A note on sex difference in variation in attractiveness

Given the recent discussion of sex differences in variation in traits, The Audacious Epigone has a quick post based on the GSS (you guessed it). He finds:

The weight finding is probably related to female self-starvation for beauty purposes (anorexia) which creates a stronger left tail for women but not for men. The physical attractiveness is surprising, but I have a guess why. Some years ago, OKCupid famously published some internal data about how sexes rate each other in terms of attractiveness (0-5 scale apparently, though I recall it as 1-5 on the site) and how this is related to contact behavior. It looks like this:

I.e. men rate female attractiveness on a normal curve (close approximation) with a modal/mean/median of close to 2.5, while women's ratings of men are extremely skewed with a modal rating of 1! OKCupid apparently later deleted the page, but the internet remembers. Anyway, you can now see why male ratings might have less variation -- they are all skewed towards the left and the floor effect reduces the variance somewhat. It would be better to obtain machine learning based ratings of men's attractiveness, and then see if these finds a larger male variation.