It's impossible to get a completely comprehensive look at a developer with just two interviews, but with this set of two interview sessions conducted earlier this year, Gamasutra hopes to present a more rounded look at Seattle-based Bungie, the studio behind one of the most successful franchises in history - Halo.

To begin with, Gamasutra presents the full version of a discussion with environment artist Mike Zak and community and PR director Brian Jarrard, originally excerpted in Game Developer magazine. It touches on everything from the interplay between art and engineering to the mood at Bungie following its October 2007 emancipation from Microsoft.

This is followed by a discussion with Chris Butcher, an engineering lead at Bungie, who was responsible for implementing the innovative and robust online features in Halo 3.

This is a pretty exciting time at Bungie, I'd imagine. Not only did you just ship one of the best-selling games of 2007, but you've claimed independence at the same time.

Mike Zak: Yes. It's a great time to be at Bungie.

Halo has always stood out from a lot of first-person shooters through its naturalistic environments, and even the city or urban environments have that certain something... distinct. If you look at any screenshot from Halo, even if the HUD weren't up, you'd know it was from Halo.

MZ: Awesome. Thank you for noticing.

Hey, it's not brown.

Bungie/Microsoft's Halo: Combat Evolved

MZ: (laughter) That's really important to us on the art staff. There's a precedent set by Halo 1, and there's a lot of gritty, realistic games out there that do that really well, but for us, we're more inspired by slightly more imaginative spaces and more awe-inspiring vistas and ideas.

Personally, it's a dream franchise to work on in environment, because I get pretty tired of bricks and rusty pipes. Not that I don't admire a lot of games that have really impressive bricks and rusty pipes, but you get the opportunity to let your imagination go with Halo environments, and really build spaces there. You might want to hang out there, to not necessarily fight in, but to have a picnic or go for a hike.

Do you think that serves an important gameplay function, too? The way you build your environments -- how does that affect and enhance the gameplay?

MZ: I think they give better choices. I wouldn't know if those inform the gameplay as much as the tone, so it's not in a direct way. It depends on how you want to define gameplay, but if you want to talk about gameplay as the specific mechanics of character interaction and the fighting mechanics.

Probably the most direct way that the aesthetics and the level environments match up would be, for example, the Forerunner architecture style really lends itself to building playgrounds that are architectural and structural, but are inscrutable enough in function so that you can really go to town with things that might be a little too gamey.

For example, in a contemporary shooter where you're going through bombed-out streets. You can't have platforms and ramps and weird conduits the way that we can. In some ways, the theme really supports that, but at the end of the day, we can do a lot of that with rocks and terrain and whatnot. You can have the earth serve the same mechanical functions.

Do you work primarily on the multiplayer or single-player environments, or do you just mix them up?

MZ: Personally, I'm primarily focused on Campaign, but during the main title release, I tend to pinch-hit occasionally in multiplayer, with just a quick collaboration and to check it out. Sometimes I'll hop over and finish out a space. The time I really focus on multiplayer is in downloadable content construction. On Halo 2 and 3, I actually got to make a map, which was really fun. I'd love to work on multiplayer full-time as well, but I can't do both necessarily.

Well, the scope at which a game like Halo is at, it's difficult to even own a large chunk of anything.

MZ: It is, and it's becoming more collaborative at every step.

So let's move on to the interplay between design and technology - they're each limited by each other in a lot of ways. How did you reconcile those? 

MZ: Absolutely. We're constantly evolving those targets and the development pipeline... evolving the art bar, is what I mean by "target", and that art bar is defined by budgets that we're engineering now. Especially with the 360 and the GPU and the CPU and the complexity of current-gen shader systems and physics and so many things going into the game. There's no simple answer.

When I started in the industry, it was pretty much, "What's my poly count? What's my texture budget?" Now, I can't even get a straight answer from the programmer, because there is no straight answer. They have the most elaborate profiling tools and are constantly tweaking performance based on, like, "Are you throwing a lot of grenades?" That's very different if you were driving six Warthogs, or if it's just you and a sniper rifle.

So do you sit down with the design team?

Brian Jarrard: It's like trying to wrap your head around this problem of... you have the engine. The engine can support either a lot of things like Warthogs or rocket launchers or something, and you have to design a map.

What kinds of things you put in the map are influenced by the design of the map, and vice versa. So if you're making a map for Warthogs, you'll have hills you can jump off of or whatever. But at the same time, the technological limitations might be based on how many Warthogs might be in the level. They're so interdependent that it might be difficult to find that spot.

Bungie/Microsoft's Halo 3

MZ: Absolutely. Again, we had the benefit of... in Halo 3, we had a pretty good idea of what our sandbox was. Certain things were more clear than others. The Warthog behavior was pretty similar, but the Scarab, for example, became a pretty different beast. In Halo 2, it was not really real. We faked it in a way that... in Halo 3, in the environment, it's AI-driven, it's stomping around, and if it stomps on a Warthog, that Warthog has physics and explodes.

And you can get on it and ride it while it's moving around. That's not something that you necessarily know what the target's going to be early on, so you prototype spaces and test that stuff early if you can. Some things come on late, but you can proceed with a lot more certainty with a space that's using Warthog combat that you've done for two games, whereas the Scarab is crazy and scary. And then you decide to throw two of them in the level and you're like, "Wow! Even crazier!"

Obviously the network requirements of Halo are... I don't want to impugn anyone else, but they're probably the most robust of a shipped shooter, and the complexity of what you can do with recording matches and having people in and out. 

MZ: I understand it's pretty sophisticated.

And that has to affect things in a way, too, with performance issues.

MZ: Performance, absolutely. As far as our actual networking, on the environment art side, it doesn't impact us that much. There's certain things for split-screen that we have to think of, and we test thoroughly obviously over Live connections and system link and all that.

BJ: In regards to safe zones, though, one of the things that you guys used to joke a lot about was... giving the player the ability to go anywhere they can in the world at any moment in time. All of the old tricks of the trade kind of used smoke and mirrors. You guys can't hide anything.

MZ: That's very true. Well, no, there is somewhere to hide. It's beyond the soft ceiling.

BJ: It's like the inside of the Pelican is now fully rendered, and you can fly inside if you want to, even though as a player you'll never go there.

MZ: In a way, we were initially kind of dreading that. Like, "Oh man, you're going to be able to peek at all of our laziness." But actually at the end of the day, I think it was kind of good, because it actually more often than not... the screenshots I saw where I would be afraid of exposing a bad angle... users were actually coming up with better viewing angles than we did.

A lot of times, they're actually getting to see stuff that we had to model for that one percent edge case, where somebody could grenade-jump up. We were like, "Okay, I have to model that and texture that, but I know no one's going to see it." But now, tons of users see it, because they're flying the monitor up there, and that was awesome.

We didn't necessarily author a lot more content than we did on Halo -- well, we did more than Halo 2, but for different reasons -- than we otherwise would, if we didn't have safe zones. In some ways, it's actually kind of satisfying, that some work paid off that you didn't think would.

I remember when the beta came out, and somebody clipped through Master Chief's helmet somehow, and found the face texture. Those are the things that must be hard to anticipate.

MZ: Basically our golden rule of making something bulletproof is that it will never be bulletproof. The players will always find a way to break your game, especially when you expand to the player population that we have. We have an amazing test team, but there's no way they can compete with several million users, you know? It's just impossible. There's a lot of smaller games that get away with a lot more because of that.

At the same time, having millions of people play your game is the reward as well.

MZ: Oh, absolutely. I'm not complaining.

Yes you are! I have it on record.

MZ: (laughter)

BJ: There's a whole population that thrives just on breaking Halo, and doing stuff in Halo 1 kind of built this hardcore following, because these guys are trying to break the physics by launching Warthogs in the air and doing these things that would never happen intentionally. In Halo 2, I think a lot of that went away, and in Halo 3, people feel free to explore and be more strident and do things they couldn't do before.

MZ: I think you could argue that it's that sort of adventurous spirit of the user base that has inspired things like Forge. We want to empower that as well, because it's way more exciting.

Something I want to talk about is the interdisciplinary nature of the team. Some developers have teams sit in interdisciplinary pods, and you guys don't. Your colleague Chris Butcher says he doesn't like the idea, because you can't find the people at hand to solve a problem.

Say a programmer's in a pod with other people like an artist and a designer, and he has a programming problem, and he doesn't have another programmer to his left or right, and he can't grab them and get an answer out of them. What do you think about those kind of team structures?

MZ: We do have interdisciplinary pods at times, not to contradict Chris.

I don't think he said absolutely not, but typically.

MZ: We sort of cluster by general discipline, but our desks are angled at 30 degrees so we can build... we don't have a grid system, basically, in our office. We have a giant, open layout, and we can cluster people.

So we generally have programmers sit in the same general area, and art and design and etcetera, but we try and pair up... like, a programmer who is working on animation systems should sit near the animators. So he'd be on that fringe of the programmer side. And environment designers need to sit near mission designers, because they're constantly iterating together. So that we definitely try to figure out, but there's...

BJ: Like a gradient kind of dwelling.

MZ: Yeah, and we shift it around a lot, too. If somebody changes position or assignment, then they might move desks. We try to keep it fluid.

I think he said that they have wheels.

MZ: Yeah. All of our wiring is in the floor. We have a false floor and it just comes up through there, so we can just unplug, move, and... the IT team must... it happens all the time.

BJ: Actually, the IT team is doing it while we're gone. When we go back, we're going to have a whole new work area.

MZ: Yes. It happens all the time.

Do you think that's good? What do you think is the best situation?

MZ: It takes some people a while to get used to. They're used to having their own office, or sharing an office with like two people. I came from a studio that had separate offices, but I would never allow that. If I ran a studio, there would be no question. I'm a firm believer.

How you segregate or integrate in an open plan... there's no perfect way to do it, but I absolutely believe in an open plan. There's just so many conversations that never would have happened, or so many people who wouldn't know each other well enough... emergent ideas that can bat around.

There is definitely a problem with being able to concentrate at times, but you just learn to love the headphones. Half the time my headphones are on and I listen to music, because I need to concentrate and slightly dull the noise. People tend not to include you, when they come talk, because they know they're interrupting. Simple systems that emerge.

BJ: That was the real motivation for moving to our new space, primarily. We were in a cubicle farm, a generic office space like the rest of Xbox. Like Mike was saying, we didn't even have a big enough space for the whole team to congregate. People would go a whole work day and never see somebody, and they'd sit in the corner behind some wall.

MZ: And this was after we knocked down all the walls.

BJ: Yeah, we knocked down all the walls in this building, and it still didn't work for us. Now our space has the custom design of being completely modular and totally wide open. I will say that anyone who has come to visit us from other developers and partners... everyone seems jealous and envious of our space. It may not work for everybody, but everyone's like, "Wow, this looks really freakin' cool. I wish we had something like this."

MZ: The other thing I really like is that it democratizes the feeling in the studio. Producers and studio management sit right in with everyone. There's no corner office. Well, Marty's got his ivory tower, but you could argue that sound actually does need to be isolated.

I think there's a credible argument there, yeah.

MZ: But beyond that, it feels good as a developer in the trenches when your producer, who doesn't actually have a deliverable for that milestone, is still there crunching with you or sitting right next to you.

Is it easier to get the "all hands on deck" feeling?

MZ: Yeah, definitely. It allows you to feel like it's okay to have conversations with people you might otherwise avoid. It's definitely a cultural thing.

The consensus seems to me, from talking to developers, that interdisciplinary working is so key to making quality products.

MZ: Absolutely.

They're all difficult problems to solve, which is why we like to talk about how they're solved at Bungie.

MZ: The precedent was set a long time ago at Bungie. There's this desire to in some way preserve the "ten guys in their garage making Minotaur"... that feeling, even though we're a hundred people. We know it can't be exactly the same, but it's sure a hell of a lot closer than if we had our own corner offices.

BJ: A much bigger garage.

MZ: Yeah. Exactly. Closer to an airplane hangar.

What is the mood of going independent again? What is the mood like at Bungie now?

MZ: There's an enthusiasm, in a way. You really feel like you're controlling your own destiny. Being part of such a big organization such as Microsoft, where you're a part that plugs into this massive, massive organism that has revenue streams well beyond anything you're ever going to have visibility into... you're just...

The Indonesian version of Windows, or something...

MZ: Well yeah. When you're all sitting in one room and know you're all working on the next great thing, you get serious about what you do with your time. Not that we weren't serious before, but it definitely gives you a different voice in the back of your head. I feel more diligent and have a sense of ownership and pride. We only have ourselves to blame if we don't deliver the best we possibly can.

I've heard a little bit from different people who know what they're talking about. There was a sense Bungie could never fit properly into Microsoft. Did you ever feel that way? It's not necessarily a negative thing; it was just a reality.

MZ: Well, yeah. I don't know if that statement... it could be as true for any discipline, because Microsoft is a giant umbrella, like I said before, that has all these different quote-unquote "cost centers".

Microsoft is an engineering company, first and foremost, and games are entertainment and content-driven. So that was a new business that they're ramping up. But that doesn't necessarily mean that Microsoft is not capable of having an incredibly successful entertainment segment to their business.

BJ: To their credit, they did a really good job of getting Bungie our own building. Essentially, we were in a bubble. We were mostly independent the entire time, with all of the perks of having a partner. They respected what Bungie had culturally.

When they bought Bungie, they didn't try to assimilate us and break our will and make us all drink the punch. They let the team maintain what made Bungie, Bungie. I think that's one of the smartest things that they did.

You hear that echoed by John Riccitiello now. I don't know if any of you guys caught what he said at DICE, but he basically said that EA's failure had been to try to manage outward to the studios they acquired, and their success had been when the studio's culture flowed inward to them.

Maxis and Black Box were success stories, and Westwood and Origin were failures of the EA way, and their new organizational metaphor is a city-state system. They're not going to tell BioWare what to do. Well, I don't think that they could, but... they could, but you and I both know that's not going to work. And they know too now, hopefully.

MZ: In a way... well, I'm not the right guy to answer the question. I have worked at other studios, and honestly, just as just a guy in the trenches making art, the culture at Bungie... there's a little voice in the back of my mind that's like, "Hey, it's just us now." But the fundamental culture was there when I started with Microsoft, and it didn't change in a significant way.

And I never felt, even when we were on the main games campus before we got our own building, and we were eating at the cafeteria... I didn't feel like they were running our lives. In fact, I worked for a studio where, even though we were off in our own building in our own city, and the next office was halfway across the country, I still felt like there was more heavy-handed publisher-down first-party management than I felt sitting in Redmond.

And who knows? I'm in the trenches just making art, but there could be all sorts of shielding going on in the management levels, but at the end of the day, I can't imagine the relationship was completely antagonistic, because both parties are enjoying the success.

Bungie drove the success of the Xbox platform at a very basic level very early on, and certainly I'm not trying to insinuate there was some sort of fight. It's just interesting to me. I mean, you've got to know that everyone tries to speculate why the hell Microsoft would let Bungie go, and why Bungie would want to go, but I'm not expecting you to answer that.

BJ: We sort of have. Why we wanted to go, I think Mike already touched on when he talked about the team. We really wanted to empower ourselves and control our own destiny and really feel like whatever next big thing we made, that we were all in it together.

Whether we make or break, we own it, and we control it, which I think is important for Bungie to get back to where the company started, culturally, and I think its reinvigorated the team and given us new perspectives.

On the Microsoft side, they still have a great partner in Bungie, and a great relationship. We have a team now that's reinvigorated and also made great games for their platforms. It really was a win-win. It doesn't seem like it really adds up to people, but both groups got what they wanted out of the deal. We're happy that Microsoft allowed this arrangement to happen. It really is a mutually beneficial arrangement.

I think it's just that we're coming off so many years of people being f*cked over by publishers that people find it hard to believe. I could be wrong.

MZ: Sure.

And I don't think Bungie ever was. Maybe it's more natural from your perspective -- that things worked out for you.

MZ: I mean, it definitely felt unprecedented.

I'll take you back to art. One thing I wanted to talk about was what I touched on earlier, with the look and feel of the Halo world. How long were you there? From Halo 1 on?

MZ: Halo 2.

Look how people responded to BioShock. They brought on an art deco aesthetic, which as Ken Levine said in his presentation was familiar culturally, but unusual for games. I think that people underestimate the way aesthetic and world-building can affect players.

MZ: Yeah.

How do you think about that process?

MZ: Well, I think this kind of touches on your independence question. Fundamentally, I think any game developer wants to build things for themselves. As much as there's a target audience out there that I could identify with some market research and tailor a game design to, we're fortunate to work in a profession where you're playing all day, even though it's work. I haven't met a game developer that is punching the clock because they got stuck in a dead-end job. Everybody wants to be there. It's a really unique industry in that respect.

Naturally, you're building spaces that inspire you. I can get my satisfaction out of playing something maybe a lot grittier or realistic, but when it comes to being an artist and sitting down to work on it, every day I'd rather stretch my imagination a little more. I'd rather go on a fantastic journey, to some extent.

All of Bungie's games have had an element of the fantastic, whether it be sci-fi or fantasy or whatever. There's always something bigger, and I think that's something that games can do. They can offer you experiences that are larger than life in a way. It really comes down to that, building things that inspire you.

It seems like it's getting harder and harder to sell people fantasy. I'll quote David Jaffe, who said to me that in the PlayStation 1 era, it was maybe 80/20 for real/fantasy, then with the PS2 it became 90/10, and now it's nearly 100%.

MZ: Yeah. I don't know if there's... I was just in a future of the MMO panel earlier today, and somebody asked the whole "sci-fi versus fantasy -- what's more marketable?" I think the answers on the panel were pretty unanimous, and I agree with them.

You make a quality game that you believe in and that you think is brilliant, and make it the best of class... Star Wars proved a lot of cynics wrong about sci-fi in the '70s, and Lord of the Rings was a smashing success. Do a quality product, and people are going to think there's enough taste out there that you're going to find an audience.

I'd say that consistency and believability of the theme are probably more important than what the theme is.

MZ: Absolutely. I think there's internal consistency. I agree. It comes down to the quality and the integrity of your IP, more than which one you choose. And fashions will come and go.

Chris Butcher

From an engineering perspective, what was your contribution to Halo 3?

Chris Butcher: Coming off of Halo 2, I was responsible for the networking and UI and the multiplayer side of the game -- that kind of online experience, from a technical standpoint. For Halo 3, I was responsible for doing early contributions to the online side of things.

And as a lead engineer, I was just helping out with various random things. The last thing I did was I made the Chinese font display correctly. (laughs) Stuff like that. You do a lot of general-type stuff.

I think that game was a nice leap forward, in terms of its online implementation, and will probably become influential. The Forge stuff opens things up a lot to the players, but in a very accessible way. Was that a challenge to get that working?

CB: It was one of the things that we placed the largest technological investments in, in Halo 3, besides obviously bringing the game to the Xbox 360, which was one of the large technological pillars for us. One of the others was online multiplayer.

We really wanted to take the experience and make it very relevant to the player, and immediate in that sense. One of the overall principles we had when we started, was that everything the player does in this game, they should be able to do with their friends.

Everything is cooperative. Everything from the single player -- you can play through cooperatively and score with it -- you can watch a saved film, you can invite your friends to watch the saved film with you. Even if they don't have it on their box, you can be narrating it and showing them all the cool stuff that's going on, or you can play Forge together and edit maps collaboratively.

We really wanted to make that social experience part of the underlying fabric of the game, because that's what multiplayer is about. It's about connecting with people -- your friends or strangers -- and it's a big part of Halo 3.

One of the things that can be difficult in development is, say, designers working with engineers to realize their ideas. How does that work at Bungie? How are you able to implement thesehighly technical ideas, from a content perspective?

CB: One of the guiding principles that we have is, "The game is the thing." It doesn't matter if something you're doing can be as technically smart or elegant in design as possible. If it's not in the overall service of the game, it doesn't really matter.

I think everybody at Bungie -- from engineers to designers to artists -- understands that it's a collaborative process, and you have to be able to work really well in small groups to enable other peoples' potential.

What that means on the engineering side is that if there's something that's very easy to do technically or very hard to do technically, if it's in the service of the overall design it's the right thing to do. And vice versa. You may have something that you think is technically awesome, but if it doesn't fit well with the design of the user interface and you don't easily communicate to players how it will fit, then it's not really worth doing.

Do you guys sit in multi-disciplined pods, or do you have meetings? How do you hash these things out?

CB: Well, it varies. If you ever come to Bungie, one of the things you'll notice when you walk in the front door is that it's completely open. We have a big floor. Actually, the last thing this building was used for before we moved in was a set of batting cages, so it's just a big, cavernous space. It was a hardware store before then. It's a single big floor with desks all around, and the desks are all on rollers, so you can move them around the space and reconfigure as necessary as people move around and shift jobs and stuff like that.

But we don't sit in multidisciplinary pods, generally speaking, unless it's very early on in the concepting phase of the game. When you're concepting something, you've got five or six people sitting in a room together, just a pressure cooker of ideas. But after that, when it's time to actually produce the game, we organize people by discipline.

The reason for that is that the things that fall through the cracks are most often the kind of informal collaboration that has to happen between engineers, like, "Oh man, I'm having a lot of trouble with this." "Did you try such-and-such?" "Oh yeah, that fixed the problem." That kind of informal collaboration is important, and if you split the disciplines up, that tends to fall through the cracks.

Whereas if you have engineering, design, and art and they're all together, they can share their own best practices, and communication between the functional groups is something you can track and make sure it's happening, because it's something you can quantify. It's very hard to quantify those informal relationships that make you better programmers, because you're all just sitting together and sharing experiences.

Do you rely on producers to act as the go-betweens?

CB: We do have a number of producers at Bungie, and they do help with adding structure to the development process, but mostly, the way we do it is there's a bunch of teams dedicated to specific things.

For example, in the early part of Halo 3, there was a writing team of four or five people who would get together regularly, and then that turned into the mission fiction team, who would be a different group of people whose responsibility is cross-discipline, to make sure each mission reflects its place in the overall fiction and the story arc and the narrative experience. In the same way, there would be a group for characters and a group for streaming and stuff like that. It's a pretty common process.

Yeah, it is. But I think when people look at a wildly successful game, they want to know, from a development perspective, "What we could be doing to emulate that?"

CB: It's all about empowering the individuals you have. We've always focused on hiring great people, and then giving them the tools that they have to do their job, and making sure that everybody is set up in order to make a meaningful contribution to the game.

That doesn't necessarily mean that every artist is going to be designing a new weapon and putting that weapon in the game because they think it's cool -- obviously it has to fit with the overall design vision -- but they should have two things.

One, they should have a clear path to communicating with people who are tuned to the vision of the game and be able to have input that way. The other thing is that they always need to have the tools and the scope of creativity that they need, to do something that they really care about passionately.

Obviously, networking is highly technical, and what you can do within the context of networking is extremely bounded by the limitations of both what kind of connections people have and potentially by the Xbox Live service itself in some ways. How does that push-pull work on deciding those features and implementing them?

CB: It is a challenge. Online is a really multidisciplinary thing, in terms of the engineering technology that's needed to go into it, and the overall vision for what the community is that you want to create.

And also, really, the artistic ability to create an interface that can entertain players as they're using it, but it also feels natural and transparent. It doesn't feel like a burden to reveal the features that are hidden within it.

It's a really, really hard problem. It's something we have quite a lot of experience doing, because we had a chance to do it on both Myth games and on Halo 2 and Halo 3. So we understand, I think, a lot of the hard problems. It doesn't make them any less hard.

When you're coming up with ideas for the networking and things you can do with the multiplayer which push things forward, at what point do you, as a technical person, have to say, "Well, that sounds great, but that is not going to happen?" Is that something that happens a lot, or is it so collaborative that it doesn't happen?

CB: A little bit. I think there are some things that you might want to do. Let's say we want to do 16-player co-op, for example. That's not something we came up with Halo 3, it doesn't really make sense with Halo 3's design.

You have to look at both, "Why are you trying to achieve this in the first place?" and "Is this a core part of what your vision is for the game?" If it is a core part of the vision, then we get to consider it on the same level as all the other core features, so it's something that might require significant investment. But sometimes you do have to say, "We're going to have to find a different way of doing that."

Almost always, if you take a specific request from designers that seems crazy, you can drill back down to them and say, "Well, what we're trying to go for is this subjective feeling in the player, the feeling of epic scope," or so on. It doesn't actually mean we need 500 AI-controlled Warthogs.

Maybe we can just have something that looks like a whole lot of Warthogs off in the distance that makes the player feel like they're part of this big battle and conveys the same emotional response without necessarily being that specific technical implementation.

Halo 3 had ambitious four-player co-op. That must require a lot of testing, and if the levels aren't really basically ship-quality, then you can't test them accurately.

CB: Yeah, Halo 3 is a very difficult game to test, because it has so many features and so many challenges. Like, what if I'm playing Forge with 16 people and the host leaves and then the host migrates twice, and then we save a film of that whole experience? And if I'm watching that film back at a party with three other people... how do you test that? That's like a ten-stage process.

There's just a tremendous ball of all of these features that interact with each other, which is one of the things that's great about the game, but also makes it very difficult to test. The only reason it was really possible for us was... I guess there were two reasons.

One was that we have a really great test team that understands from a very early stage all of the features that go into the game and all the challenges they will be for testing. They're able to help us understand what would be necessary in order to test the game.

The other thing is that we have a pretty strong focus on automated testing of the game. We have stress farms of 400 Xboxes that just run the build overnight. I like to tell people that the E3 announcement build of Halo 3 that was in 2006, we finalized the build about five days before E3.

In those five days, it had two years of testing on it, just because of the number of Xboxes we had running continuously for 100 hours before we showed it. So we were pretty confident in that build before we showed it. It was going to be solid.

When you talk to some developers, they don't quite have the lavish resources that a first-party game has thrown at it. You have room to be ambitious with it, but at the same time, I think there are obviously things to learn from that experience that can filter down in a different way on other titles.

CB: Yeah. I think there's resources and there's resources, right? We're always big believers in investing early in technology in such a way so that it can make us more efficient, so we can do more with the people that we have. We really don't want to grow to a team of 200 people, for example. We're pretty sure that our processes would not scale to that level. So we like to invest early in technology to avoid that early on, to try and avoid that.

Really, these investments of technology can be very cheap, when you look at dollar value cost. The total investment, in terms of man hours, and taking our light mapper and making it not just run on one artist's machine, but then run on a back-end server farm... it's actually probably equivalent to just hiring maybe one or two artists extra. But instead, it was able to take all of our artists and multiply their power some fraction, which helps them to be more efficient, and ultimately helps us make a better game.

That's what it comes down to for us. Efficiency doesn't necessarily mean, "Well, now we get to make more content!" What it means is that now we'll be able to go through more iteration and polish passes on every iteration that we want to do. I would say instead that this investment in technology and in process is ultimately a cost-saver for us.

You were primarily focused on network, and other engineers oversaw other aspects of the game. Do you feel that it's necessary to have people stick to what they know? Is it good to have people who are really dialed-in to one aspect of the game, from an engineering perspective, and then to follow that through?

CB: It's certainly true. You need experts. A lot of what is done on games these days is very technically demanding and requires a lot on one particular sub-aspect of engineering. But on the whole, what we strive for is to have people who can be a little more globally focused.

Those people who understand not just, "I'm going to write this great shader system," but also, "This is the role that it plays in the entire creation of the game. Here's how it interacts with the artists. Here's what my shader system has to know about audio, so the right sounds can happen when objects hit this particular type of shader."

There are all these global kinds of dependencies, and the more that we can train and find people that have that kind of global awareness, the better the game is as a whole, because that way, you can have these cross-disciplinary interactions that just arise naturally for people doing their jobs.