Sponsored By

Making Halo 4: A Story About Triple-A

Whether you love the Halo franchise for its expansive lore and emergent gameplay, or hate it for its sci-fi marines vs. aliens video game-ness, it's difficult not to be intrigued by Halo 4 developer 343 Industries: a group of true believers who didn't even really have a studio when development of the game began.

Kris Graft, Contributor

April 26, 2013

22 Min Read

The video game industry in 2012 was marked by the rise of the individual video game creator. The small teams, the new ubiquitous, accessible platforms, development tools that put a paintbrush in the hand of anyone who is moved to create, and crowdfunding that put money in the pockets of anyone with a good idea and a knack for marketing. New kinds of video games are coming along, attracting new kinds of audiences. The long-talked-about "democratization of game development" sure seems to have finally arrived.

Halo 4 is the antithesis of all of that.

Whereas this new breed of game developer is zipping around on a motorbike or gallivanting about town on a penny-farthing, there are games like Halo 4 that are veritable tanks; tanks assembled by hundreds of people, tanks made with specialized, proprietary tools, made with headset-wearing men in mind, and created with millions and millions of dollars under a giant, monolithic corporate entity. The tank is heavy, has lots of moving parts, and gets to where it wants to go by brute force.

Halo 4 represents high-budget, high-production, totally unapologetic, proper "A-A-A" game development. Whether you love the Halo franchise for its expansive lore and emergent gameplay, or hate it for its sci-fi marines vs. aliens video game-ness, it's difficult not to be intrigued by Halo 4 developer 343 Industries: a group of true believers who didn't even really have a studio when development of the game began.

"I'll just play the first level, and see how it goes"

It's not totally unheard of for a studio to develop a game within, or even take over, a popular universe that it did not create. Treyarch did it with Call of Duty, People Can Fly did it with Gears of War, Ninja Theory did it with Devil May Cry, for instance. 

But Kirkland, Washington-based 343 Industries is a bit different, as it was founded for one sole purpose: to take over Halo after series creator Bungie left former parent Microsoft in 2007. The studio simply would not exist, if it weren't for Halo.

Josh Holmes (@JoshingtonState) is creative director on Halo 4 at 343, but he did have a life in game development before he arrived at Microsoft. He started as a game tester in 1995, working for another game industry giant, Electronic Arts. Holmes moved up the ranks to producer and designer, eventually working on sports games.

"One of the ideas there that took flight was the original NBA Street, which I was lead designer on," recalls Holmes. "We went back into incubation and worked on a couple of concepts. Then I was pulled into kind of a brainstorm think-tank about how to save a dying project that was in trouble. It was a wrestling game at the time, where they lost the WCW license. That rapidly became a game, because I made a mistake with my creative partner of brainstorming the concept of 'urban culture meets fighting and wrestling.' 

"It became Def Jam Vendetta. The marketing team was like, 'That's a brilliant idea!' and we were like, 'No, that's a terrible idea! Don't make that game!' But we had a lot of fun." Holmes actually ended up working on a sequel to Def Jam before leaving in 2004 to co-found Revolution Interactive, which would be purchased by Disney a year later under the studio's final name, Propaganda Games. Shuttered a couple years after Holmes left, it was known best for 2008's Turok reboot and a Pirates of the Caribbean game that never saw the light of day.

He eventually found his way to Microsoft in 2009, when he joined the company's internal Halo team, which at 20 people was relatively small. Holmes' interest in the Halo franchise had already been simmering for years, even during the Def Jam days at EA.

Josh_Holmes.jpg"As a designer, I've always been interested in complex systems interacting with one another, and the emergent gameplay that can come from those interactions," he explains. "I've been a long-time fan of shooters. I grew up playing games mostly on personal computers -- Commodore 64 and the Amiga. I never wanted to be a 'PC guy,' because I was this die-hard Amiga fan. Then I remember seeing Doom for the first time on a friend's computer. I went out and bought a PC the next week [laughs]. I converted immediately. I spent a lot of time doing PC gaming with shooters."

When Halo: Combat Evolved came out for Xbox in 2001, his first encounter was the same as with many who experienced the game for the first time. There was a purity in his early experiences with Halo. Like most everyone else, he experienced it as a player, with no real expectations.

"Halo was that first title that really convinced me that shooters could work on a console," Holmes says. "I remember going and buying Halo, buying an Xbox, and bringing it back to EA where I was working crazy crunch overtime. I plugged it in and thought, 'I'll just play the first level and see how it goes.'

"I ended up staying up all night and finishing the game, because I was just so completely enthralled by the universe, and the sandbox systems that were at work. I was enamored with this idea that you could have a shooter with systems of a sandbox nature working together, where different solutions were possible, and emergent gameplay would come out of that. That completely captivated me.

"I finished the game, and I remember finishing it at nine in the morning or something, and I hadn't slept, and I just started it all over again [laughs], because I wanted to go and have that ride a second time. So ever since I had that experience, it really changed in my mind what a shooter could be, and in a lot of ways, what an immersive game experience could be.

"It influenced a lot of my thoughts as a designer."

"Loyalty is a weird word" 

If you were an avid reader of video game journalism in the 90s and early 2000s, you probably came across some articles written by Frank O'Connor (@franklez). The former Official Xbox Magazine editor was one of the more prolific writers about games during an era that was largely defined by the "hardcore gamer."

He left the world of game journalism and joined Bungie in 2003. He lives and breathes Halo; he doesn't appear as a man obsessed, but his actions say he might be, just a bit. When the franchise went to 343, his loyalty was not with the talented creators of Halo at Bungie, but with Halo itself. Now he's franchise development director of the entire franchise. Speaking to O'Connor, there's a sincerity in his voice about his commitment to this fictional universe.

"I came into Bungie during the beginning of Halo 2, during a time when they were really resetting what Halo 2 was going to be. So I got to see the dark days and the good days, and the culture that they built up over the years.

"But let me be clear -- I came to Bungie because I loved Halo," he says. "So when we spun out, I knew the next thing was going to be incredible. My passion has been for the characters and the scenarios that we'd been creating all those years. That was ultimately where my loyalty lied -- and I think 'loyalty' is a weird word to use in your career. It's a strong aspect of how you should behave, but I don't become emotionally attached to a corporation, for example.

"I just felt like we were on the verge of turning it into something really special; mostly through the work that other people at Bungie had done, but that was the path I wanted to follow."

O'Connor was present during the entirety of the Bungie-343 handover. During the transition, he knew that Halo would soon have another home. He was already talking to people like Kiki Wolfkill, who would be executive producer on Halo 4 at 343, and Bonnie Ross, 343's general manager about the franchise's life beyond Bungie.

Ross, Wolfkill, and the small 343 team at Microsoft were already asking O'Connor, the Halo guru, questions about how they should take care of the franchise. "It wasn't the same as farming out your property to a stranger," says O'Connor. "[Bungie] knew where it was going, and they knew we wanted to do our own thing. Whether or not [343] would do the right thing or not, Bungie didn't know. But knew that was our intention. We cared about it."

"There were a lot of mistakes"

"Look around you," says Kiki Wolfkill (@k_wolfkill), explaining the challenge the core team was taking on. "You don't have a team, at all. You don't have a triple-A studio. You have 12 people. You have 13 computers, and no office space." 

By now, Wolfkill can probably be considered a Microsoft video game stalwart. She joined Microsoft Game Studios in 1998, serving as art director. After about 10 years, and shipping a couple dozen games, she embarked on the 343 venture.

"On top of having no studio, we had to solve that problem in tandem with dealing with the other challenge," she adds. "It just didn't look possible on paper."

That "other challenge" is the most salient one: making another entry in the Halo franchise that would live up to expectations that first took root more than a decade ago.

The way Halo 4 was made was unnatural, says Wolfkill. 343 started at around a dozen people, ballooning to about 200. With contractors, the number of people who put their hands on the game amounted to 350. That growth, and all of the problems that came with it, took place while developing the game.


Kiki Wolfkill

"There were a lot of mistakes we made along the way in which we knew weren't necessarily the right way to do things," says Wolfkill of the steep learning curve. "But given what we had to deliver and our timeframe, we accepted that these are necessary mistakes, and we acknowledged and cataloged them.

"We started off with a number of people who had a ton of industry experience, and thought, 'We're going to do everything right! We've seen all these mistakes in the past, we'll avoid those, because we're smarter and we have the experience!' The reality is, circumstance forces you down a path, and your ambitions collide, and there will always be catch-up that you're doing. That's where most mistakes are made.

"There were production realities that made us build things inefficiently," she continues. "I think there was also the learning curve of understanding how to work together."

There was inefficient prototyping -- the team didn't clearly define and communicate the parameters of successful prototypes early enough in development, which slowed the process. The team also started to realize that Halo 4's narrative, rooted in volumes of sci-fi lore, was at times too inside baseball, and wasn't self-contained enough. It was an accessibility issue that needed to be addressed.

Sub-teams would get too close to a singular component of a game, such as a new enemy design, and not think of the design in the larger context of the game’s mechanics, lore and narrative, leading to inefficiencies in the overall development process.

343 also struggled with balancing familiarity with reinvention, as the studio wanted to please a large fanbase, but at the same time bring something new to the series. While the game received high scores, some critics pointed out a feeling of sameness.

Speaking to Holmes, O'Connor, and Wolfkill, there's a common theme, or tone in their voices, that recurred over and over again. For all the opportunity and potential they saw in this project, there was some mind-numbing dread of screwing up. Not just screwing up the game, but screwing up your team, your studio, your career, the franchise itself.

Wolfkill laughed, correcting me, saying it wasn't exactly "mind-numbing dread," but merely "mind-numbing fear." Luckily for 343, the studio happened to be working on one of the most recognizable brands in video games, and was backed by the substantial resources of Microsoft.

"We hired people who hated Halo"

"Having the Halo franchise was burdensome in a lot of ways -- meeting expectations, for example -- but it was great for hiring," O'Connor admits. Many of 343's problems were big, practical, logistical conundrums having to do with growth and recruitment. The studio needed to attract top triple-A talent -- talent that was in high-demand, and probably already employed at other triple-A studios. All of 343's staff came from triple-A; the studio's staff now represents over 25 triple-A studios. 

343 actually couldn't tell interviewees that the studio was specifically working on Halo 4, just that the studio was working on something involving Halo.

"We had people who we hired who hated Halo because of 'X,'" says O'Connor. "But what that really meant was, 'I feel like this game could be awesome because of 'Y input' that I'm going to bring into it. I want to prove it, and I'm passionate about proving it.' So we ended up with a bunch of people who were genuinely passionate about the product. That is a huge advantage, and that helped in hiring and forming our team."

The growing pains threaded throughout the development of Halo 4, as the studio came to terms with firing up an motor while trying to build up the rest of the car around it. For Holmes, the growing pains were familiar, and ones that he encountered when he co-founded Propaganda.

"As a leadership team, we'd go from being able to have everyone sit in an area or a room and organically talk about the experience we're building, because we were small enough to do that," says Holmes. "When you've got multiple missions, five missions in flight, and all of those teams are trying to rapidly turn things around, there's a point at which all of the feedback and interaction starts to bottleneck, and you're not able to move quickly enough."

In February 2012, just nine months before Halo 4's ship date, the studio had to address this bottleneck that was brought on by the rapid growth. Project directors found themselves handling too many line-level decisions, which was causing "inefficiency and frustration" within the team, says Holmes.

"To address this, we introduced a new production process and restructured the team around feature teams, which focused on creation of vertical game experiences, and foundational teams, which focused on game elements and experiences that support multiple features or vertical experiences.

"For example, the campaign was a feature team and the audio was a foundational team. These teams worked toward monthly goals as established by the project directors, but were empowered to make day-to-day decisions and adapt production processes to suit their individual team needs. The project directors checked in with the teams on a weekly basis and provided daily feedback on builds, but we tried to drive as much decision-making as possible down to the teams. This gave the feature and foundational teams a high degree of autonomy in pursuing their project goals, which was important in allowing our large team to remain agile, preventing the directors from becoming a bottleneck to decisions on the floor."

"E3 was sort of a validation of all the things we wanted to do"

Finding the appropriate talent for Halo 4 was one thing, but getting everyone on the same page creatively and process-wise was a different challenge. With so many people from an array of different backgrounds, there were communication issues and cultural incongruities that arose that needed to be fixed.

"There are a lot of technical mistakes that you literally fix. There are cultural things too, but you can't fix your culture, you have to evolve it in a healthy fashion," says O'Connor. "I think we got there through this crucible process. We forged a culture, and a real one.

"We said, 'What are you like? Bring whatever you have to the studio. Bring the reasons we hired you into our culture, and form it naturally out of your persona, your contributions, and the atmosphere you bring into the studio. I think that's the healthy way to build a culture from scratch.

"Maybe 10 years from now we'll be a little less malleable. But I hope we remember that process as we grow as a studio, and remember that we are successful because of what new people bring to it, and not in spite of it."

One of the most interesting moments of making video games is when a developer or a team comes across those moments of epiphany when they realize that they are going in the right direction. Creating anything for an audience is in many respects a blind endeavor until you get to a certain point in which you convince yourself that you have something special on your hands.

These epiphany moments are uniquely interesting to hear about in triple-A projects that involve so many moving parts; sub-machines that are working in parallel with one another, which at some point need to come together in a Voltron-esque fashion and inter-operate seamlessly.


"It's during that time you're questioning yourself: 'How is this going to work, will it be as I envision it in my head?" says Holmes. For Halo 4, he says there were a few epiphany moments that helped boost the morale of the team. One of the earlier ones that Holmes recalls was when the team completed a small piece of the Halo experience that he described as a "very traditional" Halo. User research showed that people thought it was a lot of fun, and it showed that the team was capable of making a Halo game that was true to what the series was about.

343 scrapped it, Holmes says, as it was too traditional. But that first build showed the new team that this amalgamation of different studio cultures could work together and achieve a common goal.

A year later, 343 had finished the first mission for Halo 4. Later, the team injected the new enemy, the Prometheans, into the encounter space, displaying their new AI behaviors, representing a culmination of a year-and-a-half of iteration. Even later, online multiplayer was in place -- members of the team could play the game against each other from their homes. It was another morale booster, a tangible signifier that things were coming together.

But when I asked Holmes, Wolfkill and O'Connor what the moment was, when that enormous, collective sigh of relief occurred, all of them had the same answer. I had interviewed Wolfkill and O'Connor in person, together. When they replied, they looked at each other and replied simultaneously, "E3."

"At E3, at a purely production level, was the first time we were able to express our intent with what the experience was going to be, and have an entire segment of the team delivering on the experience as opposed to delivering on their pieces of the puzzle," says Wolfkill.

"On an emotional level, it was the first time we showed anything publicly, so it was sort of a validation of all the things we wanted to do, and the team really needed that at that point in the cycle."

At E3, Halo 4 was at last in the hands and in front of the eyes of a large audience, in front of press and in front of fans. Reaction was positive, and Bungie staff at E3 were able to relay the feedback back to the team in Kirkland.

Holmes concurred that E3 was the prime epiphany moment. "You have an idea for this experience, but you're not exactly sure how people are going to react to it. Yes, you're doing user research testing, you're having groups of people come in and play, you're analyzing the results, but those are really just small slices of people at different moments in time. ...We'd worked the last two-and-a-half years on this game, wondering how they would react. We think they'd like it. We hoped so. But we weren't 100 percent sure."

"It's all very personal"

November 6, 2012 was the release date for Halo 4. Holmes and the crew at 343 were glued to their computers, awaiting the trickle, then eventual flood of feedback from the press and from players. He escaped into the bubble of the world wide web, absorbed.

"My wife has learned that I'm just not in a state of communication with the rest of the universe at that point," laughs Holmes. "I'm just obsessively reading stuff to see what the reactions are from fans and critics alike."

What Holmes and the rest of 343 saw from the press was generally high praise, which earned the game a Metacritic score of 87. Scores ranged most from perfect ("Trust me, you want this." - Joystiq) to excellent ("It holds the series' standard high." - GameSpot), with a few middling ("I can't escape the feeling that Halo needs to try a bit harder." - EGM) and one quite bad ("A shiny old dog without any new tricks." - Tom Chick). That's on top of all of the varied fan chatter that was happening in comment sections and forums.

"It's all very personal, whether you're getting great feedback and seeing how people are loving the game, or seeing the criticisms about the game," Holmes says. "It's something in which you really pour your heart and your soul into. So you care very deeply about what that feedback is. ... As a general rule, we try not to overreact to the loud, vocal minority."

An obsessed 343 watched everything unravel. The game generated $220 million in global sales on day one (higher than Halo Reach's $200 million launch day), and by all accounts, that provided a pretty healthy amount of validation for the team.


"I wouldn't want to do it again"

Now months have passed. The game is still a work in process, with its heavy concentration on online components and new digitally-distributed content. 343 has started work on a new project with one important luxury: having an actual studio. 

"We did some clever things, we made some mistakes, and we learned really rapidly from those mistakes, and tried not to repeat them," says Wolfkill. "And we sort of pulled it off. We pulled off both things. We created a studio with a natural, organic culture, which is a worthy source of pride to us, and the studio created a worthy source of pride with the game itself.

"We did it -- I wouldn't want to do it again, that's the honest truth, but now we have that team in place, so we don't have to."

Read more about:


About the Author(s)

Kris Graft


Kris Graft is publisher at Game Developer.

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

You May Also Like