A Study of Gender Presentation in Resident Profiles in Second Life Public Spaces
When I returned to SL a couple months ago, I noticed that there seemed to be a lower percentage of apparently male avatars than I remembered from 2010-2015. I wondered if my chosen destinations were female-leaning, so I began jumping to areas that were not personally interesting. The imbalance was there too. Having an anthropology background, it was time to get more methodological.
In counting 1068 avatars across 58 destinations in SL, I observed 50% female avatars, 27% male avatars, and 23% avatars that were not binary or were of unknown gender. In fact, the majority (86%) of venues had male avatar percentages of 0-38%. 7 locations were between 50-58% male, and one outlier (Star Wars Legends roleplay) was 90% male when I visited. The quantity of male avatars was significant at some places — there were 18 male avatars at a ballroom I visited — but they still only constituted 31% of the crowd. Of binary, clearly gendered avatars I observed, 35% were male and 65% were female.
I visited 58 destinations: 9 General, 30 Moderate, and 19 Adult.
I tried to minimize my bias by choosing locations to count in a variety of ways. Venues directed specifically at one gender, such as single-sex clothing stores or gay/lesbian clubs, were not visited. Locations were found by selecting from the SL website destination guide, links in group notices, event and destination listings, keyword searches (airport, beach, Brazil, chat, city, combat, Deutsch, family, furry, sandbox, and Turkey), looking for avatar clusters on the map, and suggestions from friends. I sought places with varied activities and audiences, most of which I had never visited before. A region had to have a minimum of four avatars present to be included.
Because of my availability, most places were visited between 4:00-9:00 PM SLT (Standard Linden Time, which is US Pacific time). However, I did get to 11 venues between 8:00 AM and 1:30 PM.
The sex of an avatar was determined by resident profile presentation (more on this in the Discussion section below). To gather profiles for this study, I visited an area, popped open the Nearby list, and opened the profiles of everyone there. I then sorted the profiles on my screen into three piles — obviously male, obviously female, and undetermined. Profiles in the undetermined pile would get more scrutiny and then I would log a final count. I did not consider avatars that arrived or left after my sorting began.
As I explored, I noticed an interesting phenomenon. I’d rez into a location, look around, and think, “Wow, there are a lot of men here!” Then I’d count and find that female avatars outnumbered male 2:1. Since I tend to seek the companionship of men, my perception was off. On the other hand, I spoke with a number of male residents who didn’t notice that the balance was so tilted either.
Determining the sex of an avatar was somewhat complex. In the era of mesh bodies and heads, binary gender presentation has become more extreme, which meant that the resident’s main profile photo was my first sorting clue. Most of the female human avatars I counted were blatantly gendered: full breasts and hips, narrow waists, makeup, long hair. Many of the male human avatars had facial hair and broad muscular shoulders.
Display names were only considered if they had a gendered title that cleared up other ambiguity. For example, it was not uncommon to see profile photos that featured a couple. That image plus “Mrs. Jane Doe” meant that I’d count the avatar as female. I often combed through text to find gendered self-references, such as “I’m a sweet country gal”. Sometimes a partner provided a useful clue. No assumptions were made around sexual orientation, but I’d visit the partner’s profile to see if there were gendered references to the avatar I was counting. “I love X, he’s my knight in mesh armor!” and that sort of thing.
I encountered several residents who described themselves as shape-shifting, androgynous, non-binary, or explicitly said that they may sometimes appear male and sometimes female. I’m not familiar with the furry subculture and was relieved to see that many gave clear indications of gender in their text sections. However, most of the avatars I counted as neither/undetermined had nearly empty profiles and in fact, I think most were bots or NPCs in roleplaying areas. An avatar named GangMember but with a blank profile was logged as undetermined. In a longer study I would try to parse out the active residents from the rest, but it is fair to say that the number of avatars listed as neither male nor female is strongly inflated by bots.
That said, my categorization of an avatar as male, female, or neither/undetermined was subjective and I’d consider other methods if this was research with a goal of publication. I used profiles rather than visual inspection because of the ease of changing avatars for effect, a joke, an event, a mood shift, or a particular venue. Relying on interviews would have made the effort much greater and would have introduced another type of selection bias.
Counting at Prehistorica in my pteranodon avi
Destinations where I found the highest percentage of male avatars were:
- Star Wars Legends (roleplay, M)
- Yiff (furry, A)
- Santa Ramona Valley (roleplay, A)
- Maui Swingers Resort (sex, A)
- Prehistorica (dinosaurs/infohub, M)
- 2Raw Extreme (racing, M)
- The Looking Glass (hangout, M)
I can only speculate why the numbers are skewed like this. It may be that nowadays, SL has more activities that appeal to people who present as female online. My husband’s suggestion that it’s easier to have a nice-looking female avatar may be a factor. There may be plenty of male avatars in SL but they prefer same-sex venues to the open areas I explored. People who had male avatars may have moved off to other virtual worlds, games, or virtual reality. People may have selected female avatars to try to ply their wares in the digital sex trade (though I see an absence of clients across the grid). It could also be that the avatar gender balance has always been disproportionate, but it wasn’t apparent enough to raise my curiosity.