John Staats and I share a couple of important things in common. We've both written extensively about Blizzard Entertainment's culture and games, and we both hail from Ohio. Staats recently self-published The WoW Diary: A Journal of Computer Game Development, which serves as both his memoir of his time in the games industry working as World of WarCraft's first level designer, and an inside look at how the genre-defining MMORPG was made.
Staats lives half an hour from me, so we met at a coffee shop to talk growing up in Northeast Ohio, the formative days of WoW's team, the as-yet-unknown medical condition that forced him to retire from development,
David Craddock: I don't often get a chance to talk with game developers in Akron, Ohio. Were you born and raised here?
John Staats: Yep, born and raised in Akron, Ohio. I went to Kent State and got a degree in graphic design, so I have an artistic background. From there, I went to New York for 10 years and worked in advertising. While I was there, I developed a penchant for making 3D levels. I grew up playing Dungeons & Dragons, so I wanted to walk around in my own levels.
From there I started making maps for first-person shooters: for Quake, and for other various FPS games. I ended up with a portfolio that I could use to apply for jobs in the gaming industry. Unfortunately, a lot of the jobs didn't look very promising. I wanted something that was stable, because I was in a good situation in New York. I didn't want to jump ship, especially changing industries, if [I was going to be out of a job in a year]. Blizzard seemed like a very stable company, so I got the interview and made the move.
Craddock: Blizzard has been relatively stable, more than most. I've noticed over the last 10 to 15 years that the games industry has taken on more of a Hollywood model of production: a studio will staff up for a game, and then everybody gets let go. They end up being nomads, moving from project to project.
Staats: Yeah. It's always been like that; you just didn't always hear about it. In the 1990s, games were smaller, so you didn't need a big staff. With greater technology, you need experts in every single [discipline], plus your budgets go up. It's crazy how expensive games are to make.
It's just a different employment model. A lot of people will work for a company, but they'll do it as a freelancer. Now some people are getting benefits as freelancers. What companies do to reduce overhead is, if you have 100 people, that's, say, two million dollars if everybody takes a month off and then their employer hires then back the next month. Freelancers can charge more money per hour; they usually make more per hour than employees. It usually evens out.
Craddock: Did you play Blizzard games before working there?
Staats: I was a big fan of WarCraft II, but I was also a big fan of all strategy games. The success of WarCraft II and StarCraft meant [strategy games] went real-time. I actually held a grudge against Blizzard because every game went real-time, and so many of them were just terrible.
But I'd never played Diablo. It was too simple, too unlike old-school RPGs [like Fallout]. I kind of went to Blizzard hoping I'd be working on a WarCraft title because that was the one property of theirs I really, really liked. They wouldn't tell me what I might be working on while I was there [for my interview], but I knew it couldn't be a sports game or a shooter. [laughs] They would joke, "It's definitely not a racing game." I was able to narrow things down when I saw, for example, how small the team was. Once I signed the NDA and made the move, they told me.
Craddock: Were you familiar with MMOs?
Staats: It was all EverQuest back then. A friend, my roommate, brought home EverQuest. I quickly arrived at a crossroads: either play EverQuest all day, or make levels. I couldn't do both. [laughs] Level design is more time-consuming even than playing MMOs. I was doing 100-hour weeks. There was an eight-month period where I didn't have a job, and I was spending crazy hours making levels.
Craddock: What was your schedule like while you were learning? Were you teaching yourself? Collaborating with others?
Staats: I remember I started with Quake 1, one of the first true-3D games. I'd spent years making awful, awful maps. That continued through Quake 2. Most of my maps were just... There were so many mistakes that took so long to correct. If a dungeon was too small, too cramped, you didn't know until you'd built it. You couldn't just resize it; you had to throw it away and rebuild. That was your option. People didn't have patience for that, but I wanted to do this badly enough that I would redo my levels over and over and over again. It takes a lot of desire.
I had no idea I'd come out the other side with a portfolio. It was just something to do. I worked with other modders, and we'd check out each other's maps. I'm doing the same thing now with board games. Actually, this place was a hangout for a lot of board-game designers. [Points] That table was where we'd playtest each other's games. We'd critique each other's work.
For level design, there were also lots of tutorials. That's a sub-sub-sub culture of designers: people who do nothing but make tutorials for how to do things. They're doing that now on YouTube.
Craddock: Since you started out making 3D levels, it sounds like you had your eye on working on 3D games even though you liked WarCraft II, a 2D game.
Staats: Oh, yeah. I played around with 2D stuff with WarCraft II [map editor], but it was all drag-and-drop, super-simple stuff. You don't need an artistic eye for that, really. To some degree you do, but 3D level design means you need to be an artist first, and you have to like architecture, and you have to like roleplaying games enough that you'd spend crazy hours working to make a play-space immersive. It takes a lot of time.
Craddock: Especially at Blizzard. With their policy of "we release games when they're ready," you have to be prepared to spend lots of time refining and polish.
Staats: Very much, and if you aren't at the conjunction of all those disciplines, plus enthusiastic about gameplay, you're not going to be a good level designer. Some designers are artistic, but they're not good level designers. It's a hard position to fill. I think I was the only person who was a genuine level designer on the [World of WarCraft] team when I started there.
Blizzard calls them "environmental artists," now. They don't really have level designers in the first-person-shooter tradition of the discipline where they do everything, because they know everything. Right now [at Blizzard], the artists just do art. They don't do layout or gameplay.
Craddock: So you were the first level designer on WoW?
Staats: Yeah, because they were having trouble hiring people. Nobody who was any good [at level design] wanted to go to Blizzard because they had no reputation for 3D games. WarCraft III, at the time, was a joke. They'd scrapped the engine a number of times. Diablo II was all randomized, 2D tiles.
Level designers traditionally don't jump from job to job. Typically, if you leave them alone, they're happy. That was true with me. You don't have to pay them a lot. I was working for nothing. My opening salary was $50,000, which is really low for level design.
Craddock: And for living in Irvine, California.
Staats: Oh, yeah. They were making six figures in Texas on Star Wars: Galaxies. Working at Blizzard had an appeal, but you might have had to live in someone else's house because you couldn't afford the cost of living on your own. Their level designers were all the drag-and-drop guys [on 2D games]. They weren't a super-specific specialist they needed; [management] didn't really know what to do with level designers when they hired us.
Craddock: How far along was World of WarCraft when you got to Blizzard? What did you start out doing, especially considering the role of level designer sounds like it was undefined?
Staats: My first day on the job, they showed me the game. It didn't look good. It was a guy in his underwear--there was no armor in the game--running around on AstroTurf. That was their first grass: gritty, green AstroTurf. There were really crude trees, nothing like Blizzard-quality stuff. But that was what they had in-game, and that was pretty much it.
I remember there was an ogre, and he said, "This is how combat works." The player runs up to the ogre, types something. No swinging of swords, no hit reactions, nothing. Someone told me, "Looting was working last week." There was nothing working at all. I was thinking, Okay... this is really weird.
That was my first exposure to an unfinished game. When I was making levels for Quake and Quake 2, I had the discs with all the finished artwork, all the textures, all that stuff. You could use scripting systems. Everything was done. When a game is being built, nothing is done. Everything is ugly. It was like, oh boy--this isn't fun.
We used the Quake 3 engine to build our first maps. They were little caves and goldmines built from Quake 3's geometry. We all had Quake 3 discs. We'd have the Quake guy with a rocket launcher in these early WoW maps, as if we were making deathmatch maps.
Craddock: And running around at, like, 90 miles per hour.
Staats: Yeah, and that's actually a very good point: Going that fast is no way to tell how big your map really is. It would be really hard to have an idea of how long it takes players to go from point A to B. We were modding Quake 3 just to check out our maps, but they didn't look anything like what World of WarCraft would become. Our maps were all hard shadows, really dark, hard edges on everything rather than the soft WarCraft looked.
We worked that way for six to nine months. Then we threw all that geometry out the wind. That was after working 80- to 100-hour weeks. We decided, "This isn't the right way to build. Let's throw it all away and start over." That's what game development is. Everyone's got that story, not just the WoW team.
Craddock: Was there any expectation at that time that you were working on assets that would make it into the final game? Or was everyone just experimenting?
Staats: We were just building in our own editors. You couldn't even see anything in-game. Programmers had code on their screens, but the game wasn't there. Everything was bare bones.
Craddock: What was the first area you worked on that, even though it likely underwent lots of iteration later, ended up making it into WoW?
Staats: It was a raid dungeon. That was the first thing I built that went into the game. It was terrible, inflexible geometry. I was learning a new editor and building everything wrong. The texture artists hated my textures because I didn't build them correctly, and I hated them because there weren't a lot of textures made for the game, so I couldn't make one area that different from another.
No one liked the game back then. Even the texture artists didn't like their textures, but they were still figuring out what worked. Eventually all those textures got scrapped. I think we hired somebody a couple of years later who repainted everything. The goldmine was probably the very first [area] that survived. The textures were mushy. It was just a different style from the final project. Everything evolved so much. It takes months to do a single dungeon.
Craddock: How long were you the first and only level designer on the project?
Staats: I'd say about a year. We hired a couple of people from the FPS community, but neither of them stayed on long, especially after we switched away from the BSP editor. That's what they were used to. They liked what they knew, and that's how people are.
Everybody left after the game shipped; I was the last one there. It was just burnout. The problem was it was a game that took so long to make, and it wasn't changing. Usually you ship a game, you're done, and you go to the next title. With WoW, it was rinse-and-repeat: updates, expansions. Just the same thing. Usually people jump ship to go to art director positions at a smaller company or something.
Craddock: I've talked to a lot of developers from Blizzard Entertainment and Blizzard North about that. When a Blizzard game comes out, it's a global event. Everybody knows your game, and that's a great feeling. At the same time, you're looking at peers in the industry who shipped four or five titles in the time it took you to release one.
Staats: That's true with MMOs. On a larger project, at the beginning, you come on board and you're not wearing as many hats. You're pigeon-holed: instead of working on a team, you're working for a department. You don't learn as much as you would by going to a team with cross-disciplines that need lots of people. Smaller teams are just easier to work on.
Everybody gravitates to that, every studio. Once you hit 30 to 40 employees, people break down into tribes and departments, just because it's easier to keep track of people that way. That's true of advertising, game development...
Craddock: Any industry.
Staats: Yeah, and any company. Just growing pains.
Craddock: As a level designer on the original World of WarCraft, did you shape your levels around a story or quest being told?
Staats: Usually, that came from a creative director. Sometimes you come up with your own idea, run it by a creative director, and he or she gives it a thumbs-up or thumbs-down. Some are a lot easier to work with than others. Some are jealous of anybody who tries to add to their vision.
Luckily we had a great creative director. I've heard horror stories about people being standoff-ish and protection of ideas. Chris Metzen from Blizzard was the exact opposite. As long as your idea didn't conflict with something, he was always saying, "Yeah. Go. Rock and roll." He was not details-oriented. As long as your idea fit the general vibe of the game, he trusted people to just go.
Craddock: Those tend to be the best managers, right? People who hire employees and let them do what they're best at doing.
Staats: Oh, yeah, for sure. Every management job is different, but once Metzen knew somebody got it, he could just say, "Okay, this cave is filled with monsters. It's basically the Isle of Dread from Dungeons & Dragons. You got it? Good? Okay, I'm outta here." Literally, in a couple of minutes, that's what you'd get. Then you were free to come up with ideas around that.
That was one way of [leading]. There are a million ways to do it, but I think companies who hire book authors are just doomed. It's a very different discipline. I even talk about this in my book. Lore, at least at Blizzard, is the first thing that changes. I've been reading up on George Lucas and Walt Disney, just studying up on their takes on narratives. I recommend the book Triumph of the American Imagination by Neal Gabler. It's a really long book, but it's an exhaustive account of Walt Disney. He's pretty much the first person in the entertainment industry who instituted a hyper-polished approach to content. Blizzard has that reputation in games.
Craddock: How do you define storytelling as a level designer? What's your role in that position?
Staats: Immersion is what people really enjoy about a game. That's what they mean when they say "storytelling." People say Half-Life has the best story, but if you look at the narrative, it's as bare-bones and corny as any other shooter. But when you look at artificial intelligence, level design, scripting--those work together to immerse the player so they can project themselves in the game.
Storytelling in games is a Borg weapon. It has to be simple because you're getting chat messages, combat messages, you're focused on the UI, their health, the combat, traveling. You can't have a subtle story with all that going on. It's just not going to combat. Players would rather stay alive than really appreciate the subtle things you can do. In a book, you can do that, because there's only the book. Same with a movie.
When you bring a storyteller from books or movies into a game, they're surgeons. They're these fine artists who will come up with a fine idea, but the programmers will say, "We can't do that." That author could have spent 16, 32 hours, who knows, on that idea, and it's shot down in a second. That's what game development is. Story isn't the lowest common denominator, but you have the determine the bandwidth you have to tell a story.
Somebody made a dragon at the end of World of WarCraft. I think he spent four weeks making this dragon, and usually you're spending maybe five days on something like that. So, okay, our first raid boss is a dragon. That's the story. It's not some grand arc. I've heard of some studios where art directors rule that way, and they end up getting not as much bang for their buck. They have tons of people working to pull off this one story element, and okay, it's great, but that comes and goes. Then you've got the other 98 percent of the game to make.
You have to get a lot of bang for your buck. That's why story is the first thing to change. In this book about Star Wars, I learned there wasn't originally a Princess Leia. Someone pointed out, "George you need female characters." Luke Starkiller was going to rescue a male friend, so Lucas changed that friend into a princess. Classic fairy tale stuff. This was going to take place on Cloud City; they wouldn't be going to the Death Star. It was only when a Fox executive said, "No. We're not making a Cloud City. Use the sets you have." Lucas said, "Well, I suppose that's better. We have some sets for the Death Star." Lo and behold, they came up with the Death Star.
That's how stories are told in collaborative mediums. A lot of people don't realize that.
Craddock: That's so true. I've had some novels published, and I've also written for games. As an author, I'm hidden away in my writer's cave with an endless budget--
Staats: [laughs] Yeah, yeah.
Craddock: --because I'm like, "I can write whatever." But when I worked at a game studio, I remember going to one of the directors and saying, "You know, the way NPCs give dialogue, it's just one line in the dialogue box at a time. You have to keep clicking an arrow to get the next line. It's so tedious. Could I just get a single dialogue box?" Engineering came back a few days later and said, We don't have the budget for a scroll bar." I was like, "... oh." That was a huge lesson for me: In collaborative storytelling efforts, you have to consider every discipline and all resources, not just your own.
Staats: Yeah, so you get it, right?
Staats: But a lot of people who want to get into games aren't programmers or artists, so they say, "Well, I could just write the story." The value of a storyteller in most games is, I'm sorry to say...
Craddock: They'll write on their resume, "I am skilled in Microsoft Word."
Staats: [laughs] Exactly. It's a tough position because you have a lot of studio heads who want to be the storyteller also. It's just a really hard gig to get, and you're only relevant for the last five percent of the project, especially if you're writing your own engine. If you licensed an engine, then okay, you can ramp up. Usually, the story is whatever elements get created. I've seen studios that work the opposite way. Bigger companies that work on console games have console budgets, which are 10 times what computer budgets are.
Craddock: You'll also see games that play more like interactive novels or movies, and gameplay is a certain type of interaction, usually making decisions that move the plot and character arcs to the next beat. I believe with Telltale's The Walking Dead, they had a story more or less hammered out first.
Craddock: You mentioned immersion as key to telling stories as a level designer. What are some details you add to levels to do that?
Staats: Sometimes you can't. Sometimes you don't have the art assets. If you can show an abandoned caravan, great. If you're trying to establish a setting, you hope you have the art assets to do that. You can make it somewhat on your own if you have the skills to also make art assets.
Level design is kind of weird because there are all kinds of level designers. There are scripters, who are the storytellers. They place entities in the map that walk, talk, and so on. I'm more of the architect: I'm dealing with architecture and solid stuff. Dealing with, say, scripted monsters throughout a level is a completely different set of skills. They're not artists; they're more on the engineering side whereas I'm more on the art side.
Those guys, when they see a room that I've made, they'll look at the elements. If I built an altar--usually because I was trying to think of something to put in the room--they'll say, "Oh, the altar's part of the story. What can I do with that? Well, there are alcoves nearby. Maybe there's a relationship between the alcoves and the altar." That's where a lot of stuff happens. Sometimes you'll have somebody make a blueprint and hand it off to the environmental art team. One guy may work in Adobe Illustrator making blueprints for the art team to implement.
All stories are so different. Some just tell themselves. Others are revealed and happen in a specific room. Some dungeons I made didn't have much of a story. You're saying, "All right, let's just do a cave and we'll put a boss monster at the end." It's more or less finding the milieu for the story and jamming out ideas that fit that style. I did tombs for World of WarCraft, all the crypts and stuff like that--anything for the Undead. They said, "Oh, if you add a few more rooms, we can make that into an instanced dungeon." That became Scholomance. I had no idea they were telling any story at all. Sometimes they look at the things I'm creating and say, "We could tell a story in this spot."
So, you can start with a story, or end with a story, or retrofit one. I think a bit of both is the right way to do it. Whether your setting is the Wild West or a pirate cove, as long as you have an identifiable genre, you can come up with a lot of ideas for it. If someone says, "We've got a bunch of black panthers with tentacles that also vibrates so it's hard to see. Try making a dungeon for that." It's weird because you don't really know what you're doing. Should it be a cave?
Probably the Undead is the easiest [theme] to create for. People respond to skeletons, spiderwebs. We all have very real-world reactions to those things. We don't often see pirates. We're not that familiar with the Wild West. But we know what a coffin is, and why it's scary.
Craddock: Did you just gravitate toward creating environments for the Undead race?
Staats: We volunteered. We had so many dungeons to build. The producers would say, "Who wants to build this?" and one of us would say, "Yeah, I could make something like that." Take the Undercity, the Undead city. That was the one area in the game that made me quake in my boots. I didn't want to build it because I had no vision for that whatsoever. I had no idea how to build, for example, a sewer system for an underground city.
That was one instance where the dungeon designers took over the idea and changed it from Chris Metzen's idea. They were so passionate about their vision, and they came up with an area that was so Gothic and beautiful, and I'm so glad they did that.
Everything else I made for the game was something I wanted to build. To be honest, I wanted to build everything. That's why I built half the dungeons. I volunteered for everything. I just had a whole bunch of ideas for this stuff, plus I had a history with Dungeons & Dragons.
Craddock: What are some aspects of level design that players may not notice or be able to put into words, but that you, as someone from that field, know are super-important to any area?
Staats: Immersion is the one thing I really care about. Some people will say art style as a way to put their personal stamp on something. I like interesting layouts, and being able to reuse rooms. If two big rooms are next to each other, I like to break the wall between them. If you give players a window that looks into another room, you're providing them with a different viewing angle that excites the eye.
You're trying to excite the eye as much as possible. Often you want something in the foreground, mid-ground, and background. What you get is a parallax effect. Players can tell they're moving forward if there's a parallax effect. By walking, they feel like they're making progress. If they're in a long hallway and there's nothing good to look at, it's not as fun. Even though you're just walking in both situations, walking through one area feels more rewarding than another because you're giving people a sense of moving forward.
I also like making rooms that haven't been seen before. There's one room in Blackrock Depths where I made the ceiling and floor, but I got rid of the walls and made a lavascape: river in the background, a bridge far-off. There's a lot to look at, and it's claustrophobic, but beyond all that is this huge scene. It's a refreshing change from the hallway-room-hallway layout of a dwarven city. Having varied geometry is important.
Craddock: We've talked about the hours you worked. I've spoken with developers who didn't mind crunch when they were younger. They didn't have families and were happy to spend all their time at a job they loved. When did you start feeling burnout? Or did you?
Staats: I was used to it. When I started, I was 30. I'd spent 10 years in New York's advertising industry. People would stay at work even when there wasn't work to do. They were the second or third generation of account executives. That's how their parents worked, so that was just what you did. If somebody else was working, you'd hang out at work and keep them company in case they needed you.
Southern California had a totally different culture. I worked longer hours because I wanted to. Other people worked 40-hour weeks. There was one guy who was from Painesville, Ohio. He said, "I want to clock in and clock out. I don't care what game we're making as long as they pay me." That was fine. He was a great texture artist, and that's what he wanted to do.
Craddock: I know a medical condition forced you to leave the industry. In fact, I believe you have trouble even playing games for prolonged periods. What is that condition, and how did you get it?
Staats: Man, it's undiagnosed. Let's say I'm modeling a palace. I'm creating a wireframe, and I'm working on one corner of the palace. The screen is flat, so when you're looking at corners, I can't tell the difference if it's a far corner or a close corner until I skew the distance; then I can see the parallax of vertices moving quickly or slowly. When you're 3D-modeling, you're constantly [skewing perspective] just to make sure you're manipulating the correct vertices instead of something else.
You're also using hotkeys a lot. Both of your hands are constantly clicking and tapping. It's way more intensive than just typing. When you're typing, you pause and think. With modeling, I was doing that all day long. When I write, I can work maybe four to six hours before I'm burnt out. I started out using Dragon Voice to Text. It's really good, way better than dictating to your phone, and Microsoft's [Word] voice to text. That's what I used for my first pass on my book.
For editing, I can type, but I've got maybe two to four hours a day in me before I want to take a break. Level design is 12 hours, 14 hours, 16 hours. It requires a long day to get anything done at all. And I don't play games. I'd rather use energy to answer an email than play games. It's an either-or scenario.
I've had experimental drugs, everything for arthritis, but I don't have symptoms of that: my hands aren't puffy. Right now, as we're talking, I haven't used my computer today. I've tried tablets, and I'll notice that the tips of my fingers, if I'm texting a lot, start aching. Both hands are affected. Nobody knows what it is.
Craddock: I understand you started writing your book at Blizzard. What compelled you to start documenting the process of making World of WarCraft?
Staats: Things were such a mess. No one had answers for anything. BSPs weren't working. They told me 3D Studio Max wasn't going to work, although that turned out to be a good tool. Until programmers can spend four months testing something and then say, "Yeah, this will work," we had no answers.
I was amazed that one of the top game companies in the world was figuring out everything. WoW was crashing and burning. WarCraft III was crashing and burning. Everything was such a mess. It was also my introduction to game development, and I said, "This is fascinating. I want to tell this story from a man on the street's perspective on what it's like to work in this industry." The book isn't technical. I'm not a programmer. I wanted to dumb it down for anybody to understand what it's like to work at Blizzard, and what it was like to make World of WarCraft.
I took notes every month. I'd go into my email program and see all the messages about things that had happened: We were right about this, wrong about that. I also worked on Saturdays just because I didn't like Southern California, so that's how I spent my weekends. I could get a lot of work done. One Saturday a month, I'd take notes. I also interviewed people to educate myself, and I'd tell them what I was working on. I talked to the CEO, the founders.
They were cool with it. At the time, Blizzard was only around 200 people. They weren't too worried about employees going off the rails and saying things. I wanted to publish after the game was done, rather than when the game launched. Blizzard has this credo of nobody taking credit for everything, so everybody can take credit. That's one reason I waited so long: the bodies are dead and buried. It's a totally different team, so nobody cares who was the art director, who came up with transparent water. I don't want to be the person doling out credit. I think the book is pretty accurate, but I didn't want to be the person who got things wrong. It's such a collaborative process.
Craddock: How'd you choose to use Kickstarter?
Staats: I definitely wanted to publish on my own because I wanted control over it. I wanted to have credibility. I didn't want this to be a PR piece. I designed the book to look like a textbook. It's not colorful. It doesn't even look like it has anything to do with computer games. Even though I'm sure this cost me a whole lot of sales, I wanted the credibility of, "This isn't Blizzard writing its autobiography focused only on the sunny side." There's a lot about the squabbles, the debates. I wanted to humanize the story.
I wrote the whole thing and took it to Blizzard. Their chief of staff was my old roommate, and I'd played poker with Mike Morhaime and all those guys, so they gave me a sweetheart licensing deal. But it took them nine months to get two pages of an agreement written. But they need time to get everyone on board. They have to make sure I'm not saying anything that would damage the company. They didn't change anything except for one quote, because they couldn't verify whether the person had said that. So I said, "I'll just stick it in prose and make it look like I said it."
The first Kickstarter failed miserably. I thought I could just say, "Hey! Blizzard guy over here!" I think I got 300 backers. I changed the scope of the Kickstarter, and I knew nothing about it before that. I hadn't even bought anything off Kickstarter. I sat down with Gerald King, who's a consultant for Kickstarter campaigns. One of the things he told me was, "You need to get [coverage lined up with] websites early, six months ahead of time at a minimum."
Running a Kickstarter is as much work as writing a book, if not twice as much more, but you do have control over things. I did the video, the PR, everything. My next project is a board game, and I'm going to do everything myself again.
David Craddock posted a new article, After the Fact: John Staats on Storytelling in World of WarCraft