links for 2009-11-24

  • OK Cupid, the hetero dating site, keeps crunching numbers about their members and blogging them. It's pure stats-porn for numbers geeks like me. In this installment, they point out that men's ratings of women collectively form a normal distribution, while women rate 80% of the men on the site as below average. However, the distribution of e-mails to men from the women basically follows that distribution curve—that is, about 80% of the e-mails go to the below-average men. On the other hand, two thirds of e-mail sent from men to women go to the top one-third of women, appearance-wise. Women are pessimistic about men, but realistic in their aspirations. Men are realistic about women, but optimistic in their aspirations.
  • "Bohemian Rhapsody," as performed by The Muppets. There are two kinds of people in the world: the kind that need no further information before they decide to follow this link, and the kind that are thinking "ehh…I don't know…"
  • Apologies to my friends who actually work for ONE/DATA, but this is just plain funny.

7 thoughts on “links for 2009-11-24”

  1. Huh, that’s interesting—I thought, like most dating sites, they only matched up people of the opposite sex. Well, smart business move.

    And you’re right, IP—I used the wrong word there. I worked my little brain into a pretzel trying to sort out that simple punnett square kinda thing. :)

  2. Where does your information on “most dating sites” come from, Waldo? I thought eHarmony was unusual among dating sites in *not* handling homosexual relationships. Certainly does.

  3. Waldo: OKCupid is also a bit different from the others in that it uses a fairly unique matching algorithm. Users answer multiple-choice questions, indicate how their “ideal match” might answer them (multiple selection allowed), and select how important it is that their match answer this way. Then statistical analysis is applied on the intersection of the answered question for each pair of users. I haven’t used any other dating sites, but I think that they are mostly based on keyword searches.

    I was initially drawn to the site due to their statistical approach, and I’m glad to see them *finally* performing some fun analysis on the user data they’ve accumulated.

Comments are closed.