Reuters and the Flawed Research Value of Virtual Worlds

Have a puzzling economic problem or a social program you'd like to beta test before dropping it on the American people? Reuters recommends you turn to virtual worlds as a possible test-bed for real-world policies.

In a recent article, Reuters jumped on Edward Castronova's well-trod "Metaverse as a laboratory" bandwagon, even calling up the respected Indiana University virtual worlds researcher for his thoughts on the utility of virtual worlds:

"We can do controlled experiments in virtual worlds, but we can't do that in reality," said Castronova.

"Controlled experimentation is the very best way to learn about
cause and effect. We are on the verge of developing that capacity for
human society as a whole."


To find out why Reuters is out to make Castronova skeptic Dusan Writer squirm, read on below the fold.

Wealth of Virtual Nations

Reuters points out the utility of virtual worlds due to the captive population therein, even citing some interesting statistics that ultimately contradict its case.

By Reuters' own research, virtual gamers definitely trend male (80% in the virtual world compared to 50% in the general population), and are slightly younger (31 years old, on average) than the population as a whole. They also tend to be more educated.

What does this mean? Virtual societies are not accurate representations of societies present in the real world. Any information gleaned from social experiments in the virtual world, whether they involve purchasing patterns or social interactions and norms, will be skewed to disproportionately represent a young, well-educated male perspective.

What's more, American virtual worlds players are, on average, wealthier than average Americans:

"Everquest II" players are also wealthier than the general population
with an average mean household income of $84,000 versus $57,000 for the
general population.

This shouldn't be surprising. Any user able to spend $50 on game software and a minimum of $14 per month certainly isn't hurting for basic needs. The point is, virtual worlds are a very privileged subset of the population as a whole. Cataloging their responses to economic or social questions may be interesting, but it is of limited use in the broader world in the near term.

Misguided Science

What makes the Reuters article so flawed is its primary focus on the economies of virtual worlds. As we've seen with Second Life and World of Warcraft, and as is likely true with Everquest, Aion, and any other online game with a variable reward system – over time, assets concentrate among the top few players. 

Why? As Wagner James Au wrote in his Second Life exposé, a small minority of players account for the majority of time spent in-world and Linden currency bought and sold. Despite having 600,000 active accounts, Second Life rarely approaches a server-compromising number of users online. For every user who spends 40+ hours per week in any virtual world, there are dozens who log on for 2 or fewer hours per week.

These players will likely not be called upon to participate in a study merely because they won't be online to take notice. Is this the kind of selective and flawed science pro-Metaverse advocates want associated with virtual worlds?

Some steps should be taken by leading virtual worlds researchers to standardize sampling for virtual communities. These standards should take into account the differences between virtual and real communities on every variable from gender to income to race in order to provide a sample response that is relevant to the real-world community.

Anything else is just indulging in fantasy.

3 thoughts on “Reuters and the Flawed Research Value of Virtual Worlds”

  1. Reuters is biased towards Second Life, after having failed at their attempts at it. They hired a development firm who charged a fortune for their island build, they sent in one employee to cover virtual news and events, and when Reuters pulled the plug, the reporter became jaded, and Reuters blamed Second Life for their failure.
    Personally, I would not put any trust or faith into any Reuters articles relating to Second Life or virtual worlds. They are skewed, biased, opinionated, counter-productive, and have an overall slant towards a narrow-minded perspective.

  2. Nice breakdown of the issue.
    Behind all of this is that most of these articles (even the Reuters ones which aim to promote SL) continue to use the same models and measures for Second Life and “virtual worlds” as they do for games.
    A more appropriate comparison would be with social networks.
    Seriously old school, get with the program here: There aren’t games. Duh.

Comments are closed.