Melissa Yeuxdoux responded to my Primitar post. Because she brought up several points and used lovely, wonderful data, I'm going to respond directly.
First of all, thanks for mentioning Marvel Comics, Melissa. That primitar looks like he stepped straight out of the pages of a comic book, with all of the attendant distortion. Little head...check. Long legs...check. Massive, bulging chest and shoulder muscles...check. He's kind of a cool dude, but his shape is just wrong. It's too bad he turned out to be the template.
Now Svetlana Pankratova is an interesting case. She is 6'7", by the way, according to a video interview with her. The popular media seems to have been shrinking her by increments (Is that reflective of our discomfort with really tall women? Is it an attempt to diminish her in some small way?).
Her legs are almost 4'4" long, but that is measured from the top of the hip bone, at about the level of the belly button, rather than from the crotch. That measurement is almost 66% of her total height, while the average woman has a rough hipbone-to-floor proportion of 50% and mine is about 60% (37" from hipbone to floor, 62" tall).
Incidentally, Svetlana's proportions are consistent with Marfan Syndrome, a connective tissue disorder which causes extreme height and very long legs. Note Svetlana's long flat hips, narrow sloping shoulders, and extremely thin ankles. Compare them with this photo of a mother and her two sons. All three have Marfan Syndrome.
Although the only numbers I found for Dji Dieng appear to come from a publicist, those numbers indicate that she has legs that are 4'1/2" long (same measurement, floor to top of hip bone) on a body that is about 5'11 1/2" tall. She is 67.6% leg.
Keep in mind that all of the live (not photoshoot photoshopped) pictures I have found of her show that she is around 8 heads tall, perhaps a little more. With 48 1/2" for her legs and 8 1/2" for her head, that leaves 14 1/2" for the space from her chin to her hip bone. Let's just say that I am extremely skeptical of those measurements.
The library avatars average 63.7% of their height from the floor to the top of the hip bone (I have an Excel sheet if anybody really wants it...I love data).
While there is an inverse correlation between leg length and heart disease and diabetes, we're looking at a difference from short to long legs of just a few inches. There is nothing to indicate that longer and longer legs confer greater and greater benefit. In fact, Marfan Syndrome, which I mentioned above, is associated with high risk of heart problems--aortal tearing, for instance.
My point is that avatars in Second Life are not shaped like idealized humans: they are shaped like disordered humans, varying so much from the human template that it requires explanation. And real world aesthetics isn't it.
A study done in the UK on women's body shape and optimal attractiveness showed that body mass index had the greatest correlation with perceived attractiveness. Waist to hip ratio showed a very slight correlation. No other factors, including the ratio of leg length to torso length, showed any correlation. Study Excerpt
Indicentally, the magic numbers were a BMI of about 19 and a waist to hip ratio of .75 to 1. Numbers lower or higher than that were considered less attractive.
Avatars tend to have apparent BMIs much lower than 19 and the women tend to have waist to hip ratios much more extreme than .75 to 1. A female avatar with the numbers for optimal human attractiveness will look plump and thick-waisted in comparison to those around her.
What seems to me to have occurred is a kind of digital speciation. A new set of aesthetic standards has been put in place for avatars, which diverges remarkably from human standards. Although it is fascinating that this can occur, my concern is that, as the use of Second Life for business and education becomes more widespread, it will become yet another source of distorted body images that young women, in particular, will measure themselves against. Thoughts?