The Making of World of Warcraft

Eleven million players. Two expansions. $1.2 billion a year in revenue. 16GB of your hard disk. Billions of hours played. Acres of news-print. A Sam Raimi-directed movie in the works. A billion-dollar grey market in gold and items. An episode of South Park. Therapists creating in-game characters for addiction counselling. Eleven million players.

The sheer magnitude of World of Warcraft is staggering. There's nothing else like it in videogaming, a fact which makes perfect sense to its players - but is frankly baffling to non-players, and those who don't understand the appeal of the online worlds presented by massively multiplayer online role-playing games (MMORPGs, or simply MMOs). Perhaps because WOW is so huge, so impossible to ignore, it polarises opinions.

Yet all of this can sometimes serve to obscure the most important thing about World of Warcraft. Strip away the hype, the numbers, the media coverage and the debate, and what you're left with is simply a superb game. A game crafted with skill, love and obsessive perfectionism by an exceptionally talented team at the company that built it, Blizzard Entertainment.

If we want to talk about that game - rather than the baggage it has accumulated over the years - it makes sense to go all the way back to the beginning.

World of Warcraft first popped into public view in September, 2001. Blizzard executive Bill Roper had flown into London to make an announcement at the ECTS trade show. Anticipation was high. The already legendary studio had revealed its (as yet unreleased) strategy game Warcraft III at ECTS two years previously, and the announcement of a new Blizzard game was a big deal. Even then, the studio was known for doing few games, very slowly - and very, very well. The smart money was on a sequel to its other strategy hit, StarCraft.

An early sketch of a gnoll [left] defined the look of the game, says Didier.

It wasn't StarCraft II. That afternoon, Roper announced a whole new direction for Blizzard - the company was going to make a massively multiplayer game, letting people roam the Warcraft world as their own characters. Three playable races - humans, orcs and the bull-like tauren - were revealed, and each was going to be fundamentally different. You would be able to play from first-person, third-person or zoomed-out, isometric perspectives.

The reaction was mixed. There was excitement and intrigue, certainly, but there was also a sense of confusion - even disappointment. Why would Blizzard, a strategy-gaming giant, be choosing to muck around with this niche genre? Didn't they know how few people played MMOs? Did they understand what they were getting themselves into?

"It felt like a natural progression," recalls Blizzard's grandly-titled vice president of creative development, Chris Metzen, casting his mind back almost a decade. "We had been working on Warcraft III or different iterations of it for a couple of years at the point when we really started to think about World of Warcraft, and a lot of the creative vision really translated from the Warcraft III experience."

It took a bit longer to get murlocs right, however.

Sam ("Samwise") Didier, the company's art director, interjects. "I'm not sure if this is exactly where it started - but at one time, we had a behind-the-character camera in Warcraft III, much like you see in WOW now. We were thinking of a slightly different, RTS-slash-RPG vibe for the game.

"We ended up going back more to the RTS side, but I remember seeing those first builds of the game: you're running around with the Archmage or the Blademaster, right behind him. You see the horizon, and the enemy camps in front... I think that helped to establish the feeling that, wow, our game would look awesome like this."

It wasn't just this natural progression, however, that inspired World of Warcraft. It helped that the team was spending a lot of time playing certain other games.

"It's funny," admits Metzen, "at the time a lot of us were playing [early MMOs] EverQuest and Ultima Online... Well, we had a secondary development team working on an unannounced project, which we thought was cool, although it was still taking shape - but given that we were such fans of games like EverQuest, the discussion at the time was, well, maybe we can do one of these!"

Playing EverQuest did more than just inspire Blizzard - it helped with hiring, too. Rob Pardo, vice president of game design, had been involved in running a well-known EverQuest guild called Legacy of Steel alongside another player, called Tigole, who was noted in the community for posting incisive commentary on EverQuest on the guild website. Hired by Blizzard, Tigole - real name Jeff Kaplan - would go on to become one of the lead designers on World of Warcraft, and a key figure in the genesis of the game.

Blizzard would soon learn that playing and talking about EverQuest was a world away from actually building its own MMO.

"We knew that it was a mammoth undertaking, but I think, at the time, we had no clue just how big and complex something like this would be," Metzen says now. "Perhaps it was a little bit of naiveté - 'oh yeah, we can bang out one of those, no big deal' - and certainly as we got into it, we would learn that these things are just... very, very complex in their scope."

Even the initial vision for the game wasn't entirely clear. When work started on World of Warcraft, the team wasn't sure of many things, including when it would be set and what it would look like. Many early concepts were much darker and more gothic than the cartoony look that's now so familiar.

"For a good stretch of the early period of WOW, certain components of the dev team had pitched the idea of placing WOW a hundred years in the future of [Warcraft's world] Azeroth," Metzen reveals. "The wars of orcs and humans were long-done, it was a much darker game. At the end of the day, there was a lot of cool art churned out - and this is a little-known story - but as we stood back and looked at it, we just felt that there was a disconnect. It didn't feel like Azeroth."

The interface as it might have been.

Didier takes up the story. "We'd been trying some different things," he says. "We were pushing it a few years ahead, into the future of the world, or rearranging the way that some of the characters would look - and it just wasn't biting for a lot of people. It's like changing Darth Vader from a big scary death knight in the future into a small child. Oh, never mind, sorry, they did that already," he jokes.

"Basically, what happened was that one of the artists on the team just took the Gnoll character from Warcraft III and made it look good at the camera angle that we were using. He kept the same proportions, the same bright colours, the same silhouette with the exaggerated features - and we all looked at that and said, yes, this is what World of Warcraft is going to look like.

"Then he proceeded and did the Kobold, and from then on, that was... Well, it was like, why are we trying to reinvent this thing? Everyone loves Warcraft - let's just make this Warcraft. We don't need any futuristic spin on it, or scary versions of this character, compared to the Warcraft III one. Just make it Warcraft III - that's what the fans are going to love."

Tauren were one of the first confirmed races alongside humans and orcs.

The look, which Metzen describes as "equal parts epic and cartoony", is defining - and it allowed a game with what he freely admits "wasn't necessarily the greatest engine you'd ever seen" to achieve a visual style that could survive years in the spotlight. "Any game will age," he says, "but there's something timeless and Disney-ish about WOW's look. That was a funny secondary benefit of really embracing the Warcraft III look - something it bought us which we hadn't anticipated."

As development on World of Warcraft progressed, the team's lack of MMO experience must have started to make itself felt. Blizzard advertised for designers with experience of the field - a relatively small pool of people in the early part of this decade. One of those people was Tom Chilton, who had previously worked on Ultima Online.

Now a lead designer on World of Warcraft, Chilton recalls that while big questions like the art style or the setting may not have been locked down from the outset, certain game philosophies had been there right from day one.

"One of the core philosophies for the game was that World of Warcraft was going to be more accessible," he says. "That was absolutely a philosophy right from the beginning, before I even came to the team. It was going to be easier to get into, it wasn't going to have the kind of foibles that other MMOs had. There wasn't going to be experience loss on death - that sort of thing."

Other ideas which became vital parts of the WOW experience took longer to crystallise. "The idea that you'd have something to do all the time and the quest-driven nature of the game is something that evolved - it wasn't one of the early philosophies," Chilton recalls. "We built these awesome starting zones, places like Northshire Valley and Elwynn Forest - and then, as we continued to do zones after that, the quests just tapered off. We actually originally expected the game to go into more of an EverQuest-style free-form, where you go out there and you fight monsters until you get to the next level.

"What we found was that all the feedback that we got from our alpha testers was that once they ran out of quests, the game got boring. They were like, 'I don't know what to do any more, and I don't really feel like playing any more once I run out of quests'. We came to that realisation that, wow, this quest thing really works. We need to do this throughout the entire game!"

In the end, these beefy troll models were replaced and wouldn't reappear until the Burning Crusade.

For Metzen, who oversees the fiction of all of Blizzard's worlds, the focus on quests was the solution to a problem he'd grappled with since the start of the project - how to tell stories in the static, restrictive world of an MMORPG. "It was a bit frustrating early on," he confesses. "We were so far out of our comfort zone, constructing something we'd never attempted before - this idea of having to create a static world. It was constraining, not being able to chase a more organic, player-specific story."

"At the end of the day, we concluded that if it does have to be more static, then we sure as hell better input as much heart and character and humour and flavour into the quest experience as we possibly can. It's frustrating as a story guy, sure, because the broad player experience is essentially on a track - but what games like this allow for is the sheer scope of the world, the breadth of the kingdoms and the races, with their own internal strife and their own internal stories, the overarching villains as well as the regional conflicts...

"It was such a big project that needed so much story - so I always felt even if we couldn't get there in terms of unique player experiences, we certainly built something that has tidal waves of story. We felt pretty good about the sheer scope."

Early versions of the game had few quests beyond the starter zones.

Quests became the basic component of the World of Warcraft experience - a decision rooted equally in gameplay and storytelling, Metzen argues. "It keeps you immersed, it keeps a carrot in front of you, it leads your sense of exploration. We wanted story to be the vehicle through which you experience the world, as opposed to chasing down monsters in a field and exploring the world by going from A to Z and getting lost. That was the paradigm in early MMOs - you would simply head out, and if you found yourself in a place that was above your level, you got clobbered and learned not to go there."

The quests delivered another element of gameplay which the team hadn't expected. "We called it the Christmas Tree Effect," Metzen recalls, chuckling. "You walked into a new town, and suddenly there were five or six characters standing around who all had these big gold exclamation points. There's something very rewarding about that - like, ooh! Check it out! There's all sorts of stuff for me to do here! In a purely videogame way, as simple as Mario jumping to catch a coin, that felt really good. It created a fun player emotion."

Chilton joined the team a year before World of Warcraft would launch in North America. He was thrilled to be working for Blizzard, a company he had been a fan of for years - but with 12 months to go, the amount of work left to do on the game was daunting, to say the least.

"I was very surprised by how much stuff hadn't been done," he recalls, laughing. "I arrived and I was like, oh my god! We have to do how much more? We have how long to do it?!"

Not only was the game lacking in content - the level cap at this time was level 15, with little more than unfinished artwork existing for the higher-level zones - but decisions still had to be made about major game systems. "It was still very rough and nebulous about what a lot of the [character] classes would do," Chilton remembers. "A lot of classes were being redesigned, and our entire combat system got redesigned some time after I came on board.

"Huge numbers of game systems didn't exist. The talent system didn't exist, so there wasn't a lot of distinction between characters of the same class, other than what hair or moustache you'd picked. The auction house and the mail system didn't exist. PvP [player-versus-player combat] didn't exist and none of the endgame ideas had been evolved - we knew that we wanted to have raids, but that's all we knew. There was still a ton of WOW that hadn't been built and wasn't really known."

The final twelve months, then, were defined not only by frantic work to get the game into a complete state - but also by a succession of "eureka" moments, as the team hit upon the ideas and solutions that would turn World of Warcraft into the game it is today.

Imagine a world where raids weren't instanced and it was every guild for itself. Blizzard came close to making it reality.

Chilton fondly recalls some of those milestones. "When we finally settled on what the talent system was going to be, that was pretty defining for us," he says. "Before that, there wasn't enough character differentiation. People didn't feel different enough from everybody else. The talent system made a big difference in terms of how people felt invested in their character.

"A eureka moment for our itemisation was when we came to the conclusion that if we were to split stats up on an item, we needed to increase the total stat budget for the item. It sounds like an arcane or a really nerdy thing, but it was a big deal, because we were struggling with how to get people not to just collect items of the same stat. How do you make warriors want items that aren't just pure strength, for example? It was a decision that has defined the way our itemisation works ever since."

Other big decisions were more contentious with the rest of team. The large-scale dungeon raids that characterise an MMO endgame were the topic of many discussions, with everything from the cap on the number of players (figures from 20 to 100 were batted around, with 40 being eventually settled upon) to the question of instancing being argued about passionately.

Blizzard had to cut Battlegrounds from the original release, meaning the only place to fight was out in the open.

Instancing - separating the dungeons from the world itself, so that teams of players could work through them without interruption from other players - turned out to be especially controversial. "Oh yeah, that was a really hot topic," Chilton says. "At some points we even had the idea that some raids wouldn't be instanced and others would be. We were really torn - we wanted to capture the magic of guilds racing towards server firsts, and that sort of thing."

The decision, in the end, went back to WOW's core philosophy of accessibility. "As we thought about it more and more, we really felt like people had an expectation of being able to do the content," he explains. "We would still be able to get some of that feeling of people getting server firsts - they didn't necessarily have to be denying each other access to content in order to be able to do that."

What Chilton remembers as the single most contentious issue, however, was the split between the Horde and Alliance factions - with the sides not only competing against one another, but being prevented from communicating with each other.

"That was contentious right up until the time we launched," he recalls. "The team was pretty divided on that. Jeff Kaplan was pretty against it - he would ask what the point was in splitting up the player-base. People are going to want to play with, well, the people they want to play with!

"Allen Adham, who was the original lead designer, was very adamant that the split was going ahead, because we needed to break the game into teams. If we wanted to have a PvP game, we would want it to be team-based - in a lot of ways, it was like Dark Age of Camelot had done. I was on board with the idea, because I felt like that was one of the advantages that DAOC had over Ultima Online, which I'd worked on.

"DAOC had pre-determined teams, you were either Albion or Hibernia or, er, the other one. Players were split into those teams, and they felt like they were a part of something. It automatically meant that you had people who were friendly to you and willing to help you, who you could join up with - and you know who the bad guys were. In Ultima Online, while not having those boundaries was interesting, it was a huge hurdle for players. People didn't feel like part of anything - they didn't know who their friends were, who they were supposed to be fighting against, what they were supposed to do.

"It was very cool for really hardcore players who could organise themselves into guilds and essentially create their own teams, but it was very intimidating for new players. The split was the right thing to do for WOW, in a spirit of wanting accessible PvP that felt like Warcraft. Eventually Jeff Kaplan came around to the idea - within the last year or so - but right up to launch, there were still people on the team that were opposed to the concept."

A scene all low-level Horde players in the early days of the game will remember - but prefer to forget.

As the launch date for the game approached, the team knew that some sacrifices were going to have to be made in order to get WOW into players' hands. "We had discussions," Metzen remembers, "about whether, if we wanted to get this thing out before we all had grandchildren, we would have to pull back, or scale back conceptually."

Scaling back on the game's ideas wouldn't have fit with the Blizzard way. Instead, some areas were pushed back to be dealt with after release. One of those was PvP - which, when the game launched, was much more rudimentary than the team had planned.

"Everyone on the team wanted PvP in the game," says Chilton, "but we just had this nebulous idea of what it was going to be. We knew that the general philosophy was that we wanted it to be Warcraft III meets Battlefield 1942. A Battlefield perspective, and an action sort of feel, with the epic warfare of Warcraft III - and that was it. That was all we knew. The team was behind the vision, it was just a question of how to accomplish that vision.

"We hadn't really resolved that by the time we launched World of Warcraft. All we had when the game launched was the ability for people to flag themselves as PvP and then fight it out for the sake of fighting it out - then we had PvP servers where people were automatically flagged in the contested territories. That's all we had time to implement by the time we launched."

WOW finally took flight in November 2004.

As the launch date approached, Chilton wondered to himself what kind of subscriber numbers the game would do. "I did feel that we had the best MMO out and that we would do better than other games - I thought that 750,000 to a million, something like that, would have been the upper boundary of what an MMO could achieve. Others anticipated more than that, but..."

They would soon find out if their predictions - and their years of hard work - would be worth anything. World of Warcraft's North American launch date was set: November 23rd, 2004. From that date, for Blizzard, everything would change.