How do men rate women on dating websites? (Part 2)


I always recommend looking at data in several different ways, to get a more complete picture of what's really going on - such is the case with the member 'ratings' on dating websites. Let's take a look at some data from a different angle...


In a recent blog post, I analyzed which age men & women the opposite sex rated most attractive. The graphs indicated that men rated 20-year-old women the most attractive, whereas women rated men closer to their own age most attractive. This sparked quite a bit of discussion (such as the comments in the cross-posting of the blog on

So I decided to look at the ratings data in a different way - this time ignoring age, and just looking at how men and women rate each other in general. I found some histograms on p. 16 of Christian Rudder's new book Dataclysm that showed almost what I was looking for, and I then used some graphs from his blog to estimate the data so I could create similar charts in SAS.

Whereas the men of all age groups consistently rated 20-year-old women the most attractive (which produced a very lopsided chart), their ratings of all women in general produced a very symmetrical chart. In Rudder's book he even describes it as "close to what's called a symmetric beta distribution - a curve often deployed to model basic unbiased decisions." Therefore it appears that men are very unbiased/honest in the way they rate women.


By comparison, women rated men very poorly. Rudder mentions that women only rate one guy in six as "above average."


What causes this huge difference in how men and women rate each other? Is one being more honest than the other? Are they rating based on different criteria (perhaps men are rating based on looks, and women are rating based on whether or not they think the men would make a good mate)? Perhaps women are hesitant to rate a man highly, because they know that will trigger okcupid to send that man a message letting them know which woman rated them highly? What other factors are perhaps influencing this data?

Feel free to leave your thoughts & theories on this topic in the comments section!



About Author

Robert Allison

The Graph Guy!

Robert has worked at SAS for over 25 years, and is perhaps the foremost expert in creating custom graphs using SAS/GRAPH. His educational background is in Computer Science, and he holds a BS, MS, and PhD from NC State University. He is the author of several conference papers, has won a few graphic competitions, and has written a book (SAS/GRAPH: Beyond the Basics).


    • Robert Allison
      Robert Allison on

      My plan is to start with some dating-website data that people can easily relate to (such as simple rating data), and if the blogs are well-received then move on to other dating-website data. Feel free to leave a comment with the kinds of data you'd like to see analyzed, and I'll keep an eye out for that data as a potential subject for future blogs! :)

      • Ha. Playing with actual data about dating could get very contentious/strident in the comments section very quickly. :)

        I like the ratings-vs-messaging from the blog post. Those kind of mismatches are very interesting to my eye.

        What about priorities for dataing selection (both stated and acual use)?

        Do we have access to hard data?

        • Robert Allison
          Robert Allison on

          I wish I had access to all the underlying data - but I only have access to what I can 'intuit' from the graphs and text in their blogs & book.

  1. Rick Wicklin

    One explanation would be that the men on OkCupid are not as attractive as the general male population. This would result if the men who use OkCupid are not a random sample from the general population but were a biased sample of less-attractive men who have trouble finding dates through traditional means.

  2. One thing I noticed on dating websites is that men, regardless of age, seem to be looking for someone younger (in some cases, a LOT younger), while women, regardless of age, seemed to be looking for someone their age, plus or minus a certain number of years. This is anecdotal evidence, so I'd be interested to see if a plot of gender, age, and age range of desired partner supports that.

  3. Rob,
    There was a social study I heard recently on an interview that looked at women rating men and men rating women that found a difference in rating if the woman was taking birth control pills. Both genders were affected. It might have been in a Neuro biology article..

  4. Hmmm...having been on dating sites recently and discussing the process with the men it is a different process for women then men. If a woman looks at any profile, you get a response. That can be quite full-on! You want to feel safe so behave differently than you would normally and only look at a few 'safe' profiles. Men do consider more women as an option than women consider men. They are open to more choice but, in the end, they tell me they tend to go for attractive women on the same wavelength as themselves.

  5. Women are biologically programmed to be more selective than men about who they will mate with. Given the enormous disparity between the number of children a woman can give birth to in her lifetime vs. the number of children a man can father, it makes perfect sense that men and women would evolve to have different mating strategies.

Leave A Reply

Back to Top