Wednesday, December 05, 2007

Virtual Labor Lost

The failure of a highly anticipated, multiplayer game shows the limits of academic virtual worlds.

By Erica Naone

In-game experiment: A team led by Edward Castronova, an associate professor at Indiana University, designed the virtual world Arden around Shakespeare characters, in the hope of using it for education and social-science experiments. One Arden character is shown above. Credit: Synthetic Worlds Initiative, Indiana University

Multimedia •  View images from the game.

Academics are flocking to use virtual worlds and multiplayer games as ways to research everything from economics to epidemiology, and to turn these environments into educational tools. But one such highly anticipated effort--a multiplayer game about Shakespeare meant to teach people about the world of the bard while serving as a place for social-science experiments--is becoming its own tragedy.

The game, called Arden, the World of Shakespeare, was a project out of Indiana University funded with a $250,000 MacArthur Foundation grant. Its creator, Edward Castronova, an associate professor of telecommunications at the university, wanted to use the world to test economic theories: by manipulating the rules of the game, he hoped to find insights into the way that money works in the real world. Players can enter the game and explore a town called Ilminster, where they encounter characters from Shakespeare, along with many plots and quotations. They can answer trivia questions to improve their characters and play card games with other players. Coming from Castronova, a pioneer in the field, the game was expected by many to show the power of virtual-world-based research.

But Castronova says that there's a problem with the game: "It's no fun." While focusing on including references to the bard, he says, his team ended up sidelining some of the fundamental features of a game. "You need puzzles and monsters," he says, "or people won't want to play ... Since what I really need is a world with lots of players in it for me to run experiments on, I decided I needed a completely different approach."

Castronova has abandoned active development of Arden; he released it last week to the public as is, rather than starting up the experiments he had planned. Part of the problem: it costs a lot to build a new multiplayer game. While his grant was large for the field of humanities, it was a drop in the bucket compared with the roughly $75 million that he says goes into developing something on the scale of the popular game World of Warcraft. "I was talking to people like it was going to be Shakespeare: World of Warcraft, but the money you need for that is so much more," he says. Castronova also says that he was taking on too much by attempting to combine education and research. He believes that his experience should serve as a warning for other academics.

Ian Bogost, a video-game researcher and assistant professor at the Georgia Institute of Technology, agrees. "It's very, very hard to make games in the best of circumstances, and a university is never the best of circumstances," he says. "I have serious doubts about not just the potential for success but even the appropriateness of pursuing development work of this kind in the context of the university." If researchers are going to build games for the purposes of research, Bogost says, he thinks it's important to look at the process realistically, and with a scientific eye. "In most disciplines, it's okay to fess up to what worked and what didn't. In laboratory work, you do this all the time ... If this is really research and not just production, then of course there are going to be these kinds of surprises."