The Wideload Way

Game publishers offer an interesting incentive plan: “Work like a slave for two or three years to ship this game, and as a reward, we’ll fire you.” It’s a morbid joke among artists, animators and quality assurance testers. The bulk of game production nowadays requires a large team of 90 to 150-plus people, but as soon as the game ships – or anyway, after the first couple of patches – there’s seldom a good business reason to keep the team together. Did you break new ground, maybe help create an instant classic? Yeah, great job. Buh-bye!

Sure, if the company is lucky and foresighted enough to have another project well-along in development, at exactly the stage where it needs bodies, individual employees can move over. (That’s assuming they’re a good fit. Maybe the guys who just spent 18 months designing spaceships don’t necessarily want to texture orc armor.)

But for a small studio, with only enough resources to handle one game at a time, “transition” equals “layoffs.” The early stages of a new project – concept art, placeholder code, design docs and blue-sky brainstorming – work best with small teams. So the execs fire everyone else, until it’s time to ramp up once more. If they try to reassemble the same bunch later, guess what? Everyone has scattered to other jobs. It’s just one more way the industry is broken.

Yes, this does affect you, the player. It isn’t a studio that makes the games you love, it’s a studio team. Whether or not the studio stays in business, the practice of “ship and lay off” means the team, that irreplaceable combination of talent, is gone. Whether or not you know it, this has already affected you, and will again.

But there is another way to make games, one not only less disruptive than the current cut-and-run model, but also cheaper, faster and more efficient: the Wideload way.

In-House Design, Outsourced Production
Alexander Seropian co-founded Bungie Software with programmer Jason Jones in 1991, during his senior year as a Math major at the University of Chicago. Their early releases, the complicated multiplayer action fantasy game Minotaur and a spin-off called Pathways into Darkness, were modest successes. With the Marathon and Myth series, Bungie became the leading Macintosh game developer, for what that’s worth.

Actually, it turned out to be worth a lot: After starting Halo, the founders sold out to Microsoft in 2000. Flush with funds, Seropian left Microsoft in 2002, returned to Chicago and soon grew bored. In 2003, with half a dozen other Bungie alumni, he started Wideload Games.

From day one, from the ground up, Wideload was built to implement a new method for producing games. It sprang from a set of commandments Seropian outlined in a talk at the Game Developers Conference in March 2006:

1. Establish your own creative direction.
2. Own your intellectual property.
3. Be no one’s bitch.
4. Keep your overhead low.

Number 4 led to the Wideload approach: Design the game in-house, then outsource the entire production. In a Newcity Chicago interview with Mike Schramm in November 2005, Seropian explained:

“To make a game these days, if you want to do a console game with some story and depth to it, it takes a lot of people. And that’s really where all the expense is.” Most of the idea work (as opposed to the production) is done by “above-the-line” talent. “The big idea with Wideload is that that’s who we are,” says Seropian. “We’re the above-the-line talent. We’re the guys who are coming up with the idea, fleshing it out, getting the technology that’s going to drive it, designing, prototyping, getting the project into production.” And everyone else necessary to make the game is hired contractually. “And that’s a very, very different way of doing a game than has ever really been done before.”

The Wideload method’s strength lies in its low burn rate. “If the team is small, the overhead is low,” Seropian wrote in a Game Developer magazine postmortem. “Time equals money, so low overhead gives you lots more time to experiment and prototype (good for originality). Additionally, every project starts small and ends big. But if you think of each project as a cycle of life, your company goes extinct pretty quickly when you have 75 people wrapping a project and then you only need ten or so to start the next one. Staying small was the key.”

Wideload’s first project was the comedy-horror Xbox game Stubbs the Zombie in Rebel Without a Pulse, released by Aspyr Media for Halloween 2005. Production took 18 months. Wideload’s own 11-person staff handled the game code, level design and writing, but everything else was outsourced. Principal contractors included The Animation Farm in Austin, Texas; audio and post-production by Post Effects in Chicago; a motion capture studio in Chicago’s Hoffman Estates; and an art production house in Bangalore, India. Aspyr created the inventive Stubbs soundtrack, which restyles ’50s pop standards as acidulous modern rock. (The game also features a malevolent barbershop quartet.)

Seropian tells The Escapist, “We contracted out character modeling, environment modeling, motion capture and animation, sound effects, music and voiceover, 2-D art for user interface, and the shell programming. Some of the contractors were individuals that we worked with previously at Bungie. Some were art outsourcing or post-production houses. Most were in the U.S., but two firms were overseas. In total, about 65 people [who were] not [full-time employees] of Wideload worked on Stubbs.”

Does It Work?
Yes. An original and atmospheric game built on the Halo engine, Stubbs the Zombie received good reviews, though many thought it too short. Stubbs has sold well enough to fund Wideload’s next project, as yet unannounced. The company will use the same method for that game.

“When I did the budget analysis prior to beginning production on Stubbs, I projected a 35 percent cost savings compared to staffing up with full-time positions,” Seropian says. “However, we did have a schedule overrun of four months. Because of our low-overhead/outsourced production approach, we were able to hit our budget in spite of the delays. In hindsight, I expect our model saved us 45 percent on the production budget.” He expects to realize similar or greater benefits on the new project.

But the Wideload way has a learning curve. Stubbs the Zombie had no producer, and that caused trouble. They had trouble getting accurate bids from contractors, trouble with underperforming workers, trouble training artists to use the engine. “The big point here is that our model works best when the iteration process is as efficient as possible. That means the cycle of assignment-production- submission-review-revision is really clean and tight. When you start dividing that process among people in different places using different tools, it can get cumbersome. We gained great efficiency once we got everyone using the same communication tools, production tools and previewing tools.

“If you are going to have contributors spread out in different locations, it’s critical to have communication and production tools optimized for non-face-to-face work flow. That means you need to be super-organized and over-communicate. You need tools to preview and share ideas and direction. The outsourced model also requires different skills. For instance, artistic talent alone will not get you good results. Managing contractors takes management and direction skills you can’t afford to be lazy about.”

History Doesn’t Quite Repeat Itself, But It Rhymes
You know who’s reading this and saying, “So what’s new?” Everyone in the movie business.

Hollywood makes every theatrical release the way Wideload makes games. A relatively small production company conceives a film project, gets a studio to greenlight it and then assembles a production team on the instant, a whole special-purpose company with hundreds or even thousands of short-term contractors. The enterprise exists long enough to complete the one project, then disbands.

The film industry enjoys a huge, free-floating population of highly skilled technicians as well as standardized working arrangements and a common culture. The game industry is still in early days in these respects, but it has key advantages over film: Few of a game’s creators need be in the same location, and they are all, pretty much by definition, heavily networked. The Wideload way requires a broad spectrum of web tools. Seropian says, “We rely on IM [instant messaging], source control like Perforce or [Visual] SourceSafe, a visually-oriented game engine with easy previewing capability (i.e. artists can preview their work in-engine before submitting) and also a Wiki for organizing and sharing ideas.

“I think our work flow, process and philosophy can work for both small and large projects. We’ve used it on a multi-million dollar project like Stubbs and also on smaller projects. The biggest hurdle for a developer or publisher to start working this way is simply culture. Groups with a DIY [do-it-yourself] mentality find it difficult to trust and respect outside contributors.”

Now Seropian is trusting contributors from way outside the gaming field. As a partner in the media company Spectrum MediaWorks, Seropian has helped create a multimedia property with the working title X Quest. X Quest is a reality TV show produced by Ron Howard’s Imagine Entertainment airing next year on Fox TV, and X Quest is also an MMOG for next-gen consoles and mobile phones that ties into the TV show. Players will interact with cast members, and events in each medium will affect the other.

How does Seropian find time to run two companies? Does the efficiency of the Wideload way make it possible? “I am working with some really talented people at Wideload and Spectrum both. That makes things doable for me personally on a creative and production level. Yes, our production method at Wideload is efficient, but what makes it possible for me to be involved in multiple projects is the outstanding people I have the good fortune to be teamed with.”

Allen Varney designed the PARANOIA paper-and-dice roleplaying game (2004 edition) and has contributed to computer games from Sony Online, Origin, Interplay and Looking Glass.