Sponsored By

Interview: Blizzard's Afrasiabi On WoW's Cataclysm-ic Expansion

Gamasutra sat down with World of Warcraft: Cataclysm's Alex Afrasiabi to discuss the expansion, how Blizzard has learned to use less text in WoW -- and how a crucial storytelling tool was borne out of a simple bug fix.

Chris Remo, Blogger

September 24, 2009

10 Min Read

Blizzard's most recently-announced World of Warcraft expansion, Cataclysm, is said to radically alter nearly every part of the game's world, which makes it the perfect game for the studio to show off its storytelling chops. Cataclysm's lead world designer Alex Afrasiabi says that over the game's five-year history, Blizzard has learned a great deal about how to convey information to the player using less text, in service of the well-worn advice "show, don't tell." For example, one of Blizzard's most crucial storytelling tools, phasing -- a system by which a particular player's perception of the game world differs from that of other players based on his or her accomplishments in the game -- was the inadvertent result of a simple bug fix. Gamasutra sat down with Afrasiabi to discuss Cataclysm's extensive scope, how Blizzard prioritizes content development, and how a bug fix became a game design linchpin. I imagine "lead world designer" is a particularly involved role in the new expansion, since it seems to be the most significant redesign of existing World of Warcraft content yet. Is that the case? Alex Afrasiabi: Oh, absolutely. You might go so far as to say any MMO ever, for an expansion. I would say every zone in the old world is hit one way or another at varying degrees, from complete redos like Darkshore and Azshara, to moderate questing changes like Feralas, to moderate redos of the terrain and the quests, to light -- but even "light" is debatable -- [modification] in Loch Modan. Every zone is hit by this cataclysm to some degree. The cataclysm starts out with rumbles, and what those rumbles are are the stirrings of Deathwing beneath the world. He's in this elemental plane of earth locked away in Deepholm. When he finally breaches into terrestrial Azeroth, it causes that gaping wound on the surface of the world -- a cataclysmic shockwave that hits pretty much everything. It's Deathwing, the world-breaker, who is the chief source of this destruction. So you've got that lore. But how do you determine from a development standpoint how to translate that into design, asset production, writing, and so on? How do you determine what areas are higher-priority for more extensive recreation? AA: I don't want to say we play it by ear, because we really don't. We know our game really, really well. We've had a lot of time now -- five years, more really -- in development to hone our skills. Each expansion, in my eyes, gets progressively better. We become better designers of the content. We understand what the players want from our quests and our content, and we try to provide that. We really know our game very well, and that includes [level] one to 60 [zones]. The first thing we did when we set out to do this was prioritize. You basically get that big list of zones, and you give them that -- "This [zone] is a five, the worst. This is just a mess." Like Darkshore. And then, "This [zone] is a one. Moderate work." We basically make this huge prioritized list, and then go through it. Other mitigating factors come into play, of course. What do you do with Silithus? It's a [level] 55-to-60 zone in this expansion. Is it as relevant? We almost have to triage the zones. We know what our production schedule is roughly -- I'm not going to tell you what that is [laughs]. But we have to triage the zones. A zone like Silithus is probably not the best of zones right now. I'd go on to say it's actually pretty bad. It's not as important a zone like Azshara. So it will probably get less of the treatment, because the people who reach that point are probably going to somewhere else at that point [anyway]. It's not as important as Aziara, which is now a [level] 10-to-20 zone for the Horde. That absolutely was terrible before, and now it's got to be amazing. So as you said, you guys have been doing this for well over five years now, and you've learned a lot. What are some of the things you've learned about MMO design, particularly when it comes to conveying story in an integrated, interesting way? AA: The most important one, I think, and this is just from sitting at meetings -- any new guys who come in, they always have that urge to tell their story. "I'm going to tell this amazing story. It's going to make you weep when you read it." That's when I stop them right there. I'm like, "Stop right there. Nobody's going to read whatever you're trying to do. It could be the greatest thing since Hemingway. Nobody cares. Nobody cares. Nobody's going to read it." You have to take a different approach, and you show the player that. It's the old adage: show, don't tell. You show them. It's a different world. That's when you're starting down the right path. When we first started doing this, sure we knew it, but we didn't understand it. There's a difference, and it only really comes from practice. It's almost a zen thing with the quest guys at this point, where it's a [matter of] "Do this quest without any text." Just blindfolded. "Do this quest, and let's see if I even know what's going on. Create something. What's going on? Can I tell if I'm entering this room or entering a point of interest? What am I looking at? What is happening?" I think that's improved our design vastly over the years. Of course, we're still going to have text, but we're not dependent on it. As we advance our technology, too, with quest map [points of interest] and things like that, we'll become less dependant on it. Because right now, what we use it for is as a means of direction. Certainly, we will provide story and lore when we can, but we want to provide that in the actual act of doing the quest. The one thing we still can't decouple from it is directions -- where do you go? But we're getting there. That's certainly something MMOs struggle with -- are people going to bother with the text? It seems like with Cataclysm, that's got to be almost the whole point of the expansion almost. A huge part of the experience as the player is seeing how everything has changed. Can you talk about any design tools or methods you use to strive for that? AA: Absolutely. It's actually interesting. Initially, we created phasing as a bug fix. It was used to fix a bug with the Blade's Edge quest. That was it. Case closed, right? There was this bug, we couldn't solve the problem, and one of our programmers -- a brilliant guy -- implemented this system. Nobody thought twice about it. [Expansion pack] Wrath [of the Lich King] rolls around, and we're in early alpha. We're getting feedback from the team, and one of my friends on the team is talking to me about [the] Howling Fjord [zone], and he's irate. He's saying, "I can't believe this. I go into [capital city] Valgarde, and I keep getting trained by these [native enemies] Vrykul. I killed them, and I did the quest. Why do I keep running into them?" It seems really kind of innocuous. "Yeah, of course. That's how the game works. There's an event playing out. Even though you've done the quest, these events don't stop." But that's kind of what got me to start to more seriously approach it. It was almost a blow to the gut. I was aware of it. It was almost a challenge at that point. How could we change the world for the player so that it actually dynamically alters, so they can actually say, "I did take that quest to kill those Vrykul, and once I did that, guess what? They're gone. They're no longer there." That was all the fire that was needed. From there, it was experimentation. It's funny. If you really break down how Lich King went, the way we tackled zones, we did Howling Fjord, Borean Tundra, and Dragonblight, in that order essentially, during development. Once you get to Dragonblight, you start seeing some of those effects. You start seeing a lot of invisibility -- not phasing -- because at the time, that phasing thing still hadn't clicked. But you start seeing more and more of it. When you get into Wintergarde, you rescue captives or villagers first. Once you bring them into town, the town actually changes. After that, we went onto [the then-new] Death Knight [class], and it was almost a proof of concept at that point. How can we do this? This obscure bug fix just popped up. We were thinking, "What about that? Could that work?" Sure enough, we did a quick run through with a test, went through from one phase to the next, and we said, "Wait a minute. This actually did change, and it totally worked. Okay. We might have something here." From there, phasing was born, essentially, in its current [form]. It became a great tool for us, to be able to tell stories like the battle for the [Undead capital] Undercity. You go to [Orc capital] Orgrimmar, and it's completely phased out into another phase, and you have all these [undead] Forsaken refugees pouring in instantly. You don't need to read anything. You just look. Forsaken refugees are on the floor, begging you for help. The Horde are all rounded up. Shops are all closed -- just straight up just closed, can't use them. Guards direct you to the other cities. It's exciting. That was a big one. So our tools have essentially gotten better. Using phasing is one example, but there are many advancements like that. Vehicles have taken a lot of flak -- some good, some bad. For things that are used for that the player never actually controls, they're actually a very powerful tool for us. An example of a vehicle is the Kologarn -- that's a boss where you have the arms separate from the body. Just using the vehicle tech, he's actually technically just one big vehicle with two passengers as his arms. Again, it allows us to tell this greater story. It's no longer just that boss -- his arm breaks off, and then his other arm breaks off. The technology is definitely improved, and it's helped us tremendously, I think. Were you apprehensive at any point in taking something that was basically a bug fix, and resting so much of the game on it? AA: Well, of course we're apprehensive. But the thing to me is that advances in our industry, and in Blizzard, don't necessarily stem from ideology. Ideology is a powerful thing, and it keeps us rooted. It keeps the foundation firm. But it's ingenuity and deviance, dare I say, that pushes you beyond that. So, you take something like this where it's just this thing for a bug fix and you deviate -- you say, "What if we can do something else with it?" You have to push it. You have to. You have to chase it down and see where it goes. A lot of times, it is a dead end. You'll chase something down, and you're like, "Ah, it didn't work out. Too bad." But you have to push that bounds because otherwise, your game won't grow. That's that. In that vein, people have criticized Blizzard at times for being a relatively conservative developer, design-wise. How would you respond to that? AA: Well, like I said, we do have a very strong ideology. We are firm in our beliefs, and we won't release a game until it's done -- you've heard that said time and time again -- but we mean that. That implies a lot of things. So we'll certainly take the criticism for it, but I think in the end, the result is often great.

About the Author(s)

Chris Remo


Chris Remo is Gamasutra's Editor at Large. He was a founding editor of gaming culture site Idle Thumbs, and prior to joining the Gamasutra team he served as Editor in Chief of hardcore-oriented consumer gaming site Shacknews.

Daily news, dev blogs, and stories from Game Developer straight to your inbox

You May Also Like