When OKCupid abolished photos for one day

These were the results:

1. People responded to first messages 44% more often.

2. “conversations went deeper”

3. Contact details were exchanged more quickly.


When the photos were restored at 4PM, 2,200 people were in the middle of conversations that had started “blind”. Those conversations melted away.

That said, the people who actually used the “Blind Date App” if anything seemed slightly happier with their dates.  The full report from OKCupid is here.  Yet here is the combined chart drawn from when people score “looks” and “personality” separately.

By the way, I would never try to match you up with a book I fear you may not like, at least not without telling you or otherwise signaling that incompatibility in advance.


Interesting. Also interesting how when photos are up, those that choose to withhold photos get almost no responses.

The problem, it seems, is that when somebody could post a photo but does not, other people assume that there is a reason they don't want to show their photo.

When people know that the person is not choosing to withhold their photo, there is no such problem. Still, I think having photos on date sites is a useful tool for mate selection.

Also slightly related, a good piece on feminism that talks a good deal about the problems regarding male and female interaction:


I think Marginal Revolution should post quote or commentary that Tyler claims he likes, without disclosing the author.

Then, we should allow time for people to comment on the quote.

Then, midway through the exercise, Tyler should reveal that the author was

Paul Krugman.

Right now this blog has a picture of Glenn Beck staring at me from the upper-left. I think it's a staring contest.

Careful, if you look him straight in the eye you'll get mesmerized. I've seen it happen, believe me.

Well, it isn't as if there are all that many 'raving freakazoid nut sandwiches' available, after all.

Luckily for this web site, Beck tends to identify himself as a libertarian, so it isn't as if such ads violate any principles espoused by a chess playing Harvard graduate.

You still don't understand how money works.

He is selling a math product.

On an OKcubpid post.

You guys' weird brand of trolling/concern trolling is truly bizarre. And you sincerely don't seem to realize it.

If someone can explain that math product that would be helpful, btw.

yeah it's kinda creepy. ..I liked the cone advert more

Bill, I will bet you...ummm...street cred, that I can name a NORMAL PK BLOG post in...hmmm...3 sentences.

The problem isn't me.

Not to mention, TC does this ALL the time.


Street creed is pretty low.

Got something else to add to the pot.

Nah. People hate on Krugman for his nakedly obvious partisan bias, not because whatever he says is wrong by definition.

Hilariously, what he says changes 180 degrees depending on who is in office. You could post a Bush-era quote from him about how US debt is disastrous and I would wildly applaud. Now that we've got Obama, of course, you can simply never have too much debt, and only idiots would think otherwise.

Mixed strategy. That's how you confuse your enemies.

His enemy is the truth.

A more fun game would be to identify the year of the quote, even just post or pre 2000.

That's an interesting reply, since it means that you have agreed with Krugman in the past.


If you maintain that your position has been constant throughout the period, and that Krugman has switched his positions several times during this period,

This must mean

You and Andrew have agreed with Krugman at some point in the past.

A liberating thought.

I don't get the Knausgaard reference. Was he being sarcastic, or is he sincerely trying to say that he would never recommend "My Struggle" without warning us that he may not like it?

This blog is *sooooo cryptic......why say such blatantly opaque things to confuse us?

Tyler teaches us to read between the lines, but at the cost of our own sanity. He's discovered a sweet spot that a certain type of mind finds irresistible when he tickles it.

I still wonder what's real and what's not on this blog.

Anything relating to the rich getting richer is to be completely trusted.

The oppressor-oppressed view is orthogonal to what we believe.

Scandinavia is awesome and Tyler feels uncomfortable with national feeling are what I get from this. I'm really not sure with what Alex is up to.

The fjords are better in NZ.

I don't think Tyler is much into nature. As to fjords it depends on what you want from a fjord, I think if you disdain Norway's fjords you haven't seen enough of them, but then I like lots of skerries so I'm partial to the inland passage and Tierra del Fuego, but then I have never been to NZ.

He's sound enough to like puffins.

He is either saying (1) "Don't judge a book by its cover" or (2) the human biases of loss aversion, endowment affect, and the tendency to shoot the messenger make the population optimum trial-and-error strategy for matching a risky proposition for the matcher.

"Conversations went deeper". Anything else?

PR stunt disguised as a study to promote a liberal message

is this a spambot or a real response?



Who needs Roko's Basilisk when we have the trolls to punish us.

Even the simulation is underwhelming.

What's liberal about disclosing that people actually judge by looks? That's applied common sense, you can't get more illiberal...

Actually, there is literature on that people scan for facial deformities and uneven facial features as a signal of poor pre-birth nutrition and mental incapacity and that this is an evolutionarily functional trait that prevented bad genes from accumulating in the next generation.

By that definition, this happy person : - ) sleeping on his/her side should make you want to sleep with him/her.


Most of these results could just be a glaring example of selection bias. Their online users dropped off a cliff during the experiment. Maybe just the people that tend to respond to first messages stayed online. Maybe those same people are more willing to share contact info quickly.

One imagines thirsty women couldn't prescreen with photos anymore and had to pretend to be human beings for a while.

Bet they wish that hadn't shown that. Oh well!

Long live Karl Ove!

I´m reading A Man in Love now and I just love it, living in Stockholm as I do makes it even better. His distain and loathing of certain swedish traits are hilarious and painfully true (conformity of opinion)

I had to call my girlfriend and read parts to her as I read it, I read 300 pages in one sitting just now.

Long live Knausgård indeed.

OKStupid gave a plausible suggestion. Site traffic went down without the photos, but the users who were left were less superficial. But I didn't get that impression from the excerpt.

If it turns out that not including the photos encourages less superficial use of the dating site, its interesting how OK Cupid should respond to that information. Maybe have a "hot or not" section of the site where viewers rate the attractiveness of users based on photos, but without any information as to the name or otherwise identifying the user. But on the user's available profile, the attractiveness rating will be shown but NOT the photo.

This gives people information about a potential date's physical attractiveness, but in a form that is digested quickly and encourages moving on to other information. The photo is probably too distracting/ commands too much attention in relation to the information it provides. People using the site could screen for the attractiveness rating, based on their preferences.


Is that the website for married couples?

Photos=superficial use is certainly the message the writer is pushing. But after you read a hundred profiles along the lines of

"I'm a walking contradiction. I'm equally at home at the opera or gutting a fish. Or gutting a fish at the opera! I like all kinds of music, and my friends think I'm great. Also, I love sports!"

your opinion of the written portion of the profile will go down a little.

Having used the site a little, I am not surprised by their results. A few observations:

1) The photo is by far the most useful piece of information on the profile. Most people aren't good enough writers for their character to come through in about 6 paragraphs. ("Movies, bands, etc I like" and "six things I could never do without" are nearly useless, and a lot of people answer "the most private thing I'm willing to admit" with "I won't admit anything on the internet". Add in the people who use the "I am a walking contradiction" self summary, and you've got 1-2 paragraphs of semi-useful information.)

2) The process isn't really about search, it's about winnowing the set of possible matches to a smaller pool that's worthy of spending more effort to evaluate. People have a limited number of Saturday nights, and a limited emotional ability to consider new romantic partners. More than one new contact a week is stressing the system.

3) Consistent with this, most people want to meet relatively early in the process. They don't want to put in two weeks of emailing, only to have no chemistry when they finally meet.

4) "Nice, but no chemistry" is probably the most common outcome of a date. Note that, except for "complete psycho stalker," this is actually the worst outcome -- you put time and hope into one person, and you end up disappointed.

Put all those things together, and it's clear that the photo is the single most useful piece of information. It tips people off about clearly unsuitable matches, and gives a hint about the likelihood of "nice, but no chemistry."

Lacking the photo, people use the site a lot less, and those who do put more investment into inferior forms of winnowing such as exchanging emails. When they go on dates, they have generally positive experiences, but as I said, this is already the most common outcome, and in some ways the worst because it takes a lot of investment to reach a negative result.

The OKCupid blog consistenty pushes a "looks aren't everything" line -- either to encourage people to expand their horizons or to reassure people that they aren't doomed. But if they want that aspiration to be true in practice, they need to increase the value of the other information in the profile. In particular, they need to increase the ability to winnow people based on personality rather than just looks.

If they want to increase the value of the non-photo information, I would suggest more aggressive screening of spammers and creeps. When I've asked girls about their experiences on the site, they're usually generally positive, but with a lot of unsolicited requests for sex, plus a few people who seem normal online but act weird or overly sexual on the actual date. That goes double for attractive women and extra super double for really attractive women. Since OKCupid's users are also its product, it seems like they're giving the worst user experience to the very people they want most.

Sounds to me like a simple conflict between sex drive and yearning for companionship. Most of the people using the site would be more successful if they concentrated on looking for companionship, but they can't help themselves prioritizing sex. So with photos they narrow down their targets to the sexually attractive, strike out, and feel bad about themselves. Without photos they have to assume there's no sex potential and focus on finding an interesting opposite-sex friend, and if they're willing to do that, they're far more likely to succeed and feel good about themselves.

I have a new model for the dating website.

One of your parents serves as the proxy who writes your description or edits your profile, the profile goes under your parents name, so that the other person who responds to the website (the parent of the other person) passes on whether you go on a date.

After the date, both parents report back to each other the results of the encounter as heard from their young adult child.

Better as a screening, monitoring and deterrence device than Christian Mingle, which apparent attracts predatory males.

Comments for this post are closed