After a particularly difficult project in 2004, the RTS team at EA Los Angeles tried something new with Battle for Middle-earth II. "We created 15 teams instead of one."
Each of those teams was assigned a mini-project. Verdu said the only way this works is for each sub-team to have leaders with unique identities and a degree of empowerment. "They are the equivalent of small design teams.
"We had 15 projects flying in formation, instead of one monolithic organization." Verdu describes it as "all of the magic of small teams" but working on one big project. Even if it was the user interface team, it had a designer, it had an engineer and it had some artists. "They really were their own little project."
"The creative work that came out of those teams was amazing," Verdu said. There's a delicate balance to maintain when guiding the teams and giving them the freedom to be creative.
The team leads would meet daily, with the goal of keeping those meetings as short as possible. They found that good communication would "keep the planes in formation." And when the senior development director took a leave of absence for a family illness, the de-centralized teams were able to keep going.
Verdu began to notice something that looked like flocking behavior, with an instantaneous change in direction when one team broadcasted a discovery to the other teams. "Everybody turned on a dime. It was organic and almost hard to understand. ... It's like we tapped into this thing we didn't entirely grok."
Frequent integration was also key. "It almost felt kind of like open source." On Battle for Middle-earth II, the results were a better game that took half the time as the first game in the series, without severe crunch time. "People had fun with the project. They felt like it was their product."
And it was very well received. "I considered that a big success." Verdu's next project was Command & Conquer 3, and he used the same small-team method with an even larger number of developers. Again, the developers shipped an RTS game that everyone was proud of - "passionate, in fact" - and it was a critical success.
"Organization design with these big teams and this open-source method of design is really part of game design," Verdu said, who believes that the days of the giant teams are over, even for EA. "Our teams are ramping down in terms of size, and we're giving them more time.
"Let's take a smaller team, work for longer. ... It's a lot more sane. I probably don't understand it yet, but it has just amazing power."
"Organization design is game design," Verdu said. "I've really become a believer in this. From now on, when we put a project together, we're going to be thinking a lot about how people work together, because it's kind of meta design - how you design your organization will drive the design of the game in a profound way."
N. Evan Van Zelfden expects great things for the future of games. Games are the greatest art form to date, he asserts. This is why he plays games, writes about them, and continues to work in the industry of games.
