GDC: Creating Spielberg's Boom Blox

According to Louis Castle, VP of Creative Development at EALA, Steven Spielberg was first tickled with the game design itch when another entertainment luminary, Nintendo's Shigeru Miyamoto, introduced the film director to the Wii. There, Spielberg saw the potential for making something fun, family friendly, and accessible. When Castle found out about Spielberg's game design itch, he saw something more: a chance to do something outside the standardized way of developing games at EA. “It occurred to me that there might be a different way to start the whole video game development process,” he said. Boom Blox represents the first of a multi-game deal between Steven Spielberg and Electronic Arts. Besides the uniqueness of being a game designed by a celebrated film director, Boom Blox also represented a departure of the EA development team from the EA Game Development Framework. A process proprietary to EA but similar to systems in place at other large companies, it imposed a highly organized process to game development through phases, gates, and deliverables; in Castle's words, these were simply described as, “What to do, when to do them, and who to talk with.” For Spielberg's bashing and breaking action game, the development team started from the white board... and stuck with it. With a relatively small prototype development team gathered around a white board, the game became more and more fleshed out. “It let us explore the space around the Wii remote and what this game could be,” Castle said. And while encouraging a lot of focused creativity and fast iteration, it also had a number of cons: no documentation, poor version tracking, poorly defined pipeline, and an inability to scale. All of this was very risky and exactly the things that the EA GDF system is designed to curb, but “it worked really well for months and months and months... and then it came back to haunt us.” Castle showed the GDC audience some of the experiments that they came up with to test the possibilities of the Wii remote, many of which will not find a place in the final game. With a small team composed of many skilled contractors, they quickly found the fun gameplay mechanics that would eventually become Boom Blox. They found many things that resulted in great fun, which allowed the team to “edit, edit, and edit and only pick the stuff that worked.” Problems started when they showed what the team came up with to the rest of the company. The reaction was actually too positive and too exciting. There was so much interest from elsewhere in the company not just because the result looked good but because the process used to make the prototype was so fun. “The process that we were using seemed like so much fun that people started identifying it as the way to make games, which it is not,” said Castle. Scaling the production team up to handle both the desire from people to work on the game and the general need for more staff proved problematic. Training people for integration consisted of Castle saying, “Here's this whiteboard, here's this whiteboard, and here were these whiteboards. There was not a scrap of documentation.” The people who the team ended up getting tended to be specialists, as opposed to the generalist who made up much of the prototyping team. “We ended up with really great people, and we ended up needing a whole lot more of them.” With those specialists, they were able to do a lot of work on the visuals and aesthetics of the game. There was more production capacity than before, the quality of the work was better, and there were fewer choke points. Despite the change in the team composition, “Boom Blox really owes a lot of its charm to the accident of needing a lot more people to get things done.” Castle encouraged people that the model his team used in the beginning can work. If you want to go that route, it's very important to stay below the radar for as long as you can. Even more important is to have a communicator dedicated to managing the information that moves around while below the radar. “At least there would be some information coming out and you can keep the team contained for a bit,” he said. He also noted that this works better with big companies; small companies can't really maintain teams like this because everyone tends to know where people and money move to. Castle then presented some of the current version of the game to the audience, showing the various activities and ways in which they are played. “What's really neat about the game is of the 300 activities you could do, with one button you can go into an editor... and change everything.” These resultant game types can also be shared, and new game types are unlocked as the game is played. Despite the success of the divergent development process of Boom Blox, Castle noted that that the team is now adhering right to the letter of the EA GDF. “In the end, nothing could be better than getting back to this process,” Castle said. With the drudgery of toll gates and deliverables comes good things like sanity and predictability. “At the end of the day, we really need these kinds of systems.”