Nov 212014

500px-WOW_logoTen years of World of Warcraft. Well. So many thoughts.

WoW has always been a contradiction of sorts: not the pioneer, but the one that solidified the pattern. Not the experimenter, but the one that reaped the rewards. Not the innovator, but the one that was well-designed, built solidly, and made appealing. It was the MMO that took what has always been there, and delivered it in a package that was truly broadly appealing, enough so to capture the larger gamer audience for the first time.

Don’t get me wrong; that’s not a knock on it. If anything, it’s possibly the biggest game design achievement in all of virtual world history. After all, we’re talking about taking a game skeleton that was at that point already almost a decade and a half old, one which had literally had hundreds of iterations, hundreds of games launched. None of them ever reached that sort of audience, that sort of milestone, that sort of polish level.

After a decade and a half of refinements that didn’t do the trick, Blizzard came along and did what it does best: cut to the essence, capture the core fun elements in a game system, and make them accessible and appealing to the broadest possible market.

These days, so many more people have passed through the gates of Azeroth than ever played its antecedents that many don’t even know the deep wellspring sources from which it came. Most of the defining characteristics of WoW are from a long tradition that started around 1990. WoW represents the (perhaps final) evolution of the DikuMUD model.

Dikus are one of many branches off the MUD family tree, and one of the most popular. They defined the pure hack n slash model of MUD gameplay for many. Fixed classes, a level-based grind, clearly defined zones with monster farming and (eventually) quests. A healer, a tank, a nuker in the form of a magic-user, with a heavy emphasis on grouping. Little sense of world persistence; everything always resetting, repopping, respawning. Its audience, from the get-go, was “the kind of people who find adventuring and role-playing more fun than hacking,” to quote the announcement post that appeared on in February of 1991.


A map of Midgaard, stock starting city in DikuMUDs.

Diku proved popular because of its simplicity: everything in a Diku game was hardcoded, which meant you got a working game out of the box. All content was in rigidly defined text files – almost like rigid database entries, with no support for scripts or conditionals. This meant swapping zones with another mud was generally trivial. The ease of hacking the barebones source code meant that a number of “codebases” or modified distributions became available: Alfa, Copper, Circle, ROM, and so on, each slowly advancing the capabilities of the basic experience. Diku spread like a virus across the gaming Internet, and the phenomenon of the “stock mud” was born, where you could log into twenty successive games and find yourself in the same starting rooms of Midgaard.

Many of the features we have come to expect from later MMORPGs didn’t originate from Dikus, it’s true. Housing in Dikus was a pale version of what was truly freeform in scriptable environments like MUSHes or MMOs. The flexibility of coding in LPMuds meant that most of the experiments that later became common features – from player-run guild systems to robust quests to procedurally generated environments – were first seen there, and then adapted into the far less flexible Diku codebases.

Dikus provided core inspiration to many of the earliest MMOs. Meridian 59 had its Diku players on the team. EverQuest was basically a love letter to Diku gameplay, inspired as it was very directly by the classic Sojourn MUD. Half of the original Ultima Online team was Diku players, though UO drew heavily from LPMud and even MUSH and MOO backgrounds as well, and therefore set off in its own, more sandboxy direction.

The Sims Online Newsweek cover

At the time that WoW was in development, there were two big MMOs launching. The Sims Online was featured on the front cover of Newsweek, and tipped to be the game that could take virtual online play mainstream. It had an enormous budget, and of course the power of the Sims brand. The other was my own project, Star Wars Galaxies – another powerful brand, but also far too short a development time and a quarter of the budget.

Both launched buggy, messy, rough, criticized for tedious gameplay. Both offered many innovations. But it didn’t matter: the juggernaut was coming.

WoW had been in development for a very long time, and unlike those other two titles, it was very much directly drawing its inspiration from EverQuest. EQ, the reigning king of Diku-style gameplay, was always resolutely about group combat. At the high-end of play, membership in “raids” (which at the time simply meant “big fights against tough monsters”) was the online teamwork game par excellence.

Rather than try to break new ground on features, WoW set out to instead collate the very best from every game it could, with a relentless focus on the fun. Where other games were chasing high-end graphics, they chose instead to aim low on technical requirements, opening up the potential playerbase considerably, while relying on stunning art direction that was initially decried as cartoony, but which was vivid and colorful and appealing in a way that the grittier other games were not.

And WoW indeed took many of the core features that other games had established, and by and large made them better through the alchemy of recombination and polish.

In Blizzard’s hands, raids became intricate puzzles to be solved, almost like teaching a large group a complex series of dance steps; a design idea drawn, perhaps, from boss stages in 2D bullet hell shooters. Given the influence that EQ raiders had on the WoW design team, it’s no surprise that to this day, raids are the engine driving WoW forward in so many ways.


The sheer colorfulness of WoW was a revelation; EQ had been relatively dark and grimy.

A related tool, instancing, had been invented clear back in the text mud days, but in WoW we saw it used to a degree that earlier MMO designers were reluctant to embrace. Seeing it as disruptive of the “worldness” of the game, myself and many other designers at the time pushed back against making too many dungeons into instances, arguing that the serendipity of meeting other players was core to the genre. But intricate raid patterns set up as patterns to solve can’t survive that sort of unpredictability, so WoW used instancing liberally, making high-end encounters feel ever more like embedded games or challenges in their own right. This had the interesting effect of reducing inter-guild contact, but strengthening the bonds within guilds as teams had to go through training regimens and practice sessions.

The realm versus realm combat that lay at the core of Dark Age of Camelot fit perfectly atop the Warcraft mythos; the underlying PvP flagging system was a variation on flagging systems explored in many muds, and of course famously and endlessly in Ultima Online. We forget, now, how contentious player-vs-player combat was, in those days, and how hard a design problem it was to make it feel seamless. It had been decided that the days of the PvP MMO were basically over, and yet WoW launched with what was basically a PvP MMO and managed to make it feel like a safe space for players.

The chat bubble system had been a massive design challenge on Star Wars Galaxies (chat bubbles had been used in 2D games, such as Habitat and The Realm, and had done some extremely interesting stuff with them which I happily stole, but most MMO designers, alas didn’t play There); WoW cheerfully leveraged that work, but didn’t even use bubbles very much (plenty of mods exist now to enable them), as well as redoing their camera system during the beta to be surprisingly similar to SWG’s (probably convergent evolution).

Troll quest giver,  MMO Costume Contest at Dragon*Con 2008. Photo BY-NC Les Howard

Troll quest giver,
MMO Costume Contest at Dragon*Con 2008. Photo BY-NC Les Howard

Among the things left by the wayside were features that were proven. Gone were the richer pet systems that had driven so much engagement from players in earlier games. Player housing, past and future source of endless devotion (and revenue) in other games, absent. Never mind stuff like towns and politics and the like. Crafting took massive steps backwards from the heights it had been developed into in Galaxies or even Sims Online, and went back to being more like that in EverQuest. Even the robust character customization that we slaved over in Galaxies, a system which today is in every RPG on earth, was gone. In WoW, you basically looked how you looked, and early on the game was a sea of identical clones. The seamless world, which occupied so much technical effort from so many teams, was simply replaced with by-then-old-fashioned zones with network mirroring on the boundaries: a much simpler use-case than solving dynamic load balancing.

None of this is to minimize the effort this took; the work of selecting the right features to make World of Warcraft was precisely the hardest work there was. In the name of shorter sessions, greater accessibility, and easier entry, WoW cut away any “world-like” features in favor of Game, Game, Game.

In their place was one overriding feature, a true innovation that WoW lifted not from the rich and varied history of MMOs, but from the rising design tide in AAA games that was even then up-ending the first-person shooter: the quest-led game.

FPSes in the wake of Half-Life were moving towards today’s world of cinematic narratives intercut with vivid action; Blizzard had successfully made the cinematic campaign a centerpiece of its RTSes such as Warcraft 3. MMOs, however, were still fairly aimless affairs, always struggling with delivering even a coherent tutorial experience. Despite the name, EverQuest was far more about the fighting than stories – the questing in the title was really each player’s own quest for experience points. Even in Blizzard’s own RPGs, the Diablo series, we saw an MMO-like “hub-and-spoke” design, always pulling players back to one central town.


See that SWG waypoint? Way way way over there?

In WoW, despite the word “World” in the title, we saw linearity displace the free-roaming assumptions of the earlier games. Whereas games like Everquest 2 and Galaxies had invested heavily in waypoint systems that had proven to be powerful social and accessibility tools, WoW left this feature out because the game simply never had you travel a long way – the next goal was generally always within line of sight. Helpful symbols floating above characters’ heads, which had barely been seen in MMOs such as Asheron’s Call 2 (they originated in Blizzard’s own Diablo 2), were everywhere.

And the quests, while still built on top of familiar fetch or kill-n-monsters tropes, were also everywhere. A heretofore unthinkable density, much of it deeply humorous and rich with references to Warcraft lore for those who were willing to dig in. For the casual player, the experience of WoW was literally a series of quest completions designed to lead you along a pathway that had been carefully planned to constrain you, keep you always seeking the next glowy mark over an NPCs head.


A snippet of academic analysis of play patterns in WoW in 2006 from TerraNova

This completely up-ended the old MMO pattern, which was more about hunting and gathering, interrupted with occasional quests. In WoW, leveling gameplay was instead about questing, with occasional hunting and gathering. Zones and their quests were crafted hand-in-hand, with the quest chains perfectly designed to lead you across the zone as it increased in difficulty, and birth you out into new zones just as you should reach the point of advancing to the next set of levels. It could only be done with an extravagant budget; indeed, WoW outspent all previous Diku games by a multiple of three or four. But it was incredibly powerful.

Sandboxy players found this to be a massive step backwards. In effect, there was hardly any open space; everything was dense, close together. The rich interdependence systems that were carefully crafted in earlier MMOs were absent. But players who had cut their teeth on single-player games had finally found a virtual world that was accessible to them, where they could “play alone together.” Paired with the power of Blizzard and the Warcraft brand, we watched WoW pick up more players in the first week than the entirety of the Western MMO market combined.

In some ways it was the apotheosis of this game design. Today, the influence is everywhere. To be an MMO has come to mean to be like World of Warcraft. The quest-driven advancement path. The tried and true combat mechanic and the raids. The interface must now conform, much as how FPSes on PC were forever condemned to WASD based on the popularity of specific games. WoW is now the template and the bar that everything must hit.

It’s not a hittable bar, for most. Almost no Western MMOs have ever been granted a comparable budget again, and of course these days, you have to actually compete with ten years’ worth of WoW expansions and additional content and refinements. Slowly, some of those omitted features have crept back in – WoW’s class system is today a much more flexible hybrid class-and-skill system that offers ample personal expression. Given time, the game economy re-acquired some of the characteristics of the sandboxy games – but not too much, in order to retain the focus.


I got your sandbox MMO right here

And the game that was once called, by me and many others, “the least social MMO on the market,” is now the virtual home away from home for millions, as network effects, familiarity, and its ongoing dedication to a great user experience above all makes it makes the place we always return to. It is the genre king, and likely will never be toppled by a game like it, as long as investment in it continues. We likely will not see true reinvention come to the space until there are massive changes in content delivery, such as near-unlimited cloud server power, or virtual reality displays, that both permit and force truly different experiences to be created.

The space it entered as a competitor is largely “dead” in the sense that WoW takes up all the oxygen in the room. Experiments in virtual world design began to dry up, to curtail their ambitions to being nibbles around the edge of WoW’s design. Like toothpaste squeezed out of a tube, the design qualities of sandbox games ended up finding their expression elsewhere, to great success. Their children served as the basis of genres, from Facebook farming games to DayZ-style survival games to the true heir of the MUD tradition: Minecraft, a virtual world based on simulation and crafting, where users run their own worlds and script and build adventures and are basically questless.

It’s just that no one calls them MMOs anymore. That title is reserved for World of Warcraft, and those largely similar games that strive to topple it from its seat. Its influence is such that it now defines the genre it refined. It is the best Diku ever made; the best combat MMO ever made; the thing to which everything like it will ever after be compared. World of Warcraft effectively made MMOs perfect, and in the process, it killed them.

  74 Responses to “Ten Years of World of Warcraft”

  1. Great read as usual! I’ve been wanting to write an in-depth paper on the paradigm shift in MMOs. Couldn’t get it approved by advisor, so it will have to wait, but this was an excellent point of WoW representing the last evolution is spot on. I believe that it maxed out the old paradigm, essentially throwing the genre into a paradigm crisis. I believe a new paradigm is beginning to take shape, but we are still somewhat in the aftermath. Would love to discuss this with you at some point. You also have great analytical points! I could write so many papers just from things you say!

  2. Meeh, UO was way better.

    3/10 post, wouldn’t read again, how much did they pay you for that?

  3. I remember being surprised that you could do quests to level all the way to 80 in WOW. In other games, it seemed like you got to a point where the quests just kind of ended, and then you’d be sitting there camping Giants or whatever to level the rest of the way. It seemed like a nice way to do things; leveling was a side-effect of questing.

  4. “World of Warcraft effectively made MMOs perfect, and in the process, it killed them.”

    I completely disagree with “effectively made MMOs perfect”, but regardless the last phrase should read “it killed the rest of them”. (Even that is an exaggeration.)

    There is an opportunity for something better, for sure.

  5. I worked on WoW. In the early Pre-Alpha days we were lacking level designers for such a massive project. So Blizzard held a contest where QA could “level up” to design. That is if we could complete an “area” in 24 hours. Of course I applied. 23 hours later I had an unfinished level. But it looked very nice. (looking at a few screenshots of WoW edit with my level) Anyways, I remember the WoW Editor being very esoteric. There were a lot of hidden keys for functionality I’d wish I’d known about, such as duplicating and scaling models. etc. The two designers that won clearly discovered hidden features of the editor. In the beginning it was very bare bones and you needed to be very creative to come up with interesting things. I had a lot of fun with WoW editor flora at the time. Flowers, plants and trees were the most advanced and interesting models in WoW Edit.

  6. Great writeup and very interesting. I’ve been playing RPG’s since the 70s, both paper (still have a Grayhawk set!) and computer from Wizardry and Alkabeth. Made more than a few games myself (including some RPGs and a “few” less cerebral, Crash Bandicoot being my brain child, and Jak & Daxter where we integrated that quest scheme very organically with storytelling). Then there was the 10 years of WOW playing, nearly a year of 24/7 realtime lost 🙂 So I wrote up analyses of the first 5 endgames on my blog although I’ve been trying to resist WOD and keep some time for actually working.

  7. […] Raph Koster Analyses World of Warcraft and it’s ten years of influence and growth […]

  8. […] Much is being written this week to celebrate the tenth anniversary of World of Warcraft, but few insights are as illuminating as this post by Raph Koster. […]

  9. […] Much is being written this week to celebrate the tenth anniversary of World of Warcraft, but few insights are as illuminating as this post by Raph Koster. […]

  10. […] Much is being written this week to celebrate the tenth anniversary of World of Warcraft, but few insights are as illuminating as this post by Raph Koster. […]

  11. WoW is a local maximum, but it’s not a maximum. It’s climbed us to the top of this hill, but there are mountains out there in other directions – we just need to find them. Evolution from WoW isn’t going to get very far, but revolution from earlier principles could and eventually will.

  12. […] Much is being written this week to celebrate the tenth anniversary of World of Warcraft, but few insights are as illuminating as this post by Raph Koster. […]

  13. Question for Koster/Bartle really, do you know the history of the random drop, or random loot tables?

    Really, it does seem to be an odd thing to add to a game, since it clearly breaks immersion.

    Its always irksome to kill a rat mob and find a “Vorpal Sword of Infinite sharpness +3”, where was it hiding it?

  14. […] wondered about this after reading Raph Koster’s retrospective on 10 years of WoW, and specifically when he […]

  15. […] it is by outsiders if nobody else, and they seem to have all the money… has been good for MMOs over the last decade is an open sore of a topic.  Are the stifling aspects of Blizzard’s behemoth on the industry […]

  16. […] our biggest games sustain this type of development? The MMORPG genre has come and gone; the ballooning time and cost to build a massive online RPG has made it an infeasible category of […]

  17. Played WoW from the beta till a little past panda-time — and came to it from a DIKU, pretty much, so a lot of this article resonated for me. Great piece, thank you.

  18. SteveConlan: random loot drops date back to the Dungeons and Dragons days. There would be a published table of “Random Treasure” and you’d roll dice to pick items and quantities of coins from it.

  19. […] Link of the Day: Quite an interesting read, from an historic and development perspective, regarding the 10 years of World of Warcraft, written by one of the lead designers of Star Wars Galaxies: […]

  20. @Kytolan Yeah, I used to play D&D back in 1986, but I’m wondering about the instance jump from early RPG (circa Tunnels & Trolls) to MUDs. Was this purely a steal straight from the early RPG tabletops or something else?

  21. Dikus didn’t use random loot much, I don’t think. LPs, depended on mudlib. Unsure about early MUD versions.

    There used ot be huge debates about WYSIWYG loot. See here:

  22. […] Raph Koster’s excellent Ten Years of World of Warcraft article […]

  23. […] even the great grandfather and Yoda of the MMO world, Raph Koster, said in a recent well-written 10 year anniversary retrospective about […]

  24. […] Ten Years of World of Warcraft Raph Koster Gute Übersicht und Einordnung: “WoW has always been a contradiction of sorts: not the pioneer, but the one that solidified the pattern. Not the experimenter, but the one that reaped the rewards. Not the innovator, but the one that was well-designed, built solidly, and made appealing. (…) World of Warcraft effectively made MMOs perfect, and in the process, it killed them.” 2600 Wörter, Lesedauer 6 Minuten […]

  25. […] Back in 2007, one of the fathers of the virtual world genre, Richard Bartle made a controversial statement regarding WoW and the future of virtual worlds. He said he’d close down WoW if he had the power. I agree with him. If I had the power, I would do the same but use the vast resources of Blizzard to create a fantasy virtual world that elevates sound design principles before profits and stock prices. Seven years later in 2014, video game luminary Raph Koster issued a similar statement by saying that Blizzard killed the MMO genre. […]

Sorry, the comment form is closed at this time.