Your looks and your inbox | Hacker News Reader

Your looks and your inbox

blog.okcupid.com

315 pointsthirdusername102 comments

I have loved all of the okcupid linkbait.

Much of it conforms to the old wives tales or intuitions of people, but "The women who are subjectively rated the most attractive get disproportionately less attention from men. No, really, we're not making it up, we can demonstrate this with SQL" is a whole lot more fascinating than the equivalent finding justified by, ahem, wild speculation. (Or its close cousin, "We surveyed 18 men and 22 women who happened to be enrolled in psych 101 at this campus! Sweet, sweet empiricism!")

I too find statistical confirmation to be interesting, but even moreso is where OKCupid's data have contradicted common wisdom:

- Women rate 80% of males as less than average looking. The graph is completely off-kilter from what you might expect (a pretty usual normal distribution around 50%-attractive)

- The reply rate between sender/receiver attractive is also a really interesting social phenomenon. The unattractive recipient who gets a message from a very attractive member of the opposite sex is unlikely to respond. Self-esteem at work!

I love this sort of statistical drivel (I mean that in the absolute nicest way), it certainly beats reading pop-psych books.

Total agreement. That's the beautiful thing about metrics. Anybody can convince you of what you already believe is true. The real value is when you can demonstrate that reality is different than what you think it is.

- Women rate 80% of males as less than average looking. The graph is completely off-kilter from what you might expect (a pretty usual normal distribution around 50%-attractive)

This is very interesting. It's always been almost impossible to get my gf to point out other guys she finds attractive - this might point towards a reason, although it still doesn't really provide an explanation.

I found this particular comment on the post to be poignant:

"I love that both the males and females are completely delusional but on opposite ends of the spectrum."

Or she doesn't want to wax eloquent about the attractiveness of other guys to her boyfriend?

Women rate 80% of males as less than average looking. The graph is completely off-kilter from what you might expect

It doesn't completely surprise me; it's always been my impression that the average female is more attractive to the average male than vice versa. Although I wouldn't have expected the difference to be so large.

The unattractive recipient who gets a message from a very attractive member of the opposite sex is unlikely to respond. Self-esteem at work!

And the effect is much stronger for males writing to females. I can believe that; an ugly guy who gets an email from a hot girl may suspect it's a setup, but is probably more willing to take the chance.

This would also explain why having multiple wives is much more common than having multiple husbands.

> It doesn't completely surprise me; it's always been my impression that the average female is more attractive to the average male than vice versa.

Average looking != average attractive. If asking a woman about how attractive a man is based only on a photo you will get results like that because you've withheld most of the information she'd use to judge his attractiveness in person.

Depends on the photo. If the photo shows him wearing his expensive watch and sitting in his expensive car, she has most of the information she needs.

My anecdotal evidence is that women go for alpha-ness, not wealth. There's a strong correlation between the two, but it's quite possible to have one without the other. A pick-up artist with no money will attract women, while a well-paid but socially inept geek won't.

80% of males as less than average looking

This could be partly due to the fact that especially good-looking men can find a mate easily, so they never get around to registering on a dating site. To the folks on HN it's probably a normal thing, but I believe dating sites still have a certain 'for losers' stigma in other social circles.

And on the other side, perhaps girls that are very picky cannot find a guy to their liking so they end up on dating sites in greater numbers than the average.

If this is true - why isn't it also true in reverse? (i.e., good-looking women can find a mate just as easily, so would not be registered on a dating site).

Because, with all do respect do my fellow men, most guys out there really are assholes... and unfortunately many social norms reinforce they're actions. I've spoken with quite a few females who hope to find someone who isn't an asshole through online services.

To put it another way, what's the probability of a girl finding a quality guy at a bar versus finding a guy who just wants to take her to bed? I know a lot of girls who are sick of that game.

As a result, you get a lot of guys who aren't attractive enough to pick up random chicks at bars and you get a lot of women who are sick of being picked up by the types of guys that make a sport out of picking up random chicks at bars. Those types of girls tend to be attractive. That's not to say that everyone in online dating services falls into one of these two buckets, but that is why you see such weird skews in the data.

okcupid also has a higher than usual amount of people who are in open relationships or polyamorous, so there might be more better looking dudes than you'd expect.

No, that would have the opposite effect.

I'm not trying to make a value judgement based on lifestyle, I'm just saying that not all the men on okcupid are there because they can't find a girl another way.

Actually it's exactly what I would expect. Cue the "96% of the population is undateable" Seinfeld quote!

Asking people to rate who's below average looking will generally result in the vast majority of people being rated as such. Yeah, statistically you'd expect around 50%, but once you throw human emotion into the mix all bets are off.

OkCupid isn't a representative sample, either. Even more, you have to take into account the site mechanics: a mutual 4 or 5 star rating results in a "match event", but any other combination doesn't. So rating someone 1-3 is equivalent ("I don't want to talk to them") as is 4-5, and men and women's responses to this mechanic may influence their ratings in differing ways.

Indeed. 4 stars for a guy could mean "this person is average looking, but I think I might get to fuck them" rather than "this person is in the top 80% of beauty".

"... top 40% of beauty."

No, it's only a 5 point scale.

1: bottom 20%; 2: 20% - 40%; 3: 40% - 60%; 4: 60% - 80%; 5: top 20%

Am I missing something?

Well, men (but not women) apparently give grades on a bell curve, so you'd expect 40% three stars, but only 5% 5 stars.

No, it looks like I was wrong but there's no link to delete it.

Really? And he didn't mention that in his post? Seems like that completely explains away the difference in attractiveness ratings then.

Two people mutually rating each other 1 star supposedly generates a burn message.

It sure does. Followed, in both cases of my experience, by the woman involved insisting that she just slipped, really, because she would never rate anyone 1 star, even if they were really ugly, which isn't even the case here anyway... ;)

I guess I should be flattered that none of the 2403 people I've one stared didn't return the flavor though truthfully I think I'd get really interested in whoever rated me back. :)

This is the case. Rating someone a 4 or 5 generates an e-mail to that person. So the only reason you rate someone a 4 or 5 is if you want them to come check you out.

Yep, which might also explain why the women are apparently so harsh on their ratings compared to men - they already get more messages than the guys, and don't want to be contacted even more.

I believe the 4 or 5 star match event only happens if you are using Quick Match, not just browsing your matches. I could be wrong though. Also, I don't know what portion of ratings are done in Quick Match instead of just browsing profiles.

You are wrong, it has no connection to quick-match, just rating. I rate a lot of people globaly that never show up on quick match. :)

I don't understand how the QuickMatch/profile rating mechanic can be the ratings cited in the article. IIRC, The QuickMatch/profile ratings are from 1 to 5 stars. The cited ratings are from 0 to 5.

Not sure, but I know you can click "Can't Tell" aside rating them 1-5 which might equal a rating of zero.

For those who were confused by the contradiction of this comment and the actual post, it looks like "less attention from men" was probably supposed to be "more attention from men".

For instance:

"So basically, guys are fighting each other 2-for-1 for the absolute best-rated females, while plenty of potentially charming, even cute, girls go unwritten."

(And it's doubly confusing because patio11 refers to it as an "old wive's tale".)

He's talking about the dip at the extreme right of the messaging graph. The old wives' tale refers to the observation that some women on the extremely beautiful end of the scale aren't approached by very many men because they're perceived as "too" beautiful.

Except that that statement is not true. They may not be approached by numbers proportionate to their beauty relative to the population, but they are still more approached than any less beautiful.

Approaching extra-beautiful women in person, on the other hand, can be intimidating.

The okcupid data and your last sentence contradict your first paragraph. I was over-reaching when I said "not approached by very many men", but okcupid's data implies that there exist women A and B such that beauty(A) > beauty(B) AND suitors(A) < suitors(B). This probably seems unfair to woman A, that her less attractive friend B attracts more men, due to the fact that men find A intimidating. But the fact remains that beauty and suitor count appear to be non-linear at the high-end of the scale.

That said, woman A and woman B are both probably approached by more men than a majority of other women, but their relative success is non-intuitive.

Have a look at the graph again:

http://cdn.okcimg.com/blog/your_looks_and_inbox/Message-Mult...

This shows that the most attractive people get more messages than those less attractive, uniformly. It does not provide evidence for the existence of "A and B such that beauty(A) > beauty(B) AND suitors(A) < suitors(B)".

I think you're confused by the distribution graph:

http://cdn.okcimg.com/blog/your_looks_and_inbox/Male-Messagi...

But the "dip" on the right doesn't mean what you seem to think it does. If there were no dip, and the message distribution line had a uniformly positive slope, it would mean that the majority of all messages would be going to the most attractive women. The existence of the dip simply means that the sheer number of women with average attractiveness get enough messages to outweigh the tiny percentage who are most attractive in total number of messages.

My last sentence was about physical meetings; the rest of my comments were with respect to online messaging on dating sites like okcupid. In other words, I was suggesting that the old wives' tale may have more truth with respect to face to face than online.

I don't think the data indicates that at all. According to the graph, ~6% of women are considered very attractive (5) and receive ~18% of the messages. ~15% of the women are considered pretty attractive (4) and receive ~28% of the messages. That means the population considered pretty attractive is 3 times larger, but only gets ~1.5x as many messages. Thus, those considered very attractive do indeed get more messages than those considered pretty attractive.

The graph is somewhat misleading in its portrayal, but I think if it did indicate your suspicions, the author would've surely said something about it.

A dating site for geeks would have to include automatic A/B Testing for profile pictures.

Can't wait for their analysis of what makes a photo attractive. Especially as for men, men don't seem to have a good grasp as to what would appeal to women.

Based on empirical data (walking through my local park each morning) if I ever put a profile on a dating site the best picture to use would be one of my puppy.

Or you with your puppy featured prominently? That might draw their attention away from any of your flaws, and when they think of you they'll immediately associate your face with a cute and cuddly puppy.

There's also the other idea. What if you put something unattractive near you to make your attractiveness stand out more. Like you volunteering to help a disfigured veteran (no offense meant to disfigured veterans). It might make you look more attractive in comparison.

That sounds kind of evil. It would be a great way of getting advice for dressing, hair, etc.

Come to think of it, HotOrNot probably already provides that.

An old coworker of mine at an internship did exactly this to find the most attractive profile picture for a dating site. He had a sample size of maybe 10 photos and used HotOrNot to do a simple ranking of them. The top rated ones were used as his profile picture on social networking sites.

Unfortunately, he left out one part of the experiment - he did not have the "ugly" photos as his profile picture, measure response rates, and then try it again with a "hot" photo.

I am currently running an experiment with my OkCupid profile along similar metrics - I am varying my profile picture, as well as certain parts of my profile, and testing profile views, response rates, and 4-5 star ratings. This is mostly out of frustration with the pitiful view/response rates I was having originally, but if the data is interesting I might do a write-up.

for that you want 'http://styleguidance.com/

I think A/B testing would be unhelpful or even counterproductive because of the difficulty of picking a measurement to maximize. You want a flattering but reasonably accurate advertisement for yourself, which doesn't correlate with any easily gatherable metric. If you try to maximize clicks or emails, you run the risk of making your profile marginally more attractive to lots of people you aren't interested in at the cost of looking generic and boring to the few people you are interested in. You could reduce this effect by maximizing the percentage of profile views that result in emails, but it would still be distorted by people who mistake you for someone entirely different. If you maximize emails from interesting prospects, which is what you really want, it will take a long time to accumulate a useful amount of data.

"Totally decent-looking" dudes are actually co-founders of OkCupid :)

http://www.okcupid.com/about-us

As you can see from the gray line, women rate an incredible 80% of guys as worse-looking than medium. Very harsh.

The female rating is particular interesting. Are only ugly men on OKCupid (myself included)? Or do men look worse in pictures than women? After all, remove the makeup, carefully selected attire, and fancy hair, and you have your average male. Even photogenic men don't look as good as their photogenic counterparts, in my opinion.

most guys shoot from the hip when it comes to pictures, women prep the crap out of it.

Hell I know a girl who used photoshop to tweak a photo she used as a profile picture on facebook

If you're serious about online dating you should probably be getting professional profile photos.

I disagree. I've done quite a bit of online dating, and professional photos (a) don't show your fun side, and (b) often look quite different than how you look on an average day. What this results in is me distrusting them -- before I meet anyone, I need to see a candid shot of them in natural lighting.

I think there would have to be a special kind of service: a photographer who can take photos that look as if they were casual, but aren't. Even for Twitter and other social networks casual is probably better, but few photographers (if any) understand that so far.

Any photojournalist worth his or her salt could pull that off. Might be a nice little side-income.

I am a photographer, and I've had inquiries for this. No one is willing to pay more than a few dollars, which doesn't even offset the time required or the depreciation of the equipment.

As a photographer, what is your opinion on the "going rate" or similar for a spread of five natural looking photos, compared to what you get offered ("just a few bucks")?

Curious if it's similar to what happens when people in "typical" social circles find out I'm a freelance web developer, offering as little as $100 for what could easily (after graphic design and so forth) be two weeks of work.

What would be a decent rate for that kind of work? More than classical portraits? Honestly the "social networking" photographs would be more important to me than the classical photographs for job applications. As they say, job applications are dead - but employers and clients are likely to look at social network profiles...

How much would it have to cost to be worth your while?

I don't know where you are, but around here (Australia) little mall studios for glamour shots are quite common. People seem willing to pay $50-$100 for some underwear shots.

Get professionally taken casual shots in natural lighting.

Look on the bright side: If 80% of men classify as "worse-looking than medium" then they are actually on a more level playing ground with each other.

Not really surprised about the female curve. They say attraction for males is 85% physical and 15% personality, and for women the reverse. Not sure on the accuracy of those numbers, but the general principle is probably true. And since you get no inkling of personality from a picture, it isn't shocking.

What I'd really like to see is if they listed the guys to one group of women as just photos, then to another given select bits of data, what would happen then. I suspect you'd see strong correlations, especially if income were there.

> you get no inkling of personality from a picture

Really can't agree with this. How people present themselves in self-chosen photos tells quite a bit about personality.

I agree with the OP -- from my experience, it's very difficult to tell how a person will act on a date, personality-wise, merely from their photos, profile, and non-video chat. And personality is such an essential part of a person for dating.

And that's why people never fall in love with jerks, because all they care about is the personality.

Maybe your intuition about the attractiveness of being a jerk is wrong?

Don't understand, what do you mean?

I mean that being a jerk can make someone -- usually a guy -- more attractive, rather than less. (Your comment implied that you thought that when a person falls in love with a jerk, it's because that person is so attractive physically).

It's not the jerkiness (?) that's attractive, it's the self-confidence. That's what people really mean when they say "women prefer assholes". Bad behaviour is a proxy for confidence. Self-confidence trumps all other factors from a woman's POV.

Well, there's a whole niche culture (PUA) built around this, and there are so many contradictory observations and prescriptions that I don't know that I even have a position on exactly what aspect of asinine behavior is sometimes attractive. I would note that it's not uncommon to see a girl devoted to a guy who seems self-loathing (as well as being a jerk, at least to her), so self-confidence is not always a major part of the picture.

From what I read about "the Game" a common play seems to be to make the girl feel insecure, though. Maybe other factors than just self-confidence are sometimes at play. (I have not read "The Game" itself, though, only blog articles mentioning it).

Man, ain't that the truth. My favorites are the following categories:

- The only good photos I have of myself are with friends in bars.

- My cat/dog has to be in the photo. I love my cat/dog more than I'll love you.

- The only time I look good is with my ex, so I'll crop his side of the picture out. You won't notice, will you?

- 20 self-photos of me in the bathroom mirror. Don't mind the mess in the background.

- I love to travel! Can you see that speck at the bottom of the Eiffel Tower? That's me!

Thought of a few more:

- I was a bridesmaid! Try and find me in this group of 8 well-dressed women! This says nothing about my craving to be married at all!

- I could only get my coworkers to shoot this photo of me at the office. Can you see I hate my job?

- Here's some cleavage to distract you.

- This is me & my dad. I'm not a daddy's girl at all, but get ready for the interrogation when you pick me up.

That doesn't explain rating perfectly OK looking guys as 1 or 2, which is what seems to be happening.

I would have expected 3-4 to be most of the ratings. When I ask women if sush & such is good looking, they almost always say 'he's ok.' I concluded thought that women have a huge 'ok' range & small 'no way' & 'huba huba' ranges. What I should have concluded is that they're all liars.

I've heard the theory that men mostly agree about who they consider attractive, but that women have much more varied tastes. Which of course suggests that women should approach men, not the other way around.

Seems like we should ask the okcupid guys.

I question the numbers. (Well, the female ones anyway.) I had a plentyoffish account a few years ago, and the image ratings averaged around 3.5 out of 10 for both men and women. I know that the hotornot guys had the same problem, and admitted to fudging everything upwards.

    fudged_hotness = 10.0 * ((hotness / 10.0) ** (1 / gamma))

Then just crank up the gamma correction to 3.0, and your 3.5 out of ten is now a reasonably-attractive 7.0472987320648917 :)

Favorite quote from the other posts on the blog: "This is what 'dying alone' looks like as a png." Lots of fun statistics, and some humor to boot.

Not as hilarious when all the attractive girls ignore your messages because they are flooded with spam, however...

If they're ignoring you, doesn't that mean your messages are spam too?

Sure, if they log into the site to read it. Most people got tired of bothering long ago, however.

this sounds like a problem to solve. i'd like to see a dating site that doesn't involve browsing profiles like a catalog, and encourages people to respond to messages a bit more.

kills two birds with one stone, if you have it based on a smarter matching algorithm that requires you to give private feedback on people you interact with, to help hone your personal recommendations. also makes fake profiles and spamming a much tougher prospect, and would probably give it an overall safer feel.

tl:dr Women don't have unrealistic expectations, they're just saying that the average man is ugly.

I disagree with their interpretation that the fact that women rate men as low attractiveness means they "have unrealistic standards for the 'average' member of the opposite sex."

Remember, they didn't ask "do you think this guy is above or below the average attractiveness." They asked "is this guy attractive." There's an implicit assumption that "neither attractive or unattractive" is equivalent to "average attractiveness of men."

That assumption is wrong. Men are ugly. So of course most men will be rated as more unattractive on average than women.

interesting. women rate men much more harshly on looks, yet a woman's attractiveness is much more important for whether they receive messages.

sounds like attractiveness is rated poorly for women partly because it doesn't speak to women.

Ok, I'm going to go out and say it: There is an absurd amount of pressure on women in this society to be perfect.

Now, I'm not going to refute the fact that I wouldn't message the most-attractive women as well. But if the knee-jerk reaction of a woman is to reject the message from the most-attractive men or to harshly rate decently looking guys, there are outside factors affecting the female perspective of appearance, especially if the same isn't true for men.

sure. my first thought was: maybe the average guys on the site are worse looking than average. who knows? the statistics here are ridiculous.

To expound, men may have to retreat (when they're "retreating") to OKC because of below-average looks. Woman may have to retreat because of below-average personalities.

Maybe it's not "defective" people that post on dating sites, it's outliers who aren't likely to hook up with a random and have it work out. Maybe people with strange hobbies, beliefs, intelligence levels, whatever.

Did you read the whole article? Men's attractiveness ratings of women are healthily distributed, almost unbelievably so. And women rated about 8.3/10 (rather than perfect 10/10) got the most messages.

I did, but I'm not sure what in your comment disproves the self-esteem issue. My point is not what men think of women (how they rate them), but what women do when presented with the options I mentioned (attractive guys messaging them).

If an attractive guy messages me, I would be thinking as a woman "Wow, this is validating" instead of "No, he must have made a mistake". If it's the latter, you have to wonder what initially gave her the idea that he made a mistake? Was it something deeply engrained?

Maybe I didn't read the article correctly, but please don't assume I didn't read the whole thing. Starting off your comment like that is not very welcoming and doesn't add any value to the conversation.

You didn't talk about self-esteem. What you said was this:

>There is an absurd amount of pressure on women in this society to be perfect.

I think the data show precisely the opposite - that women aren't under pressure, at least from men, to be perfect. That was the point of my comment.

Actually, in my experience, women are many times more guilty of the things they typically accuse men of than men are (e.g. objectifying the other sex, being judgmental about looks, etc.). Men are under at least as much pressure as women in these regards.

And women rated about 8.3/10 (rather than perfect 10/10) got the most messages.

This is a misleading statement. As a total number of messages, they did indeed get the most, since there are more of them, but the ones rated 10/10 still got a lot more messages per individual than those rated lower. And overall, the messages versus attractiveness curve is much steeper for men messaging women than the other way around.

I met my girlfriend on ok Cupid over a year ago within a week of signing up. The site was a lot less spammy then, and most would say we just got dumb lucky.

I've done other sites and while pictures matter, if you're not a 4 or 5 not coming across boring or desperate will make you a 4 or 5 in comparison.

As far as I know, women date by "how you make them feel?" not how a guy looks :)

okcupid is doing science. I wonder what academic will pick up this data and extract evolutionary biology conclusions from it.

I don't know why you're being voted down - it seems to me like they are doing science, and I wouldn't be surprised at all to see a reference to Okcupid in a university study or a pop science book.

If they release their dataset, it'll spark a lot of research. Datamining researchers today use a lot of 'hot' datasets, from Netflix to ePinions, which are sometimes old, but they are available. In some research domains, it is not uncommon for data sets to have huge importance, including getting you tenure and being guaranteed resources to exploit the dataset you gathered.

In this sense I meant it: that they are collecting valuable data. That their own analysis is flawed by their own bias is evident to me, but does not diminish the value of their data.

The idea of okcupid "doing science" reminds me of the quote from the video game Portal, which is appropriate, because both things involve the thought that something is "doing science" when it really isn't.

Doing science, if we are to go by Feynman's exposition of it, involves more than just gathering data--it also means accounting for everything that can make your data wrong. OkCupid is just gathering data and presenting it. This is an important part of science, and it's interesting to read about, but it is not a sufficient condition for having done science.

okcupid is doing science in the same way that Mythbusters is doing science.

Obligatory xkcd reference: http://xkcd.com/397/

Reply on news.ycombinator.com