Poor neglected blog. I shall return to you now and provide a bit of an update. The worlds of steampunk, warcraft, and research are proceeding apace, with a major milestone reached: the big ole presentation on our results to funders, other teams, and various other folks. We SCRIBE folks showed our first big set of findings from our research.
It was rather an awe-inspiring experience to see the research being done by the other teams, including some really fun stuff from Nic and Nick over at PARC like this: did you know that about 33% of men regularly use female avatars in WoW – and they use those female ‘toons about for 30% of their gameplay? BUT! Women only play male characters about 8% of the time. What that means when you work out all the numbers is that over half the female avatars you see are actually played by men. (Confusing, I know. It’s because only 25% of WoW players are women… assuming their sample of 1,000 is representative of all 12.5 million WoW players. It’s probably pretty close, if not a bit more female than the total pool).
In our (much smaller) sample in Second Life, we had to work hard to recruit men (women are a bit under half of SL players, and they are more likely to participate in studies like ours). Among our 215 or so players, only 8% used an avatar with a different gender than what they reported in the survey about the offline gender – and most of those were men playing female avis.
What these studies show is that a) there are girls on the internet; b) girls do play WoW; and c) yeah, that hot blood elf may well actually be a guy. Oh, and that SL and WoW have really, really different cultures in terms of avatar use. For example, most WoW players have more than 1 ‘toon they play regularly, but among our participants, only 45% did.
Another amazingly cool result was from folks on the team headed by the inimitable Dmitri Williams at Annenberg USC: they used their amazing behind-the-scenes data from Everquest 2 to create maps of who trades with whom – and found a way to identify gold farmers.
Another one that really caught my eye was a project that set up experimental economics studies in Metaverse (which is a lot like SL) to take a look at people’s risk-taking behavior using things like game theory experiments (that’s game theory, not theories about games).
For our part, we talked about the kit and kaboodle, far too much to go into here, but a few fun facts (that were statistically interesting):
- Gender roles in conversation and appearance are alive and well in SL: Women use 50% more exclamation points, about 40% fewer periods and commas, and 63% more apologies than men. They also used over three times as many costume items from our quest, and were 50% more likely to have a gender-idealized avatar.
- More educated players really do write better: they use 43% more punctuation and 60% fewer emoticons than less educated players. Not that I am advocating against the use of emoticons :)
- Older players (over 45) look better: they use avatars that are about 6% taller and are 50% more likely to be gender-idealized compared to younger players (18-25). Plus, older players use 33% fewer emotional phrases like “hurrah!” and correct their misspelled text 40% less often than the youngest group.
- Leaders online kind of look like leaders off line: they have taller avatars, wore over twice as many of our costume items, and used about twice as many directives (like “go do that”) as non-leaders. That’s a lot like research that says taller, better dressed, and more commanding people tend to be leaders, even in groups of strangers. People who said they were high in leadership characteristics and behavior were more likely to be men, but those who were actually voted leader of the group (afterwards we asked them to identify the leader in their session) weren’t more likely to be male or female.
And just in case you missed it, you really have to see that Second Life quest trailer video that we made for our game:
(ps promising to start contributing here again. promise!)