the same time.

But I’m going to go ahead and get braggy, because this past week we’re celebrating the 28th anniversary of this game.

It was not only the biggest and the best funded—although it was a total skunkworks project and a big brand—but frankly, it was dramatically crazier than all of the other games that were getting made. There’s complicated reasons for that.

First, remember I was in the Turkey City workshop with all these people. I had just come out of the ferment of all of those MUDs, and so did most of the team. We had team members from all of those different kinds of MUDs: the social ones, the conjuring-swords-from-code ones, the LP MUDs, all of that. We had people from all of those. And we were steeped in the sci-fi and everything.

So when we made UO, we actually thought we were building the next generation of the internet. The average age on the team was 22. Damn. Any of you here 22? None of you. Any of you within five years of 22?

All right. Goes to show you can dream big as long as some other boss hands you a big check.

The crazy thing about this is all the stuff we ended up trying to do. We had the combat of a Diku that was in there. We had all of the social ideas from games like TinyMUD and the other MUSHes, multiple-user shared hallucination worlds, and social worlds, which is why you could sit on chairs. It was us and Chrono Trigger. Nobody else had done that.

In fact, I’ll run through a list in a minute here. It was a simulated world. We couldn’t render it, but the grass actually grew. You could harvest wood from trees. The trees stayed put, but you could reduce the wood on the trees. It opened up all the politics and crap that had been happening in LambdaMOO. All the ideas from DartMUD. All of this stuff was happening. And as a result, it ended up in The New Yorker. It ended up in Wired magazine.

I’m going to give you a list of features that were popularized into video games by Ultima Online. For any given one of these, there will be an antecedent. You’ll be able to go, “Oh…” I’ll give you the first one.

Mounts. Yes, it’s true, Daggerfall beat us to market in having a rideable horse you could see in game, fight from the back of, hop on and off of. But it only worked for your avatar. We allowed it to work for everybody, and you could see it on everybody. The first modern mounts in games, UO.

Visible equipment layers on your avatar in the game. UO.

Taming, training pets. In-world houses. Decorating a house. Crafting economy. The modern crafting system. Customizable visible gear. Abstract resource properties on objects, the same idea that underpinned Thief and The Sims.

Player-run shops. Being able to rate another player. It was one of the first three reputation systems on the internet. The other two were eBay and Slashdot.

Bounty systems. The second game with a flat-fee sub. The idea of buying a game time card that you could redeem for game time. A total dead end: true 3D terrain in isometric graphics, along with multiple-story buildings.

The modern guild system, which I wrote over a weekend. Player loyalty rewards for sticking with a game. Collectible rare items in a game. eBaying digital items. This was ground zero. The birth of gold farming happened here because eBay was there at the time. It was a site for selling Pez dispensers, and then a castle went up for sale for $1,000.

Concurrency in a game larger than 250. UO supported 2,800, and the awful acronym was coined to explain the difference between that and Meridian 59: massively multiplayer online role-playing game. MM O RPG. It got shortened almost immediately.

Live in-game storytelling. Player-owned ships that you could sail. The first worldwide deployment of a VPN so that we could run servers in Europe. Community managers as a job. Dynamically spawning points of interest in the world. In-game jails. In-game player-written books.

Finally, because of a discussion on the MUD-Dev mailing list, it’s my fault. I said, “Well, we really built UO more as a sandbox than as like a theme park full of rides.” And that was ground zero for the term.

However, the most lasting legacy of it, since it got steamrollered by EverQuest almost immediately, was that it got hacked a lot. Users started running their own worlds in what were called UO grey shards, and at one point Ultima Online was accidentally the most popular metaverse platform on Earth.

There were hundreds to a thousand illegal Ultima Online servers running. You could download a piece of software with a catalog of them and connect to any of them. Long before the game launched in Asia, there were half a million players playing in China by 1999 that we had no idea about.

I show up at an event in China and everybody knew who I was. I’m like, “Why do you know?”

“Oh, we all stole your stuff.”

Bear in mind, every object you see on this screen, including each of these wall tiles, is streaming down on the fly. It is dynamic coming down on the fly, and every single one of them is scripted.

You probably have to come forward past Battlefield, past Call of Duty. Probably not until GTA Online will you find another game with as many interactive objects in this one screen as there is in the entirety of GTA. Every single one of these things can do something. All of them are dynamically scriptable. That’s something we’ve actually lost.

In parallel, the games in Asia kept developing. UO shards were popular there. Eventually, we got this game called Lineage. Jake Song was playing UO. He worked on Kingdom of the Winds. We can blame both NCsoft and Nexon on Jake.

This kind of set the guild-versus-guild PvP MMO in motion. To this day, Asian MMOs and Western MMOs play really differently.

Then a game made here locally said, “Let’s take DikuMUD, which is the really well-known, really dominant model of online games, and just freaking clone it.”

Brad, Bill, those guys, they just wanted to make a DikuMUD with graphics. And they made a great one. They actually had to sign a piece of paper saying, “We swear we took no code,” because DikuMUD was an open-source project, and they even used some of the same text messages. Very much a Diku, and a proven game model.

Frankly, it was way more moment-to-moment fun than UO, with all of its dreams and ambitions, was. Pair that with 3D graphics and it just took off like crazy.

But now, time for a new column. You have been following this, right? New column.

What if you port an MMO into web pages? Several people started doing this. Weirdly, they mostly were inspired by UO, which is why so many of the web-based games. I’ve talked to the people who made Neopets, lovely folks. They said, “Well, we wanted to make UO on a web page, but we weren’t good enough coders, and my girlfriend drew all the creatures.” That was the birth of Neopets.

Inspired by sandbox play, these started reaching enormous audiences.

By 1999, we were starting to get asked by serious journalists, “Do you think we’re going to plug into the Matrix and never wake up?” People were starting to ask those questions. People are starting to worry about the impact of online.

In the wake of Ultima Online, EverQuest, and the third of the big three, Asheron’s Call, everybody started making MMOs, and they didn’t confine themselves to RPGs. We had tries at RTSes. We had people who said, “What if we make SimCity online?” long before they made SimCity online.

A lot of these genres didn’t stick. But one worked really well for web, which was the asynchronous strategy kind of genre. There were dozens of these that popped up all over the web. It took a little while for them to really click because it would take a new platform for them to really, really work. Nobody had invented the smartphone yet, and that is where that genre really flourished.

We started getting the notion of, “What about a secondary world that intersects with the real world?” This was the birth of the alternate reality game. First with The Beast, the marketing campaign for the movie A.I., and then Majestic.

People started saying, “What if we put sandboxes on the web?” RuneScape was born as a UO clone running in Shockwave. Huge audiences.

But people were saying, “What if sandboxes can’t be successful?” because there were a lot of high-profile disappointments. The Diku model clearly made more money and players had more fun. The Sims Online was on the cover of Time, but it couldn’t make back its budget.

There was a backlash against huge crowds. Neverwinter Nights instead allowed individual players to run microservers so they could each run their own little D&D games with just friends, which is basically the private server movement of today.

A couple of folks who were modding RTSes came up with this idea of modding an RTS into a MOBA, and we got a new genre.

It took a long time, but somebody finally did build a new MOO, called Second Life, that explicitly wanted to be a metaverse. It was also explicit, but the less said about that, the better. They actually did release open-source versions of their client and their server, and to this day you can hop around OpenSim worlds.

MMO FPSes: getting the latency down enough to allow people to shoot each other in the face was a huge challenge. That was really, really difficult. Also pioneered right here in town: PlanetSide at Sony Online.

There.com started doing essentially the same old kind of social world. All of these are struggling in the face of the straightforward appeal of shooting somebody in the face and leveling up.

Even this one, which is my fault—although not the engine part, that’s not my fault—we started stealing all the best social design bits whether or not they fit the Star Wars license. But the fact of the matter is that sandboxy games like this one, which also said, “We want to be UO in space,” really struggled during this time period. The straight-up things where you just killed monsters and leveled up were succeeding. Projects like Wurm Online, where you had to literally build the entire game yourself starting with dirt, were having trouble. They weren’t accessible enough. They were too hardcore because their features were fashion, farming, crafting, and item trading. Clearly way too hardcore of a feature.

Kids’ worlds, however, started booming. They were mostly combat games. In Toontown, you shot corporate cogs in the face to defend cartoons. It was a DikuMUD, right? But kids’ worlds got to be extremely popular. People weren’t paying much attention, but kids’ worlds were racking up tens of millions of users long before mainstream MMOs were.

Then came this thing called Facebook. All those social worlds just kind of went flat because it turned out the world was in the way of chatting. It was easier to just chat.

Along came the answer to the mass-market MMO. The regular gamer showed up. It turned out that the secret to being a mass-market MMO was: use a major game license, handhold players all the way through, and have a higher budget than every other online game preceding it combined. That’s how much they spent on getting it.

Made it much more of a game and less of a world, but it worked.

Pretty graphics finally started opening up this market. By 2004, we’re getting games like Travian. Web pages are pretty enough, responsive enough, that you can have the web-based PHP strategy game. This would find its ideal home on Facebook.

Console games started becoming MMOified. KartRider in Korea: because of long-standing history between Japan and Korea, you couldn’t really get a hold of game consoles. So Nexon started cloning famous—let’s face it, KartRider is Mario Kart. They started cloning console games, and then they had to go, “We can’t stick a sub fee on Mario Kart, so how do we make money, since people only play online PC games in our country?”

Their answer was to rip off the gold-farming idea and start selling the gold directly. This game is the birth of the microtransaction business. That is why it happened.

For a while, you couldn’t turn around without seeing kids’ worlds. They were absolutely everywhere. Even ones that people today forget were once upon a time kids’ worlds from that time period, such as this one, which was actually part of a different boom: the web-based virtual-world boom, all of which died in the 2008 crash except for Roblox, but which all aspired to be the metaverse.

Then something like Twitter comes along, and it’s like, why do you need a metaverse if you have Twitter?

So what we started seeing was the gradual dismantling of the old ideas. We get looter shooters, which are basically like lobby MMOs. We get kids’ games that are basically like lobby-based card-game MMOs. People take the World of Warcraft art style. It was actually an ex-UO developer who was on the original team, and he helped out some of the other guys, and they made this thing called League of Legends. It got really popular.

Social media let you have these asynchronous games. This is just a fancier Travian. And in fact, it was former MMO people who built FarmVille. In fact, former sandbox people, the people who had been in that web-based virtual-world metaverse building.

So social media really pushed asynchronousness.

From here on out, the changes get few. I’ve got three minutes, so I can show you just barely a couple.

Streaming. The main innovation here is that now devs make less money than the players. But literally, this is the first innovation in chat in 35 years, because Discord ain’t it.

This is the most important MMO of the last decade, because it didn’t just create survival games. It’s the closest thing to a whole bunch of people running MUDs. Three generations of virtual admins have grown up running it. And a guy from Wurm Online was one of the people that did it, came from Wurm Online.

The MMO thing there.

Of course, we can look at Hunger Games leading to battle royale. Clash of Clans, Minecraft starts turning into survival games. Here we’re in the realm that all of you remember, right? Even you who are within five years.

More survival, less surviving in the survival.

Roleplay didn’t die. Kids really want to roleplay, but if you ask the industry, roleplay is niche, even though kids will roleplay cops and teachers in these games all day.

There was a programmer who tried making an MMO twice. He made Game Neverending, and it didn’t work, but it had a cool photo-sharing feature, so he rebranded it Flickr and invented photo sharing. Then he made another one called Glitch that was really cool, and it wasn’t successful. So he pivoted that too, and he took out just their collaboration chat system, and we call it Slack.

He’s a billionaire now, but he’s a failed game designer. Sorry, Stewart.

Looter shooters turned out to be the thing that works for console. Discord is frankly just IRC again. We have too many ways to chat now, and they’re all the same.

The whole world is effectively going MMO. This bar at the bottom is spreading sideways because the games-as-a-service model swallows all the genres and drives the revenue of the top ten on all platforms.

Pokémon.

And then for one month, peace reigned on Earth.

It turns out that MMOs drive the majority of the hours of usage until Gorilla Tag on all VR headsets. And of course, there’s a case to be made that Pokémon Go is the first MMO augmented reality.

People finally cloned Hunger Games into a game. Lobby-based, session-based, not an alternate world, but we get battle royale, and that’s pretty much our last genre.

There have been other online games since, but has anything really changed? Has there been a true innovation?

The thing I really want people to have as a takeaway from this talk is that so many of the ideas are old, and so many of them that still have currency are still things that people dream of, are still the shining city on the hill. When we think of, wow, my game of the future, it looks like the ones back here.

There’s some sense in which you all might be 30 years old, or whatever, on Earth, but it still feels like it’s 30 years away, if that makes sense. Because the metaverse dreams, they just won’t die. Except that we might end up in a cyberpunky situation where we are completely beholden to some corporate master for it instead.

So here’s my takeaways for you as you think about game making.

A lot of the hot genres are ancient. That’s one of them.

Big hits are built on top of old hits.

Some things just don’t change. Chat’s the same. Doesn’t matter that you can embed more GIFs in Discord. Okay, whatever.

MMOs are swallowing everything because you can put every game genre in an MMO. You cannot do the reverse, because an MMO is not a game. An MMO is a virtual alternate world, and you can put as many things as you want in some alternate world. I’m sure somebody has, in some fevered fan fiction, stuck Fortnite in their Tolkien fanfic. It’s got to be, because everything’s pure.

Social might go flat, but these dreams just won’t go away.

It is amazing the degree to which people have the same misconceptions over and over again. They think sandboxy gameplay is inaccessible because there are too many things to do, even though farming, fashion, crafting, and decorating houses is, I don’t know, 80% of the entire mobile game market today. And why somebody would say that the activities in Minecraft are inaccessible.

Watch out whenever somebody says, “Well, but that’s not a real game.” They’re always wrong. Always. It got said about every single one of those.

And of course, MMOs are dead as a marketing term anyway, but history says they’re probably just getting started.

Thank you.

[Applause.]

HOST: Thank you so much, Raph. Wow. Well, we do have some games outside that we are demoing. But before we head out, I just want to say, we love having events like this. We love sharing this knowledge, and we’re here to support everybody, especially our San Diego game developers.

So please come to our monthly mixers that we have. If you guys are not on our Discord, please talk to me and I can get you on there, and we’ll be having more of these. If you have any suggestions on how we can improve and what you guys want, please let me know.

Oh, yes. We are also having a California-wide game jam called California Clash. It’ll be us against all the other IGDA chapters in California, and we’re going to win this one.

So yeah, we have a lot of games out there to play. Stars Reach is also on display. Please find your way outside. Thank you so much, guys.

[Applause.]