February 14, 2011

OK Cupid and the attraction curve

It’s Valentine’s Day, so it’s fitting that I finally visited the dating website that my favorite podcaster, Rob Long, had talked about in a recent episode. Turns out the clever nerds who run OK Cupid, a booming singles site, have stumbled on a surprising statistical truth about which members get the most attention...a truth that helps explain something I’ve noticed in our own surveys of cultural audiences for years.

Now, bear with me. It’s a big leap from dating-website data to how and why people connect to an orchestra or a natural history museum. But as Rob notes in his characteristically wry take on the data, the ins and outs of attraction are a pretty good metaphor for all kinds of human preference-related behaviors, especially in leisure-time, feel-good categories like music, art, and entertainment. “Everything, when you get down to it, is kinda like dating.”

And this post traffics in the objectification of women, a dubious first for me. The OK Cupid crew have analyzed stats from their site about which female members are considered hot and which get the most messages from other members. They promise to do the same for (to?) men soon. Meanwhile, if you’re particularly sensitive to “lookism,” skip down to the bottom and post a disgruntled comment.

So what’s the big reveal? That the women on the site who get the most attention (in the form of messages from other members) aren’t the ones with the highest average attractiveness ratings. They’re the ones with the most disparate ratings — the ones about whom opinion is divided. Lots of 1s and 5s in your ratings is better than lots of 4s. As OK Cupid co-founder Christian Rudder puts it in his post about the analysis, “Guys tend to ignore girls who are merely cute” (that is, fairly but not outrageously attractive), “and, in fact, having some men think she’s ugly actually works in a woman’s favor.”

The whole post is fascinating, and the statistical analysis looks strong, especially for that counterintuitive last bit about how the lowest attractiveness ratings actually contribute more to the attention the member receives than the second-highest ratings. (And for the record, we’re not talking about negative attention. We’re talking about the correlation between the distribution of attractiveness scores and the number of approaches that men make to female members, presumably with a relationship on their minds.) ...

Except that, as someone who spends a fair amount of time looking at graphs of how people feel about, and behave toward, cultural and educational organizations, I don’t find this counterintuitive at all. For years I’ve been sensing a disconnect between the largely positive ratings that audiences give our clients and the lukewarmness of their relationships with many of those organizations.

The OK Cupid data points to a possible explanation: that the organizations are perceived as “merely cute” — pretty attractive to pretty much everyone in their audiences, with little dissent. And, let’s face it, not much to dissent about. The satisfaction curves of our clients show either broad, general enthusiasm (like this one for a ballet company’s website)...

...or slightly more bell-shaped responses when we ask them how the organization in question stacks up to others of its type (like this one for a natural history museum).

What we never see are the barbell-shaped curves that OK Cupid says are associated with greater interest and energy. Why not? Maybe because the organizations do exactly what most dating site members try to do: play down their flaws, try to appeal to as many people as possible, even if that means becoming a little generic. I’ve often said that most cultural organizations are afraid of making mistakes or showing their weaknesses; they hide any flaws or oddities under the crisp cotton of professionalism. That’s why experimenting is often hard for them.

But human nature may not reward people or institutions that are unobjectionably, categorically attractive in the eyes of the world. We need something to get excited about, something interesting and unique. Something that somebody else may find unappealing, even ugly. The OK Cupid analysis suggests that really sparking some people’s attention and making them want to get close to you requires turning other people off.

Which could be very freeing. How many times have you heard someone in your organization say — or said yourself — that “if we’re for everyone, we’re not really for anyone.” Now we have some data to back up the thought. Create variance, as the OK Cupid staff advises. Get them disagreeing about you. You may find yourself with more passionate, and more numerous, fans.

Nina Simon — February 15, 2011

OKCupid is the only blog I had to unsubscribe from because I found it so fascinating I would get lost in the geek factor for way too much time. Now I use a "read later" function to read the posts when I truly have no other work on tap.

Gillian Savage — February 16, 2011

The big difference between the data from Cupid and the data for cultural institutions is that the data from Cupid applies to people the rater knows nothing about except a photo and a few lines.

In contrast, most data about cultural institutions comes from people who have been there, often because they chose to go. I'd say that most people who would rate a museum 1 or 2 based on a photo or description would choose not to visit. So, most potential 1s and 2s are excluded from the sample. That's why you don't get the bar-bell picture in the data for them.

Another interesting aspect is that the infamous Net Promoter Score deliberately excludes the 'OK' ratings from the scoring. They use the bar-bell.

Enough from me,

laura grace — February 16, 2011

The OK Cupid findings don't actually surprise me. I feel like we humans tend towards black and white thinking. Everything is fight or flight, good or bad, etc. People might be more attracted to the simplicity of the decision to date someone who seems easy to figure out rather than to contact someone they see as shrouded in the grey. Thoughts?

Peter Linett — February 17, 2011

Gillian, actually one of the graphs above is from community, not museum visitor, responses, i.e. it includes plenty of people who don't go, or go only rarely, to the museum in question. We've consistently found that even non-users rate cultural organizations highly: they're positively disposed to the whole category "museums" or "the arts," or at least they think they should be so disposed. Sociologists like to call this "social desirability bias." It means we need to discount the positive ratings somewhat, since they can be a little abstract and tied more to attitude than to behavior (and from the organizations' standpoint, behavior is the point). I think museums and arts orgs are complicit in shaping these diffusely positive attitudes, and are loathe to put them at risk with anything truly new or out of character.

Laura, that may indeed be a facet of human nature, but it's certainly exacerbated by the ubiquitous opportunity to rate things. Qualitative research methods tend to reveal much more nuanced thinking, something I'm often surprised and impressed about when listening to focus groups or interviews. But from Amazon and TripAdvisor to singles sites, we're asked to pick a number far more often than we're asked to reflect and discuss. With predictable results.

Thanks for the comments, and don't be a stranger...

Kristina Newhouse — February 25, 2011

Would love to be able to post this on Facebook!

Paul Botts — February 25, 2011

In fact I just did so, and also forwarded the link to members of the two theater boards that I'm on as well as some others who will find it interesting. Good stuff.

Laura Grace — February 28, 2011


I definitely agree. I worry that some quantitative research methods are skewed by the sense of power that subjects feel when given the opportunity to rate things. I tend to research and follow political studies and (cringe) polls. Polling methods seem to give people the same judgement power of the Romans in the Colosseum. We would rather give the proverbial thumbs up because of the sense of self-satisfaction we get from negative judgment.

However, I am definitely guilty of something similar. When I am sorting through product ratings I always search out the worst ones to see what people are saying, even if a bad rating is only 1 in 100. Maybe I am looking to give the issue a fair and balanced approach. ;) This seems somewhat similar to OK Cupid in that the consumer pays more mind to the lowest ratings than the second highest.

Thanks Peter, this was definitely an interesting article, and I have found myself following Rob Long.


Janis — February 28, 2011

This is interesting -- it's underlined an opinion I've had for some time that if you want to get people's attention, you will have to get their attention ... one way or another. And by doing that, just by being unusual enough to get noticed, you will get up someone's nose somehow. I've heard that Sophia Loren was repeatedly told by directors when she was a girl that she was frankly ugly. One said, "There is no way to photograph this girl and make her look good." And I don't doubt it would work for men, too. I know as many women who think Steve Perry is unattractive as think he's the second coming, and the same oddly shaped mouth that makes a lot of women like Ioan Gruffudd turns a lot of other ones off.

Being one thing powerfully will arouse strong feelings one way or another -- the only thing that won't accomplish this is a bowl of vanilla ice cream.

I can't remember who it was said that all great beauty has some strangeness in the proportion. Beauty IS an exaggeration, after all. And I'm sure that's why a lot of classical music is ... well, bland anymore. If you want to make one competition judge hit the ceiling with joy, you'll probably make another one hate you.

Post new comment

The content of this field is kept private and will not be shown publicly.