August 11, 2005
Thursday

For roughly a decade, people have used role-playing online games to conduct parallel lives. Raise another family. Start a new business. Build your own city. It's all possible in these virtual worlds.

Now, some economists and social scientists say these Internet worlds could be a new type of laboratory to study economic behavior, such as how consumers respond to inflation.

"I think there's an incredible opportunity here to run controlled experiments on economic questions," said Edward Castronova, an economist at Indiana University at Bloomington.

Castronova co-founded Terra Nova, a group blog that explores the technological, commercial and social dimensions of these virtual worlds.

Dimitri Williams, a communications professor at the University of Illinois at Urbana-Champaign and fellow Terra Nova contributor, said scientists could subtly alter the software that governs these worlds, tweaking the rules of the games, then measuring how these changes affect behavior.

One of the most sophisticated of these artificial worlds is Second Life. Based in San Francisco, it is not so much a game as it is a cross between an Internet chat room and an unscripted movie set.

The real people who play Second Life create proxy characters, called avatars, that can mimic human behaviors like flirting, gossiping or showing off. Some players, through their avatars, even make imaginary goods that they sell - sometimes for real money.

Virtual worlds like Second Life, for it is only one of hundreds of such realms, began in the 1990s as places for Internet users to act out dungeon-and-dragon fantasies with other people.

"This is a social science petri dish," Williams said. "You can see everything and they don't know you're watching"

Online games emerged in the late 1980s and early 1990s as the explosion of computer networks made it possible for many, widely dispersed individuals to take part.

This genre of games is called MMOGs, or massively multiplayer online games. More than 100 MMOGs exist today. The number of players has been estimated at between 2 million and 10 million.

The best known of these fantasy worlds include Lineage, World of Warcraft and EverQuest. Second Life is smaller and newer, more lifelike and less game-like. But as is the case with these other online realms, Second Life has its own currency, the Linden, which participants can use to buy and sell goods.

It was this trade in virtual goods that caught the attention of economists, who were fascinated to discover that much of the activity in these online worlds consists of commerce, not fighting.

What's more, players are often willing to spend real dollars to acquire the totems that confer status, power or bragging rights. In the second quarter of 2005, for instance, Second Lifers exchanged goods and services worth about $4.2 million in real money, according to Linden Lab, the San Francisco company that created this virtual world.

MMOG players spent an estimated $880 million for online goods and services in 2004, said Steve Salyer, president of IGE, a Los Angeles firm whose 250 employees help players buy, sell and trade across the real and imaginary worlds.

"This is commerce in places that don't really exist, over items that don't really exist, in tangible currency," Salyer said.

It was Castronova, the Indiana University professor, who focused scientific attention on the economics of online worlds. In 2001, he wrote a scholarly article about the commercial life of Norrath, the online home world of the popular game EverQuest.

Now, as the economic activity of these games continues to grow, Castronova wants to do more than simply observe. He wants to tinker with the economic rules of the game in a way that would allow him to draw cause-and-effect relationships between changes in rules and changes in behavior. This is possible, he said, because of the way these virtual realms are embodied in server computers.

Castronova explained that with most current games each server can accommodate only a finite number of players. As more players join the virtual world, the game host adds new servers and starts filling them with players. The result is that a game with 400,000 players might actually consist of 40 servers, each with 10,000 players - 40 parallel worlds, each with the same economic rules.

What Castronova would like to do, but so far hasn't accomplished, is gain access to the software that sits on those servers, changing some while leaving others alone. Assuming a random distribution of players to the various servers, differences in economic outcomes and behaviors should be traceable back to the changes, he said.

His notion has elicited cautious interest but no ringing endorsements.

Harvard University's Alvin Roth is noted for lab-based economic experiments. And Roth says researchers must go to great pains to avoid prejudicing the results of these experiments by the selection of subjects or design of the tests. He calls Castronova's proposal attractive because players come to online games willingly and because the underlying economics could be changed in invisible ways and therefore less likely to alter conscious behavior.

"But you have to worry about the fact that these are games, and not everybody plays them," Roth said. In short, are the economic behaviors of people who play online game the same as people who watch television or engage in other diversions?

Where all this is heading, no one can say. But insofar as people have always learned by observing the behavior of others, some say, we may now have a social feedback system that moves faster than the speed of history.

SitNews
Stories In The News
Ketchikan, Alaska