ROBLOX is a MUD: The history of virtual worlds, MUDs & MMORPGs

A lot has been written already about the history of virtual worlds, but most texts usually focus heavily on MMORPGs. One would need a very big book to tell this story in detail, so this article was done as a general timeline of what was going on, the key names, and where you can get more info.

Hope you find it an interesting read. 🙂

PART I — The first virtual worlds

Back in the 60s & early 70s, we didn’t have home computers. We just had a bunch of massive computers stored inside universities and large companies. Most were controlled by inserting perforated cards and had no monitors, just a printer showing the result of the commands you inputted.

Home computers wouldn’t have a mouse until the mid-80s!

Yet some of these computers were way ahead of their time. The legendary “Mother of All Demos” presentation, made by Douglas Engelbart in 1968, shows him using a mouse and window-based UI, clicking on hyperlinks, and chatting with a colleague via video conference while co-editing an online text.

PLATO (not the Greek guy)

One of such avant-garde computers was the PLATO (Programmed Logic for Automatic Teaching Operations). Created in 1960, it was a system with friendly terminals, designed to teach university students via virtual lessons.

The PLATO IV system, introduced in 1972, went as far as to offer vector-based graphics, a touch-screen interface, and connection to ARPANET, a pre-Internet network among US universities, research centers, and defense agencies.

University students using PLATO terminals in the mid-1970s

Students soon found that all this could be used to create games as well, and titles like Empire (1973) and Spasim (1974) began to appear. Empire is particularly impressive: it is a game where up to 30 players battle in a top-down space arena, shooting each other’s ships and fighting to control the galaxy — all this in 1973!

When Dungeons & Dragons came out in 1974, it unleashed the perfect storm: powerful computers, bored programming students, and a statistics-driven game that was begging for automation. The result was the birth of Computer Role-Playing Games, with titles like pedit5 (1975),  dnd (1975),  Moria (1975),  Oubliette (1977),  Future War (1977),  Avatar (1979), and Camelot (1981).

PLATO’s Moria, not to be confused with the popular 1988 roguelike of the same name.

All these are interesting in their own way, but some of them stand out for the purpose of this article: MoriaOubliette, Avatar, and Camelot all had multiplayer!

Of course, I just mentioned that Empire also had multiplayer, but these two had more than a bunch of people shooting themselves: they had a world to explore, with mazes, shops, secrets and monsters!

They had a virtual world!
A permanent one, that all adventurers would explore together!

The details are a bit foggy, as the games were updated over the years, but the basics are mostly the same: you create a character — sometimes choosing from multiple races, others with multiple classes — then you start in a peaceful town, buy equipment, enter a massive dungeon and must try to find a magical item, fighting monsters along the way in turn-based combat.

Multiple players (around 15) could explore this world dungeon at once, competing to see who becomes stronger, and they could also join forces in a party. Essentially, the party leader would control the movement of the entire group, but each player would control its character during combat.

While kinda simplistic and grindy (the mazes were gigantic and full of random battles), these games were extremely popular among PLATO users:

The CERL PLATO system logged 10 million hours of use between September, 1978 and May, 1985 (a period for which the most complete statistics are available). […] some numbers are known for games. Avatar alone accounted for about 600,000 hours, and Empire claimed another 300,000 or so. All told, games probably accounted for about 20% of PLATO usage during this period. — David R. Woolley

(TIP: Thanks to the effort of Cyber1, a community created to preserve the PLATO legacy, these games are still available and can be freely played.)

While a few PC games would come out from PLATO users trying to replicate that experience — such as the Wizardry series — they were mostly single-player games. Future online RPGs would have a different origin: MUD.

What’s a MUD?

MUDs began in 1978, with Roy Trubshaw and Richard Bartle, two students at Essex University in the UK. They played ADVENT (also known as Colossal Cave Adventure or Adventure), a game from around 1976 where you would explore a cave (composed of multiple “rooms”), solving puzzles and collecting treasures to gain points. It’s a legendary game, that created the entire adventure game genre and inspired countless developers. The game was fully text-based, with all interactions done by typing text commands like “OPEN DOOR” or “GET KEYS”:

You can play ADVENT online at

Inspired by it, they created their own version, but with a revolutionary twist: other players were also exploring this world and trying to progress. So, instead of being a normal dungeon, it was a Multi-User Dungeon: a MUD. (Their game would later be referred to as MUD1, British Legends or Essex MUD)

A key innovation was that MUD was open-ended in a way that ADVENT wasn’t (and adventure games today still aren’t). MUD was about freedom; ADVENT was about puzzle-solving. Although MUD did feature some ADVENT-style puzzles, they were less constrained and more non-linear. This wasn’t just for philosophical reasons, either: in practical terms, if one object (such as ADVENT’s lamp) was the key to advancing, then whoever got it first would lock up the game for everyone else. — Richard Bartle, MMOs from the Inside Out (2016)

The game is very open and mysterious, figuring out what you should do is one of the challenges and a reason to talk to other players, be it for trading tips or to help with specific obstacles, like lifting a heavy object. While the text commands were primitive, it was enough to have conversations with other players, and it even offered emotes like smiling or hugging.

Combat is simple, you just type KILL “CREATURE” and your character will fight automatically, but there are also spells and items to use in battle. Monsters can kill you permanently, forcing you to create a new character and start from zero. Gaining enough points makes your character level up. Reach the maximum level and you become a wizard — immortal and able to cast spells that affect yourself, other players or even the entire game server — effectively becoming the equivalent to a GM (Game Master) in a modern MMO. A wizard can freely teleport, see what other players are doing, freeze time for everyone except himself or even kill or imprison other players.

Some of MUD1’s spells, taken from Duncan Howard’s book ‘An Introduction to MUD’ (1985)

A landmark in video game history, MUD1 would define the genre. While the PLATO RPGs were all about combat and stats, here the exploration and social aspects were at the forefront. Combat was just a type of obstacle, or a drastic way to interact with other players, if you so desired.

Like the PLATO games, it was originally only accessible to students at the university. However, in 1983 Essex University allowed people with home computers to connect to their network at night, spreading the game to a much wider audience.

PART II — Turning MUD into money

Personal home computers began appearing in the mid-70s, with iconic machines like the Apple II, Commodore PET and TRS-80:

An Apple IIe, Commodore PET and a TRS-80, in an amazing photo from .

It’s important to remember what a personal computer meant back then. There were tons of different brands: IBM, Apple, Commodore, Sinclair, Amstrad, Texas Instruments, etc. Each incompatible with the other. Imagine Android vs. iPhone, but with 10+ competing standards.

Also, computers were still kinda useless outside of offices or colleges. They were mostly a very expensive novelty for tech enthusiasts.

They were slow and hard to use, most would just give you a black screen when powered on, requiring you to type relatively complex commands before they did anything.

Also, there wasn’t a “killer app”, a good reason for every house to have one. You could write text, make data sheets and play a few games, but is that enough? To justify such heavy investment they were sold as a machine for the entire family, showing how the son could do homework and the mother could read cooking recipes on them (lol).

In the 90s, the Internet would become this “killer app”, but there was no Internet yet. The US had closed networks like the ARPANET among a few universities and defense centers, but people at home were left out of the fun.

At home, all you could do is use your modem to dial to another computer via the phone — paying phone charges by the minute! And since dialing outside your area meant a more expensive fee, people were mostly connecting to local machines, where they would leave messages to each other — the BBS!

A BBS is like an archaic version of Reddit — you go to a BBS (“subreddit”), post a message and people will be able to reply below. Just look at these messages from a BBS in 1993 discussing the just-released DOOM:

PRESS SPACE TO OPEN THE DOOR! This is why video games have tutorials now.

Unlike Reddit, they could also host files. So you could do like the fellow above and download demos & shareware games like Doom from your local BBS, but that’s a story for another time.

(TIP: If you’re curious about BBSes, there’s an 8-episode series called BBS: The Documentary that you can watch on YouTube — Episode 5 is especially cool, showing the origin of ANSI & ASCII art all the way back to telegraphs!> Or you can visit BBSes that are still online.)

The point is, the cool kids at the universities could play PLATO games and MUD online, but all the other computer owners could just visit BBSes.
This would quickly change.

The local MUD servers

In 1978 Alan E. Klietz wrote Milieu, Colossal Cave clone with a heavier Dungeons & Dragons influence. Like MUD1, it also had multiplayer support — it’s not surprising that such concept was “invented” multiple times by multiple people across the globe— but was initially restricted to an educational mainframe in Minnesota, severely limiting its early influence.

However, in 1983 Klietz converted it to home computers, allowing for up to 16 players to play together connected via modem to a server. Renamed Sceptre of Goth, it became the first commercial MUD/online RPG, being sold as a franchising business in the US: companies could pay for the license and equipment to run local servers and charge players per hour:

INFO Magazine #10, May/June 1986

Sceptre of Goth focused heavily on delivering something akin to the tabletop D&D experience, with players rolling characters of multiple races and classes, forming a party and exploring a virtual world that was supervised by a server admin in the role of Dungeon Master, which would edit the world in real-time and even role-play as a monster.

Sadly, the game cannot be played today, but you can read about it HERE.

The rise of Online Service Providers

If small local companies charging users for online games, you can bet that big ones were also going to do the same.

Around 1979, the first Online Service Providers began to appear in the US, such as CompuServe, The Source, GEnie and Prodigy. Now, do not mistake these for Internet Service Providers — there was no Internet yet! Instead, what they offered was a bundle of online services, such as email, online newspapers, chat rooms and games. All paid, of course.

A really weird CompuServe ad about its chat rooms, and how you could meet the love of your life there.

In the battle to differentiate themselves, games played a vital role. Like console companies searching for exclusive games, so were the online service providers trying to find an addicting game — a reason for people to contract their services.

Some famous online games from this era include Kesmai’s 1988 Air Warrior, a 3D flight simulator where you could battle players online, and 1989’s MadMaze, a very popular single-player adventure game set inside a giant maze.

Another of those games was 1984’s Island of Kesmai:

While the PLATO games were still hack & slash RPGs with very simple multiplayer, and MUD1 was more of an adventure game with social elements, Island of Kesmai was already the blueprint of an MMORPG.

Sure, it had no graphics, just a tiny ASCII map, but that allowed it to offer a depth that it would take years for other games to rival.

An excerpt from 1989’s ‘A Guide to the Advanced Game v2.0’, a fan-made guide for Islands of Kesmai. Note how advanced the world already was.

Up to 100 players at once would explore a massive open world, killing monsters, doing quests and hunting “world bosses” that required a powerful party to take down. It also had an alignment system, were PvP and theft was punished by making the killer “chaotic”, marking them as a free target for PvP and having NPC guards attack them — a system still being used 36 years later in MMOs like Ashes of Creation.

Active from 1984 to 1995, IoK received multiple expansions over time, adding new areas and challenges. A key feature was finding a portal to take your character from the Basic Game (BG) areas to the Advanced Game (AG) ones. Doing so would allow you to grow more powerful, but was a one-way journey to a more dangerous land. And death could be permanent, either by old age (after a certain number of turns), or by being devoured by monsters.

Playing Island of Kesmai with Hunter, a DOS front-end from 1994. It supports player replays, if you want to watch how people played.

I’m going into a lot of detail here because Kesmai is very well-documented (you can even watch replays of old players) and I want to show just how complex online RPGs already were in 1985—mechanically, some of them were much more complex than single-player RPGs of the time, like The Bard’s TaleUltima IV or Dragon Quest.

The legendary Habitat

Another online game made for Online Service Providers, Habitat stands out not only for having fancy graphics, a big development studio behind it (LucasArts) and being extremely influential, but also for signaling the start of a great divide in MUDs, online RPGs and virtual worlds in general — the focus on social elements above all else.

Habitat is clearly based on MUD: a world composed of rooms that multiple users explore, interact and socialize by typing text commands. However, Habitat had no goals. No dragon to kill, not even a score system to track all your treasure. It was just a big world to explore with other people. Sometimes it had activities set by the devs, but mostly it was a sandbox.

The idea behind our world was precisely that it did not come with a fixed set of objectives for its inhabitants, but rather provided a broad palette of possible activities from which the players could choose, driven by their own internal inclinations. It was our intent to provide a variety of possible experiences, ranging from events with established rules and goals (a treasure hunt, for example) to activities propelled by the players’ personal motivations (starting a business, running the newspaper) to completely free-form, purely existential activities (hanging out with friends and conversing). — Chip Morningstar and F. Randall Farmer

There are many amazing stories here, like the fact that Habitat had guns and the devs let the community decide how to deal with people who killed others to steal their items. I really recommend reading more about it HERE.

Also, you can play a restored version of Habitat HERE, directly in your browser, with a tutorial and fun trivia about the game!

The second wave of online RPGs — now with graphics!

With home computer technology advancing and new Online Service Providers beginning to appear, new online RPGs followed, with a famous trio of graphical MUDs / online RPGs appearing in 1989–1991:

Kingdom of Drakkar (1989), Neverwinter Nights (1991) and The Shadow of Yserbius (1991)
  • Kingdom of Drakkar was based on an early local MUD from Kentucky, got a graphical front-end in 1989 and was then launched with top-down graphics to a wider audience as part of an online service provider called MPG-Net (Multiplayer Games Network) in 1992.
  • Neverwinter Nights was based on SSI’s extremely popular “Gold Box” series of official Dungeons & Dragons single-player CRPGs. Hosted by AOL with support for up to 200 players at once (later 500!), it was the biggest of the early MMOs, surviving until 1997 when an IP dispute closed it down.
  • The Shadow of Yserbius was a first-person dungeon crawler, where up to four players would form a party and explore a massive dungeon, fighting in turn-based combat and trying to solve puzzles. It later got two sequels, The Fates of Twinion (1993) and The Ruins of Cawdor (1995).

All these games would run for several years and form a die-hard fan-based, who spent a small fortune playing. Not only personal computers were still very expensive, but access to these games was still paid by hourly rates!

For reference, in 1991 CompuServe would charge $5 USD for each hour using their services, while AOL would charge $2.95 USD/hour. Some games would charge an additional $3 USD/hour — adjusting for 2020’s values, playing online in 1991 could cost up to $12 USD/hour! Plus the phone bill!

As industry analyst Jessica Mulligan explains:

“On GEnie during 1991, our average MMOG customer spent $156 per month, the equivalent of 32 hours at $3 per hour to play. However, the hard core players averaged three times that and accounted for nearly 70% of the total revenue. The top 0.5% had truly astronomical bills, well over $1,000 per month”

So it was still a tiny market sustained by a few hardcore players with a lot of disposable income — or “whales”, as modern marketing calls them.

PART III — The great MUD explosion

MUD is a primitive, outdated and kinda misleading term.

“Multi-User Dungeon” has a very Dungeons & Dragons vibe, but MUD1 wasn’t set inside a dungeon, and it was barely an RPG. Habitat is often called a graphical MUD, but had no monsters, levels or even goals.

Of course, players themselves are a diverse bunch and want different things in a game/virtual world. Richard Bartle has a very famous theory on this, splitting players into four distinct groups:

Taken from ‘Designing Virtual Worlds’ by Richard Bartle (2003)

(TIP: Try this fun test to find out which profile you fit. And notice how many of the questions don’t make any sense in modern MMOs anymore.)

Usually, a game designer wants all of these players to be happy, to make for a rich and dynamic virtual world. But what if they just want the “Socializers” to talk in peace? Or just want to keep making challenges for the “Achievers”?

So far we had only a handful of virtual worlds, as they were hard to create, and only Habitat focused heavily on social aspects. But in 1989, a group of students in the UK developed AberMUD, heavily inspired by the original MUD — and then released the game’s code for free across the world.

This is a landmark in online history. Hundreds of MUDs appeared across the world, with people modifying the code to change the setting, add new features, etc. These would lead to two big evolutionary paths:

PATH A — TinyMUD and customizable virtual worlds

TinyMUD was created to be a more social MUD, with a smaller world to explore but more social interactions, like the ability to privately talk to another character (before, anyone else in the same room could hear you).

But its key feature is one we see to this day: players could edit the game world, making their own rooms and objects for others to see and interact.

This was revolutionary. A virtual place that anyone can customize, build and modify, then show to others & visit others— this is one of the pillars of the Internet, of social media, of Habbo Hotel, Second Life, Minecraft, etc.

Initially, all you could do was basically create a room, write a description for it, place items inside and write a description for them. For example:

@dig <BATCAVE> 
@create <BATMOBILE>
@describe <BATMOBILE> [=<You see a super cool car>]
@create <ROBIN>
@describe <ROBIN> [=<You see an annoying kid with a yellow cape>]

So yeah, now I have a Batcave, and you can come to check all the cool stuff I have inside it. I’m oversimplifying, but you can see the potential.

Over time, new code variants appeared, allowing for players to create more complex content without needing to know how to program, such as MUCK, MUSH and MOO. We’ll talk about them below.

PATH B — DikuMUD and the blueprint of combat-oriented MMOs

If TinyMUD was for people who wanted to socialize and create worlds, DikuMUD was for people who wanted to kill monsters or other players.

The base game offered four classes — mage, warrior, thief and cleric —each with set combat roles (the omnipresent Tank/DPS/Healer trifecta). Killing monsters would earn you loot and XP, and you would return to town to level up and earn new abilities. These abilities would have cooldowns, and allow characters to draw aggro, crowd-control or even summon pets.

Yes, I am describing most post-EverQuest / World of Warcraft MMOs. That’s how influential DikuMUD is. That’s how deep of a divide would come out of this — now each player type had its own virtual worlds to play:

Social worlds went one way; game worlds went another. This was the Great
Schism that rent the concept of virtual worlds in two, and — sadly in my view — persists to this day. — Richard Bartle, MMOs from the Inside Out (2016)

A ton of MUDs and confusing MUD acronyms

Now the stage was set for hundreds, if not thousands of MUDs to appear across the world, sporting all kinds of rules, settings and code bases. To help MUD fans differentiate them, they are commonly split into a few categories:

Also taken from ‘Designing Virtual Worlds’ by Richard Bartle (2003)

Yes, there were furry MUDs. And erotic MUDs. And anime MUDs. And PvP-only MUDs. Educational MUDs. MUDs for Star Wars, Star Trek, Lord of the Rings, Discworld, DUNE, etc… Hell, Aerosmith did a “Cyberspace Tour” in 1994, talking to fans inside MOOs! Here’s the archived chat of that day.

Some MUDs were paid, but the majority was free, made by students leeching very expensive university infra-structure. There was no Google, so people would post lists of MUDs on BBSes, newspapers and magazines:

From ‘Net games: your guide to the games people play on the electronic highway’ (1994)

While I’m trying (and failing horribly) to be succinct, it’s important to point out a key event here: in 1993 a user hacked a MUD named LambdaMOO and forced another avatar to have virtual sex with his avatar.

The incident, reported on the A Rape in Cyberspace article by Julian Dibbell, sparked many debates about cyberspace and laws in virtual worlds. You can read more stories from inside the MUD world in Dibbell’s book My Tiny Life: Crime and Passion in a Virtual World. It was also around this time that Neal Stephenson wrote Snow Crash (1992), an iconic book about cyberspace.

MUDs in 2020

From here, MUDs take a backseat in our story, but that doesn’t mean they’re dead — far from it, The MUD Connector lists 627 currently active MUDs.

These games are, of course, still text-based, but have received tools like MUSH Client and Mudlet that provide maps, stats and a customizable interface. If you want to see how Aardwolf, a popular modern MUD works, check this:


Circa 1991, a kinda important thing happened: The Internet.

Remember, so far we basically only had private networks like ARPANET and people at their houses using modems to directly call other PCs, BBSes or an Online Service Provider, where they could read emails, chat & play games.

The Internet means you use your modem to call an Internet Service Provider (ISPs) such as AOL, and that service connects you to this big web of servers — the World Wide Web, full of web pages for you to access.

Of course, back then you still had to pay the per-minute modem call to the Internet Provider. And now the Internet Provider too. But at least accessing a server in New York, São Paulo or Tokyo would cost the same.

For someone like me, living in Brazil, directly connecting to a MUD in the US would cost a fortune and be as fast as sending my data by letter. The Internet allowed me to connect to my local Internet provider and then access content hosted anywhere in the world, or even play World of Warcraft on the US server without any extra cost — just an adorable 3500ms latency.

PART IV — Where we make readers happy by talking about that one MMORPG they played

First, what’s an MMORPG? It means “Massive Multiplayer Online Role-Playing Game”, and was supposedly coined by Richard Garriott to show how his online RPG was cooler than the other online RPGs.

There’s no commonly agreed number that makes an online RPG “massive”, so it’s really hard to point out THE FIRST. Island of Kesmai could support at least 100 players at once, is that massive? Meridian 59 could support “a few hundreds”, is that massive? Ultima Online’s few thousand? And are any of those really massive when compared to World of Warcraft?

It’s not important. It’s a buzzword, coined for marketing and heavily dated. Take just one thing: servers now supported more people playing together.

By the mid-90s, we get a third wave of graphical MUDs /online RPGs / MMORPGs. Like the generation before, they all had very unique styles:

The Realm Online, Meridian 59, Legends of Kesmai and Dark Sun Online — All very distinct graphical MUDs / MMORPGs released in 1996.
  • The Realm Online followed the style of Habitat, but with a bigger focus on the RPG aspect, including turn-based combat.
  • Meridian 59 presented its fantasy world in a Doom-like fake 3D first-person view. This was considered a massive leap in quality and immersion, part of why some call it “the first MMORPG”, instead of a “graphical MUD”.
  • Legends of Kesmai was a successor to Island of Kesmai, now with graphics!
  • Dark Sun Online was based on the Dark Sun: Shattered Lands (1993) single-player RPG by SSI. Very obscure, it had several development issues and was quickly killed.

I don’t have space to go into detail here, but The Game Archaeologist has great articles on all those games — I’ve linked them above. It’s a fascinating era, with developers still trying to figure out how to graphically present a virtual world with hundreds of players.

We also start to see some action outside of the US, with German studio CipSoft releasing Tibia (1997), which is still going strong in Latin America.

I really recommend you Google translate this article about South American mafias operating inside Tibia:

Things were also heating up in Asia, but let’s first travel to Britannia, the land of Ultima Online.

Ultima & the bread bakers

One of the first versions of Ultima Online

Ultima was one of the biggest RPG series of the 80s and 90s, always very bold and somewhat experimental: Ultima 4-6 deal heavily with morality in video games, while Ultima 7 tried a simulationist approach to RPG worlds, one where you could get flour, add water, roll the dough with a rolling pin then bake it to make bread. It tried to be a living world, not just a fantasy land where you kill orcs.

Ultima Online (1997) embraced that, creating a simulated sandbox where people could do traditional RPG adventures, PvP, craft items and/or role-play. All this with thousands of players. At this scale, you can create something that cannot be done by a few dozen players: a virtual society, with enough people to birth a dynamic economy — professions, worker guilds, social classes, etc.

Garriot himself perfectly explains the charm of Ultima Online here:

Taken from Next Generation magazine #48 — April 1998

It’s a fascinating game, and you can read that story about how they killed Lord British and many other fun trivia about Ultima Online HERE.

Meanwhile, there was another radical industry change that greatly helped Ultima Online and other MMORPGs to take off: monthly fees.

UNLIMITED INTERNET! The end of hourly fees

Until now, everyone in the US who went online was paying an hourly fee. That changed in December 1996, when America Online (AOL), one of US’ biggest Internet providers, changed their service to charge flat monthly fees:

Since customers were given the option of buying unlimited network access for $19.95 a month, AOL users have spent an average of 32 minutes online each day, more than twice as long as the daily average in September, said Tricia Primrose, a company spokeswoman.

This created a boom in Internet use, bought many new users and forced other providers to follow with monthly fees. Moreover, the games themselves had to follow as well and start charging monthly fees.

As Jessica Mulligan explains again:

“Until Meridian 59 launched in 1996 and Ultima Online launched in September of 1997 with flat monthly rates, billing for commercial MMOGs was mainly on a per minute/hourly basis (with a brief period of free access to AOL’s games from 12/96 to about 7/97). Thus, the number of total subscribers was less important than how long you kept your hard core players (the top 10%) in game.”

So it wasn’t just that online RPGs could now support thousands of players at once — they NEEDED thousands of players, as the most hardcore player was now paying the same as the most casual one.

Of course, not every game was ready for this. Not even AOL was ready for this, and lag, server & connection issues were very common at the time.

The genre standards are set

While Ultima Online was having fun experimenting with what an online world could behave like, with players making houses and role-playing as peasants, EverQuest (1999) would return to that old MUD divide and take the other route — it was game made by DikuMUD fans.

The first version of EverQuest, with its blocky UI.

Like DikuMUD before it, it still had some social aspects, but they were all in service of combat and player progression. Just like DikuMUD would champion a wave of MUDs focused on combat above everything else, EverQuest would do the same for its genre, laying down the formula that most western MMORPGs would follow — including World of Warcraft.

From here starts a long series of MMORPGs: Asheron’s Call (1999), Dark Age of Camelot (2001), Anarchy Online (2001), Runescape (2001), EverQuest II (2004), City of Heroes (2004), Guild Wars (2005), Dungeons & Dragons Online (2006), Lord of the Rings Online (2007), etc… It’s not that they are bad games and I’m personally attacking you for not telling you how amazing the one you played is, but from here onward they become a well-defined genre, with most games following a very similar set of standards.

So, once again, I’ve linked to detailed articles by The Game Archaeologist.

Sadly, part of those standards meant abandoning the whole “baking bread” thing, following a more guided style of play where all players had to go out and kill monsters to progress. This is exemplified by a famous article of an MMO player on how he wanted to “bake bread”.

You see, it’s not just bread I want to bake. I also want to bake armor and dresses, phasers and chairs, scrolls and steam engines, swords and plow-shears, blasters and potion. I want to bake bread and anything else that you think is relevant for me to bake in your game.

This is not to say that MMORPGs abandoned crafting, just that they abandoned the idea of a fully player-driven economy. Yes, WoW has people playing “Auction House: The Game”, but the vast majority of items and services you need are sold by NPCs or readily given by dungeons & raids.
If you want a cool mount you don’t pay a fortune to a guild of players specialized in mount training, you just kill a rare monster or do daily quests. Consider this: are you earning/spending most of your money with other players or with NPCs and automatic game systems?

(TIP: If you have an interest in economy & MMORPGs, you will love Virtual Economies: Design and Analysis, by Vili Lehdonvirta and Edward Castronov)

Most MMORPGs abandoned that concept, but not all. The “bread bakers” still had a few great titles that remain beloved landmarks of the genre:

Star Wars: Galaxies (2003) had 34 professions for players to pursue, most with no combat role whatsoever. Instead, they were a vital part of how the world worked. For example, to change your character’s hairstyle you had to hire a player with the Image Designer profession. The higher their skill, the better options they could offer. And they would set their price. Or rather the free market would. Or rather, the guilds & mafias would…

Players had formed governments. Vehicles were very popular. The early game economy, which was intentionally rocky because players had not yet developed all the interdependence infrastructure, had started to hum along. Entertainers were going on tour, and few of them were macroing, because they played entertainers because they liked it. People were building supply chain empires and businesses with hundreds of employees. Merchants were making a name for their shops full of custom-crafted gear. — Raph Koster

As such, you had players that were max level, extremely active, rich and important without ever having to kill a single creature. Alas, it was but a short-lived dream. You can read about its downfall HERE.

EVE Online (2003) is one of the most fascinating MMORPG ever made. Take everything I wrote about SW:G and take it to 11. The entire game is driven by the player economy, factions and political schemes.

Want a weapon? It needs to be crafted, with materials gathered from a miner, that needs to be transported across space, where pirates await, so better hire guards… all these are people (or bots controlled by people), with giant organizations behind them controlling the prices and fighting for dominance. Mandalore has a nice review of the whole thing HERE.

Finally, A Tale in the Desert (2003) deserves a mention. Completely devoid of combat, it’s a virtual world based on Ancient Egypt, all about role-playing, crafting, socializing and fighting for social status & power, with players in positions of power being able to pass laws and even ban other players.

Split into seasons, the first one began in 2003 and ran for 18 months. After that, the server was erased and a new season began, with new rules. While it always had a small populace, it’s still online, currently in its 9th season:

World of Warcraft

And we finally get to the big guy. Released in 2004, World of Warcraft was massive. To have an idea of its size, here’s a graph of monthly subscriptions for western MMORPGs from 1997 to 2008:

Data collected from reports and press releases by Bruce Sterling Woodcock at

You could combine every other western MMORPG together and you wouldn’t even reach 1/3 of WoW’s numbers. The graph stops in 2008, but WoW’s peak would be only in 2010, at 12 million users. All paying monthly.

Why was WoW so big? There’s many factors to be discussed, but we must consider how polished and friendly-looking WoW was, even at launch. More than just cute visuals and an accessible UI, it streamlined the MMORPG experience, giving players a clear sequence of quests to complete & level up, instead of just walking around killing monsters.

As Raph Koster puts it in this great GDC presentation (seriously, watch it):

We built bugs by the tens of thousands. We tried to simulate ecologies and created environmental disasters. We built economies reading text books to master an impossible task, because our worlds were vast experiments. Dark Age of Camelot, EVE Online, Star Wars: Galaxies, Project Entropia… many worlds were rough, implausible, buggy and unfriendly.

And in 2004, the game industry finally kicked us in the ass and brought us back to reality when Blizzard launched World of Warcraft. Polished, expert, slick, fun, it was the hack & slash game that nailed for formula down… and collapsed the possibility space. It was, in some ways, the end of history.

World of Warcraft also rode on a big IP, from a well-established developer with a lot of goodwill — and a massive marketing budget. It became a cultural event, talked about in newspapers and magazines outside the game world, a part of popular culture, with even South Park episode about it.

Is all that enough to make it the biggest MMORPG of the world? No, because it isn’t the biggest MMORPG of the world.

WoW came out in November 2004. In August 2005, Korean studio Neople released Dungeon Fighter Online, a 2D side-scrolling MMORPG:

You might never heard about it before, but it’s one of the highest grossing video games of all times, with estimated 13 billion dollars in revenue (WoW is estimated around 10 billion). It also had more than 600 MILLION registered users. The game is free-to-play, but that doesn’t change that it’s played by more people than WoW— while still making more money than WoW!

And that’s why we have to talk about Asia.

ADDENDUM — Looting, Shooting & Surviving

Raph Koster was nice enough to read this wall of text and point out that I skipped looters-shooters. Good point, so here’s an addendum.

All the talk we have had so far was about persistent virtual worlds. Ever since the aforementioned PLATO’s Empire (1973) there have been games where a virtual world is created for a multiplayer session, then resets when the game ends.

So many bombs, so little result…

That’s what happens in the maze shoot-outs of Maze Wars (1974), the airplane battles of Air Warrior (1988), in Doom (1993) deathmatches, in StarCraft (1998), Battlefield 1942 (2002), League of Legends (2009)etc. Nothing you do inside those worlds will last or be carried out. You could say that Counter-Strike’s map de_dust2 is closer to a ping-pong table than to WoW’s virtual world of Azeroth.

But then some games began allowing you to keep your character and what they earned in that temporary world. In Diablo (1997), you can create an online world, bring your character, play with other people and then keep all the XP and loot you got, even as that specific online world vanishes.

It’s not a new idea, Eamon (1980) was a single-player text-based RPG where you could explore worlds created by other players (similar to Doom WADs), keeping the XP and items earned in them. Famously, you could play in a Star Wars world, get a “light sabre” and then bring it along into any other world!

I’ve been conditioned to get excited at purple item names

Diablo II (2000) is the game that elevated loot to an art form, an addictive pursue of rare and powerful equipment, with iconic color-coded tiers. Some of its creators tried to bring that concept to a multiplayer FPS with Hellgate: London (2007), but it was Borderlands (2009) that nailed the formula of FPS combat + RPG progression + tons of loot, creating the “looter-shooter” genre.

While Borderlands offered multiplayer, it was just a private game of up to 4 players. Diablo III (2012) forced all players to be online, but you were still playing with only four people at best, without any sense of sharing a world. WarFrame (2013) was, at first, similar to that. Composed only of mission maps where you would play with three other people, get loot and then go to the next mission. Overtime they added more permanent elements, like a hub where players could meet or large open-world areas to explore.

But before any of that came Destiny (2014), setting the standards in looter-shooters for what they called a “shared-world”:

Destiny’s Tower, where the cool kids gather.

Destiny is a persistent world complete with a dynamic day/night cycle and weather effects. Your character has persistent progression across single-player, co-op and competitive multiplayer as it levels up and gains new equipment and weapons. But the most important feature to note is that players, and Destiny, are always connected to the internet. — Eric Hirshberg

It’s like an entirely different family of games, that only recently found a way to successfully incorporate a permanent virtual world — and they did it rejecting the focus on massive battles with hundreds of players that “MMOFPS” like World War II Online (2001) and Planetside (2003) pursued.

I would love to see more debates about this new wave of smaller virtual worlds. For example, survival games like DayZ and Rust have worlds that can last from a few hours to months before restarting, begging the question of how permanent must a permanent world be. History is still ongoing, there’s a lot of new concepts to explore. 🙂

Now, onward to Asia!

PART V — MMORPGs in Asia

Video game history in general is extremely US-centric, a bias that can lead to some pretty terrible oversights. But let’s be clear: any talk about MMOs without Asia is either poorly researched or trying to mislead you.

First, remember one important thing about the pre-Internet world: it was already more expensive for a person in the US to connect to another state, so you can imagine few people were doing this across different countries.

So Koreans weren’t playing American MUDs (If you want more backstory on the Korean scene back then, check this article), but they had some people creating local ones based on DikuMUD. The most famous was a MUD called Jurassic Park (yes, with dinosaurs!), which went live in 1994.

A screenshot of Jurassic Park / 쥬라기 공원

By 1996 there were over 100 MUDs available to Korean players, with an estimated 200,000 people playing them regularly. Jurassic Park was the most successful, quickly generating over 20,000 visits per day and earning $20,000 in the month of July, 1994. Its growth would eventually reach over $200,000 per month in revenue for Samjung Data Systems. — Jong H Wi (Innovation and Strategy of Online Games, 2009)

Behind Jurassic Park was Jake Song (or Song Jae-kyung [WTF they deleted his English wiki?]), the Korean father of MMOs. Inspired by the success of the game, he partnered with his college friend Kim Jung-ju to create Nexon and release their first commercial game, Nexus: Kingdom of the Winds (1996), set in the world of Kingdom of Winds (바람의 나라), a popular manhwa.

Nexus: Kingdom of the Winds proved itself popular, but Jake Song would leave the company and join NCsoft, where he would work on a landmark in MMO history: Lineage (1998).

Lineage focused on PvP, with clans fighting each other being a core element of the game. That image on the right is a PvP raid for control of the game’s castle.

Sure, Meridian 59, Ultima Online, EverQuest, Runescape… all these games are important. But Lineage was MUCH bigger. How much bigger?

Data collected from reports and press releases by Bruce Sterling Woodcock at

From 1998 to 2004, it was bigger than every single western MMO combined.

  • Ultima Online peaked at 240,000 monthly subscribers.
  • EverQuest at 460,000 monthly subscribers.
  • Lineage at 3,250,000 monthly users.

It was WoW before WoW. But why are we comparing subscribers to users? Because Lineage wasn’t “just” 13 times bigger than Ultima Online, it also revolutionized the Korean MMORPG industry. Korea is a country known for its Internet Cafés and avid StarCraft players — both of which were booming in 1998, when Lineage came out.

As such, they offered a novel business option: Internet Cafés would pay monthly license fees so that its clients could play Lineage for free:

In 2000 Internet café sales accounted for over 70% of NCsoft’s annual revenue, more than 3 times the revenue that came from individual users on home computers. — Jong H Wi (Innovation and Strategy of Online Games, 2009)

Since most players weren’t directly paying to play, Lineage later introduced an item shop, something we’ll talk more about in a minute.

With this in mind, it’s healthy to contest some of the “truths” of MMORPGs. We already saw how Lineage crushed Ultima Online and EverQuest’s numbers, but other Korean MMORPGs like Ragnarök Online, Mu Online, Maplestory and the previously mentioned Dungeon Fighter Online also had huge numbers.

MU Online (2001), Ragnarök Online (2002) and MapleStory (2003) — Note that they never embraced the style of 3D third-person EverQuest/WoW clones.

They are harder to track but, in 2005, when WoW had almost 6 million monthly users, MapleStory was already at 13 million. It would peak at 18 million users in 2008, dwarfing WoW’s 12 million.

Of course, MapleStory is free, but those are still real people playing the games, and the companies are still making a profit over them. A LOT of profit. You can read an examination by DiGRA of the free-to-play + item shop revenue model in Korea circa 2007 HERE.

Moreover, the free-to-play model allowed them to bring MMORPGs to a broader audience in regions like Latin America, where few people could pay WoW’s monthly USD fees.

Japan & The Console Modems

While Korea was booming due to a very PC-centric game culture, Japan suffered from the opposite: not only Internet was still slow and charged by the minute, but computers weren’t popular as home devices. The solution was to use consoles to play online.

A Famicom, the Japanese NES, with the Famicom modem attached.

Console companies were already experimenting with this since the late 80s, with accessories like Nintendo’s Family Computer Network System for the Famicom and the Satellaview for Super Famicom, or SEGA’s Meganet for the Mega Drive. These were mostly used to download games and were commercial failures that didn’t last long.

By 1997 SEGA would release Dragon’s Dream for the Saturn — a Japan-only RPG where players would explore an online fantasy world, gathering at a town hub and exploring first-person dungeons with up to 4 players at once. While the game disc itself was free, players needed to buy the Saturn modem, pay for a registration fee and then pay an hourly fee to SEGA — plus the modem connection phone fees.

Released in 1999, the Dreamcast already had a built-in modem, so players didn’t need to buy any additional accessories. This greatly helped the release of a landmark in the genre: 2000’s Phantasy Star Online.

Heavily inspired by Diablo, it’s an Action-RPG that allows you to play alone offline or take your character online, meeting in a town hub then exploring an instanced world with up to 4 players — closer to games like Monster Hunter and WarFrame than to massive online worlds like World of Warcraft.

While it still had to deal with the high cost of Internet use and hardware limitations like the lack of a hard disk— forcing players to buy a Phantasy Star Online v2 disc to update the game — it was a commercial and critical success, proving the viability of a console-based MMO.


Sadly, much about China is still clouded by a huge language barrier, propaganda and several political & economical reasons why their internal market is usually ignored by journalists and even academia.

For example, CrossFire (2007) is reportedly the world’s most popular video game, with over 200 million monthly users, but you won’t find a single article about it in mainstream video game websites like Kotaku or IGN.

A very praise-worthy exception is this excellent PC Gamer article about PC gaming in China. I admit to knowing little about MMOs in China, but I would like to complement the article by showing this fascinating list from 2007:

2007 analysis from China Analyst:

Not only the list is dominated by MMORPGs you probably never hear about — with insane numbers like 1,5 million peak concurrent users — but even in 2007 the micro-transaction revenue model was already a force in China too.

Fantasy Westward Journey

This was years before Farmville (2009), when MySpace was still the biggest social media website. This was before Team Fortress 2 hats or CS:GO skins, before EA added micro-transactions to FIFA 09 with the Ultimate Team mode. This was going on when Bethesda was being mocked for selling horse armor as DLC in Oblivion (which still sold well).

In 2007, Fantasy Westward Journey (which is a really bad English title for a game based on the Journey to the West novel), was already at 25 million users around WoW’s release, and reportedly has over 400 million users by now.

But it’s the second game in the list, Zhengtu, that stands out even more: fully exploiting the aggressively competitive Chinese culture and a rising middle-class with more money than free time, it removed all classic MMO time-sinks. You can instantly teleport anywhere, talk to any NPC and get any quest. There’s also no need to spend hours grinding for rare loot drops, as this wonderful article about the life of a powerful Chinese player explains:

Good equipment means money. Unlike other games, in this game there are no items dropped when killing monsters or completing missions. “We all want the best,” said Lu Yang. “You have to go to the system’s shops to buy materials, and then use the system smith to make them. Or, you could go gambling.” “Gambling” means “opening the treasure chest.”

By treasure chest, they mean what we now call loot boxes. Yes, in 2007. Keep this in mind, there are companies sending press releases about being the biggest in an arbitrary thing while excitedly taking hints from bigger games. Don’t believe the hype.

PART V —MMORPGs aren’t the biggest virtual worlds

Since we’re talking about hype, it’s easy to be dragged into it and treat MMORPGs as this behemoth market that dominated everything, but that’s a somewhat narrow — and inaccurate — view.

Habitat would leave a lasting legacy, the concept of a virtual world where you created your avatar and explored while chatting with strangers would be replicated multiple times outside the MMO wars, without any RPG mechanic.

WorldsAway (1995), Worlds Chat (1995), OnLive! Traveler (1997), The Palace (1995) and Microsoft V-Chat (1995)

Dozens of virtual worlds focused on social aspects would follow. You had Habitat clones like WorldsAway, but they eventually began to remove actions outside of chatting — becoming worlds made of interconnected chat rooms, exploring concepts like 3D worlds (Worlds Chat), voice chat (OnLive! Traveler), virtual meetings, etc.

In the mid-1990s it was not at all clear what to call these new spaces. Users of earlier text-based worlds referred to them as graphical MUDs, those familiar with immersive virtual reality used the term VR with a number of prefixes (online VR, net VR, web VR), whilst academics were coining their own less than memorable terms such as collaborative virtual environments (CVEs), multi-user virtual environments (MUVEs), and members of the press and authors of key nonfiction books at the time had put forth terms such as the metaverse (Stephenson 1992) or mirror worlds. — Bruce Damer, Meeting in the Ether: A Brief History of Virtual Worlds as a Medium for User-Created Events (2008)

Might sound silly today, but this was hyped as THE FUTURE, with big brands and artist joining in (like it would later happen with Second Life):

A 1999 news segment on Worlds Chat, showing even David Bowie buying into the hype.

Part of this virtual world explosion and hype collapsed when the dot-com bubble burst in 2000. After this, a new wave of online virtual worlds would appear, with more humble beginnings and appealing to a younger audience: Habbo Hotel (2000), Gaia Online (2003) and Club Penguin (2005).

Habbo Hotel (2000), Gaia Online (2003) and Club Penguin (2005).

Gaia Online is particularly interesting for how it and successfully navigated online trends: it began as a list of links (those were important, we didn’t have Google!) and a small forum for anime & manga fans, then added customizable avatars and mini-games — like NeoPets (1999) and later Stardoll (2004). This allowed them to give exclusive items in exchange for donations, then in 2007 start directly selling items for real money payments.

By 2007 they also began doing sponsored activities, showing movie trailers in their in-game cinema. At this time, while WoW was at around 9 million subscribers, Gaia Online already had 26 million users.

Ironically, Gaia Online also jumped on the MMO hype in 2008, releasing its own browser “mini-MMO” called ZOMG! After the success of Candy Crush in 2012 they would even add their own match-three game, Switchem (2013).

Building your own virtual world

Another evolutionary path is the virtual worlds focused on allowing players to build their own houses or worlds — spiritual descendants of TinyMUD.

In 1995, (the same company behind Worlds Chat in the video above) would release Active Worlds , a 3D virtual worlds that users could visit to explore, chat and build whatever they wanted in 3D — from houses and castles to in-game ads of real products or even convention centers where people could go have meetings.

This concept would be revisited in many other virtual worlds in the following years, from 2D games like Furcadia(1997), which is basically a graphical version of TinyMUD for furries, to the hype-craze that was Second Life (2003) — which had companies and even embassies fighting for virtual land and some thought would replace the Internet itself.

A virtual lecture inside Second Life

(TIP: If you want to read an interesting examination of Second Life as an MMO, check this post by Mariane Riis about Richard Bartle’s theory on game world design, followed by a short discussion between both in the comments.)

Of course, the king of virtual worlds where you can create an avatar and build whatever you want is none other than one of the biggest games of all time:

Going back to that Raph Koster presentation at GDC that elegantly summarize in few minutes this massive wall of text that you just read:

In the end, MMOs moved out of the spotlight. We predicted as much. We said that some day we would give way to AR glasses and mirror worlds. MMOs gave you game guilds. Gave you free-to-play. They gave you the profession we now call Community Management. They birthed the farming game that became social gaming. Would there be BitCoin today if not for gold sellers? Because there surely wouldn’t be Minecraft without MUD.

Minecraft sold over 200 million copies. It’s 11 years old now and 126 million people still play it monthly.

EPILOGUE — Roblox is a MUD

First released in 2007, Roblox is absolutely mind-blowing.

I can create a character and go explore a myriad of worlds while socializing with friends & strangers. Forget making a room with unique objects, you can make a world with entirely different rules & gameplay — go to school have lessons, tend a farm, be a pirate, role-play inside a hospital, race, work at McDonald’s, see the Titanic sinking, visit Japanese castles, go to space, shoot people or fly around firing energy beans in Dragon Ball Z-style.

It’s like if the whole MUD explosion — all the DikuMUDs, TinyMUDs, MUCKs, MUSHs, MOOs, etc — was all happening inside a single platform, and you could freely jump between worlds with your friends, or create your own.

This is what 150 million monthly active users are playingmost of them childrenRoblox is a MUD for the TikTok generation.

This article is based on the rants of a bunch of boomers who grew up with MUDs, which I translated into the rant of a millennial who grew up with MMORPGs. I hope I get to read the rants of zoomers who grew up with Roblox-likes or whatever, all laughing at my prehistoric ideas.

If you enjoyed this extremely long article, check out my other ones on the history of RPGs in China and the birth of Japanese RPGs, or follow me on Twitter.

To finish, some basic bibliography beyond all the links I’ve added:

David Kurt Herold and Peter Marol — Online Society in China (2011)
Duncan Howard — An Introduction to MUD (1985)
Edward Castronova — Synthetic Worlds (2005)
Jong H Wi — Innovation and Strategy of Online Games (2009)
Richard Bartle — Designing Virtual Worlds (2003)
Richard Bartle — MMOs from the Inside Out (2016)
Rusel DeMaria — High Score! Expanded: The Illustrated History of Electronic Games (2019)

Dionisio, J. D. N., Burns III,W. G., and Gilbert, R — 3D Virtual Worlds and the Metaverse: Current Status and Future Possibilities (2013)

The CRPG Addict —
The Game Archeologist —
Raph Koster — The Online World Timeline:
The Journal of Virtual Worlds Research:
Jessica Mulligan — Biting the Hand:

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s