I think this analysis is way off. Given two girls with similar average attractiveness, the one with the higher stdev (wider distribution, whatever) is going to have more "5" votes than the girl with a low-variance average. This is probably completely driving the trend. Guys are most-likely to message you if they think you're a 5. So, given girls with equal average attractiveness, the one with the wider distribution will have more 5's and get more messages.
EDIT: After looking more-closely, their formula basically agrees with what I've said: the number of 5's is weighted most-strongly (0.9 versus the second-strongest of 0.5). Also their R^2 value is pathetic.
We now have mathematical evidence that minimizing your "flaws" is the opposite of what you should do. If you're a little chubby, play it up. If you have a big nose, play it up. If you have a weird snaggletooth, play it up: statistically, the guys who don't like it can only help you, and the ones who do like it will be all the more excited.
This should be followed by an asterisk:
*As long as it doesn't bring your average rating down.
There's a big difference between playing up polarizing features and drawing attention to flaws in general.
The thing Christian Rudder fails to point out explicitly is how man can put this information into action. Knowing that women with tight distribution get messaged less men should start messaging cute girls with tight distributions to improve their chances.
Why should he have to point out anything at all about how men can put it into action? Presumably close to half the people who use OKC are women looking to attract men•, so the message to them ought to be fine.
Actually, as a male who uses OKC, I found it interesting to simply note "the curse of being cute." My takeaway is that there are some "conventionally cute" women who you find attractive, and who you also assume receive a lot of attention. It turns out, many others assume the same, and so end up not sending a message.
Putting this into action, then, means to take more initiative messaging someone who you assume must be receiving a lot of attention.
EDIT: After looking more-closely, their formula basically agrees with what I've said: the number of 5's is weighted most-strongly (0.9 versus the second-strongest of 0.5). Also their R^2 value is pathetic.