Interview: SOE's Yanagi Talks DC Universe Online's Birth
Sony Online Entertainment's upcoming DC Universe Online is a key project for the Everquest publisher, which has latterly been trying to push into more casual markets with its FreeRealms title.
The game is a PlayStation 3 and PC MMO which will feature dozens of characters (from Batman through Superman to Lex Luthor) and settings from the DC comic book universe, as well as customizable superhero or super-villain characters created by users.
DC Universe Online is believed to be due out some time in 2009, and earlier this year, Gamasutra sat down with DCUO senior producer Wes Yanagi to talk about the project.
During the course of the chat, Yanagi discussed the contributions of venerated comic book artist and executive creative director Jim Lee, possible revenue models, and how the studio playtests and balances a game with such diverse characters and environments.
How large is the [Sony Online Austin studio] internal team now?
WY: We're above 70 folks now.
I was just talking to [DC Comics veteran and game consultant] Jim Lee, and he's been on the product since the very very beginning. How, from your perspective, does he work into the pipeline?
WY: Jim and his team, they do a really good job setting the art style and art direction of the product, as you can probably see from the videos and the things that we've shown so far.
The entire world and all the characters are very cohesive, and that's something that can be a challenge on other games without any kind of IP or any kind of art direction from someone like Jim, in which you have a bunch of different pieces that get put together, and they don't necessarily coalesce as a whole.
I know you can't say when it will be released, but how do you anticipate the entire development process? How long has it been in development so far?
WY: It's been in development for about -- I started about two and a half years ago, and I think it was a few months before I started that it officially got kicked off. So, [around that long]. [laughs]
Yeah, it's been a real long time coming, it seems. Are you going to immediately work on expansions afterward? How long do you anticipate this development tale going?
WY: We're still trying to formulate our plan on that aspect, as we're still technically in pre-alpha right now. I imagine that we'll have a portion of our team doing the live events and any kind of live content, and then we'll have a portion that will be dedicated to the expansion pack.
Do you all the customer service and support internally at SOE?
WY: Yeah, SOE has a great customer service staff, and of course, the benefit of working at SOE is there's a lot of MMOs in the pipe. And so, everything from customer service to the platform –- you know just getting the executable out to people and patching -- all that infrastructure is all in place.
How many people do you have working in those departments, if you know?
WY: I really don't know. I can't say.
I'm always curious because usually it completely dwarfs the development team.
WY: Yeah, I'd imagine.
I talked to Blizzard a year and a half ago – they probably have way more now – they had twelve hundred.
WY: Twelve hundred?! Wow, that's pretty insane. That's a lot of guys. But then they have to support ten million folks.
You must have the network infrastructure pretty down at this point, since you've had many many years to air it on this stuff. Do you anticipate any difference for this title or is it going to be using your similar network infrastructure?
WY: We have a standard infrastructure that we share across a lot of the games, and we're really leveraging a lot of that, because it's pretty robust and developed. The demos that we've been showing right now at Comic-Con or at FanFare, those demos were all running off of a client/server environment.
We've even been playtesting over between Austin and San Diego, testing independently and seeing lag to see how that affects the action combat. So far, we've been really really happy with all the results.
With Jim Lee, there's been a lot of back and forth on getting character models to look like how he wants them to look in the comics. Where is the line on what you can do -- obviously there are budgetary concerns -- and how can you really determine what you can take and what you can't?
WY: Generally speaking, we try to get the essence of what he's talking about, if at all possible. And we'll bend over backwards to get that look because that's what his style is known for and it's an awesome style that resonates with a lot of people.
Where we can draw the line is when it runs into a technical issue, where we have something where there's too much detail, or the textures might not fit with the parameters that we have or something like that. When we run into technical issues, we'll discuss that
with him and go over what our limits are. Usually, he's understanding about that and goes, “Okay, I understand,” and moves on to the next thing.
When I was talking to him and hearing him on the panel, it sounded like he's really coming from a good artistic perspective. He seems to have more of an idea of what you can do and not do, when compared to say, directors that come in and are trying to associate themselves with a game product. Have there been any challenges in terms of what is physically possible?
WY: Not really so much a challenge, I wouldn't say challenging. Jim's a really smart guy, and once we walk him through what the issues are, he picks up on it really fast. So, there really hasn't been any issues that way. Maybe some of the more arcane technical server issues might be harder for him to understand, but for the most part, he gets everything and is very accommodating in that respect.
What kind of things are you looking at for dealing with RMT -- real money trading? Is that going to even come up in the game?
WY: We've been talking about some ideas, but those things haven't been solidified yet. We're definitely going to be talking more about that in the future.
And are you guys going with a subscription or item type model?
WY: That's the same thing as with the RMT. All the business side of it we're still trying to figure out. Right now, we're focused on making our core game, making that fun and then figuring out how the business works on top of that.
You kind of have to have it figured out in advance if you're going to go with one of those models or the other, but it's okay if you can't say what it is. [laughter] Because you can't really just build the game, and then put the business model on top of it. But anyway, what do you think of the free to play model?
WY: I think that's a huge potential, especially for a more casual market. You know, just getting lots more eyes in on whatever product or whatever game.
I think so many things have happened in the past through shareware -- Wolfenstein and Doom were huge hits based on that model, and I think that there's lots of potential in that market.
SOE is trying to do that with Free Realms, as you're probably aware of, and I think that that has a huge chance of being a successful product.
I don't think a lot of teams making MMOs have used Scrum. How does Scrum work into that pipeline? Because the big thing about Scrum is that you can meet milestones easier. You don't necessarily have those same kinds of milestones in MMOs. So, how does that work?
WY: For us, the philosophy that we're taking is publishing it internally to the team. We try to take the mindset we're live already and we have our infrastructure in place earlier on where we have our daily builds. They get kicked out, you run the patcher. When you log in, it sees if there's a new build up there and puts it up there.
On an everyday basis we have playtests. All the different Scrum teams -- we have a combat team, a content team, an environment team -- they all do their playtests, and they can see what the other teams are doing, and see how their piece fits into the whole.
I think that's one of the biggest challenges that I've seen in MMOs in the past, that they're so big and they're so massive, and in traditional development processes that I've worked on in the past, you'd have all these teams working independently.
At some point, maybe a year or two between each other, between milestones, they would try to integrate everything, and things would fall apart horribly.
For us, it's really on a three week basis they have to have a demonstrable portion of the game and all the pieces have to fit. And so they're working together really closely on a day to day basis.
It sounds slightly different from the vertical slice thing, in which you make this section and then you go on. But you're a little more compartmentalized and then integrating quickly, that's how you're doing it?
WY: Yeah, so for instance, in the demo that we showed at Comic-Con this year, we built what we're calling a world event, where there's a portion of Metropolis and Brainiac's invading. That involved a lot of different teams working together to build that scenario.
The cool thing about it was that it told us a lot about how our game should be developed, or built, from a combat point of view, you know, where are the issues that come up between, say, fliers and people that are running on the ground.
The other part of it is that the environment guys were really able to push; you know, what do we want our look and feel to be for the end product? And getting all those together in one demo was huge for us.
How do you reconcile that world? You mentioned people with different powers. You got the fliers and the people on the ground. How do you build that universe and not have it break? You got people who can do a lot of different things.
If you've got Aquaman, do you have to have the sea and how do people get into the sea? You've got the air, do you have to have the ceiling past which they can't fly, and that kind of stuff. How are you dealing with those considerations?
WY: Trying not to get into specifics and generalizing that question a little bit more -- a lot of those come out of our daily playtests. We'll put in a certain power set or certain movement type, and then because we have PvP heavily developed in this game, that encourages people to almost exploit it because people are competitive in a PvP environment.
So, pretty quickly you see the exploits come up, and then we have these notices that are on our playtest machines, "Ice powers banned today" or something like that, until the designers can go back and resolve what those issues are.
Again, using agile is great because, as those issues pop up from that Scrum team's perspective, if they're focused on balancing or getting all these powers working together, that can be their highest priority that they have to deal with.
It seems like balance is one of the biggest issues in MMOs. Like you said, there will probably always be exploits, it's impossible to root them all out. It seems like it requires way more playtesting than other game types. Do you have everyone playtest?
WY: We do have everybody playtest just about every day. Our daily routine is we have our daily stand-ups in the morning. While one team is doing their stand-up, another team is in the little play pen that has all our playtest machines. There's a group, they're playing the game, one person's taking notes as people are just chiming in with comments. When they're done they do their stand-up, another group comes in and does their playtest.
I think there's a tipping point in game development where when you still have the core mechanics being built, the game isn't quite fun yet, you have to almost force people to playtest. And luckily, we've finally got to the tipping point where our core game is fun enough that we just have to schedule it and people come. So that's a good position to be in for us right now.
Yeah, you don't have to strongarm anyone anymore. A lot of MMO making companies have a tendency to use the beta as the final QA period. Do you think that's a good thing or a bad thing?
WY: I think that's fine and good from a standpoint. There's only so much that you can balance on a smaller scale with a smaller team. Even if you have like a hundred, two hundred people on a team, you can only do so much balance. Once you get a thousand, two thousand people, then the game dynamics change that much more.
The problems that I've seen on projects that I've worked on myself is that, a lot of time you spill some of your game development and feature development into your beta time frame. At that point, it's really difficult to manage between the priority of getting a feature done versus clearly a major flaw in your game that you have to address that came out of beta, or even just dealing with bugs.
Our goal is to make sure that when we get into our beta phase, that all of our features are done so we don't have that conflict between the two. If you have to redo a feature because of gameplay feedback in the beta, then that's okay, but you don't want to be building something that you knew you had to do out of time.
It seems like it's really tempting, since you know that MMOs are never done. They're never finished, so you're always developing. At the same time, when you do that, you run the risk of showing people a game like Hellgate London. The beta didn't go that well, and people weren't really feeling it. By the time the game was actually really fun, everyone had already been like, "I already played that and I didn't like it." So that's a real danger with a beta. But I guess that's where all the internal QA comes in.
WY: Exactly, exactly. And I think even internally, we want to be really sensitive to that. For us, we're playing it with the team, with our team. It's kind of our product, our baby, and so it's always “the best product ever,” and then we can keep playing on that.
But once we release it internally to other parts of SOE, we want to make sure that there's another level of polish there, so that people are excited about that and then going on to the next step of any kind of closed beta.
About the HUD and management of items and powers and things. For me, what turns me off from a lot of these games is how complex it is. There are all these icons, and I've got to remember what all the buttons do, like hotkeys. How do you figure that out, what's acceptable for people, how much they can actually stomach?
WY: I think there's a combination of things that get based on the gameplay. The game that you're building, the game that you have. I think that a lot of it is, how many decisions do you provide the player at a time.
We give more decisions on where you should be moving and how you're controlling your character because we're a much more action-based game. And strategically, what powers you fire off if you're hiding behind something, if you're picking up a bus, or if you have to dodge a bus.
And that's much different than more traditional MMOs, where it's less movement based and it's more picking a combination of powers or reacting to powers, or watching someone's health go down and then hitting the heal button, or casting the heal spell.
By that nature, since there's less movement, you have to make more decisions or give more choices through the UI or through the HUD. For us, ideally, we have a much more simpler HUD, but the much more complex decisions phase is in running.
To take it to the other extreme, if you look at a first-person shooter, there's a very, very simplistic HUD, but you make a lot of decisions in how you're aiming, whether you're going for the headshot or just trying to shoot somebody, or what weapon you're using.
Do you also try to get some outside views? I don't know if a lot of MMOs actually do playtests, like blind playtests with focus groups.
WY: Yeah, we've actually already started that. We did a couple of focus groups with our Comic-Con build before we went to Comic-Con to make sure the public would be okay with it. I was deathly afraid that our UI and controls might be too complex going in, and so we wanted to make sure that that was addressed.
We have a usability lab in San Diego, where our headquarters are, and they've been bringing people in to play the game and giving us feedback on top of that. And there's other things that we can do with heuristic evaluations and those things. So as we get more UI online we'll be running it through that.