Jon Millward – Blog

Cupid on Trial: A 4-month Online Dating Experiment Using 10 Fictional Singletons

Cupid on Trial - Putting OKCupid to the Test

Cupid on Trial

Is online dating a different experience for men than it is for women?

Cupid on Trial

To find out, I conducted a 4-month experiment in the US and UK using 10 dummy dating profiles.

Cupid on Trial

Here’s what happened…

A line I made in Photoshop. I suck.

Only eating and sleeping could be said to have a stronger grasp on the steering wheel of our daily behaviour than the thing in our heads that is constantly urging us to find love and have sex. But even an insatiable appetite and overwhelming tiredness are no match for the sudden arrival (or breakdown) of pure romantic love, or unbridled sexual lust. These are, after all, the states of mind that inspired every one of our direct ancestors to relentlessly pursue love and sex until they succeeded at least once in getting their genes into a new generation. We are each the product of an unbroken string of successful fuckers and lovers, so it’s no wonder fucking and loving pervade our thoughts as completely as they do.

The advent of online dating, then, must have seemed like an incredible idea. Whereas in the past the pool of single men a woman could potentially meet and attract was limited by who she happened to physically be around during daily life, now it was exponentially larger. Now the number of men she could date was limited only by how far she would eventually be willing to travel to spend time with them in person. Dozens of suitors turned into thousands, or even millions.

However, things turned out to be more complicated than that. Just as freshly-online businesses, expecting to amass untold fortunes in a new, global market, found themselves in competition with internet businesses that they would never have otherwise had to compete with, so too did online daters face the prospect of having to stand out as special and attractive amongst a much larger pool of singletons than they were used to. Whereas before a man just needed to be the best looking guy at work to get a date with a colleague, now he needed to be in the top 10% of all men to get a date with one of the women in his city.

The expanded horizons offered by online dating don’t equal unrestricted access to a ready and waiting list of beautiful people. Every man and woman online still has criteria that must be met by people who want to date him or her, and every guy and girl is still in direct competition with every other person of their gender. In that case, then, is the acquisition of love and sex online just as easy or difficult for men and woman as it is offline? Or does this new social arena amplify the dating frustrations each sex has struggled with since the dawn of time?

To find out, I decided to peek behind the curtain and get a glimpse into what online dating is like for men and women who are of varying levels of physical attractiveness.

Things were about to get sneaky.

The Experiment

How many unsolicited messages do men get compared to women? And what difference does their physical attractiveness make to each man and woman’s success?

Phrased another way, do women have it a lot easier than men, and do hot people in general have it the easiest? I know what you might be thinking: yes and yes. It’s hardly the unsolved question of the century. However, at this early stage I didn’t know exactly how big the gap between men and women might be, or how different a relatively unattractive person’s online dating experience might be compared to someone more blessed in the looks department. Nor did I know what to expect to see in the unsolicited messages, because men rarely get to see the messages women receive from hopeful boys, and women rarely witness the reverse. I’d have a privileged, and somewhat immoral, view into both.

The Dummy Accounts

Morals aside (where would space travel be without the unpleasant demise of Laika the Soviet space dog?), I set about creating ten dummy dating profiles on the world’s fastest growing online dating site: OKCupid.

I’ve used OKCupid for several months for my own love/sex life, so I was very familiar with how its system works. There are three main elements to having a presence on there: your written profile, your photos and the inane interesting questions you’re supposed to answer to help the matching system pair you with likeminded people. It’s a pretty flawed concept and one that I’m sure is only there to help them serve you relevant advertising, or make you feel like you have a hand in sorting through the horde of freaks that inevitably lurk in the shadows.

Anyway, for each of the ten dummy accounts, I answered 25 of OKCupid’s questions in exactly the same way. The questions ranged from the obvious to the ridiculous.

OKCupid Questions

I also gave the ten accounts very similar sounding usernames, again, so that nothing would immediately differentiate them from each other (I wanted the photos to do that, because it was the influence of gender and appearance on the number of unsolicited messages received that I was interested in).

The Written Profile

For the next element of the accounts, the written part, I created one single solution: a bunch of answers to OKCupid’s default sections that ALL of the ten dummy accounts would have. In other words, all ten would have the same written profile, once again so that this part of the accounts wouldn’t sway people towards or away from sending messages.

The written profile I created didn’t give any clues as to the owner’s gender and it included a few ‘hooks’—mentions of party tricks and whatnot—to give people something to talk about in their messages.

Here it is, if you’d like to read it.

OKCupid Profile Writing

The Photos

Now for the interesting bit: the profile pictures. I selected five photos for the boys and five for the women that depicted men and women who I personally thought varied in physical attractiveness.

Then I gave the ten photos to three other people (male and female) who would act as independent judges of the girls’ and boys’ looks by ranking them from best looking to…not the best looking. Pleasantly, for the experiment, all three judges agreed on the rankings.

The Boys and Girls

At this point, I had ten profiles with similar sounding usernames, all with the same answers to 25 questions, with the same written profile and personal stats (all heights consistent, the same level of education, etc.), and each account had a different photo of a man or woman.

I then herded our collection of fake people onto Deception Airways and pretend-flew them to five different US cities, where they would be allocated in pairs. The best looking man and woman in one city, second best boy and girl in another, and so on.

US Distribution of OKCupid Profiles

Then I waited.

The Waiting Game

I must admit, I was excited. If you ever want to feel like a small-time god (albeit a somewhat meddling, devious one), I recommend creating multiple dating profiles.

Before we get into the results of this first stage of the experiment, here are a couple of interesting online dating facts that hint at the complexity of meeting people online.

• While online daters rate their photos as relatively accurate, independent judges rate approximately 1/3 of the photographs as not accurate.[1]

• Female photographs are judged as less accurate than male photographs, and are more likely to be older, to be retouched or taken by a professional photographer.[1]

• Men lie more about their height, and women lie more about their weight, with people farther from the average lying more.[2]

• In a survey of online dating users, over 80% of participants registered concerns that others misrepresent themselves.[2]

To be fair, my tinkering with the system wasn’t really helping with that last statistic. Anyway, science and all that! Onto the results.

Results after 24 hours

Messages after 24 hours

After being online for 24 hours, the ten accounts between them had amassed 90 messages. Remember, for this experiment, it was all about unsolicited messages—I sent no messages to anyone and never replied to ones received.

As you can see from the graph above, the women got many times more messages than the men.

• Each woman received at least one message, but the two best looking women received 581% more messages than the other three combined.

• Only one man received any messages.

Messages per profile views

If this was a sign of things to come, then it seemed a sure thing that the women would get messages without any extra effort on their part, whereas the men’s inboxes would be markedly less full.

Results after 7 days

Results after 7 days

The above graph shows the results after the profiles had spent 168 hours online.

• The most contacted woman had almost 17 times more messages in a week than the most contacted man.

• Three of the men had no messages, despite their profiles being viewed about 25 times between them.

The women’s messages outnumbered the men’s 17 to 1 (mostly thanks to the two best looking women).

• The two best looking men received 5 fewer messages than the 3rd and 4th best looking women.

In summary, when it comes to receiving unsolicited messages based on gender and photos alone, women wipe the floor with men, and very attractive women sandblast the floor with the fellas. They kill. Their inboxes heave with hellos and how are yous.

To make sure it wasn’t just an American thing, I refuelled the Deception Airways jet and relocated all of the profiles to the United Kingdom for a much longer stay. 4 months in fact.

Changing Countries

UK Distribution of OKCupid ProfilesBy this point it was obvious that women on OKCupid, and probably all dating sites, get a lot of messages from men, and if a woman happens to be very pretty, she’s swamped with attention. This is not breaking news, because most women who have tried online dating quickly discover what it’s like for females online. It’s the same as offline, except exaggerated. If they are hot, the girls can pick and choose which men they interact with. If the men are hot, they will get some unsolicited messages, but the attention they receive will be several orders of magnitude less than their female counterparts.


The worst looking men and women, unfortunately, are in a similar boat to each other—the ‘Not Much Attention’ boat, which is scheduled to arrive at Love Island, but no one can be sure of when.

I left the 10 profiles dotted around England for over 4 months while I spent time on other projects, like analyzing the last words of 478 death row prisoners, then I returned to the accounts to see what had happened in my absence.

Results after 4 months
The Final Results

As you can see, the results after 4 months echo those from a week into the experiment.

• The women as a group received over 20 times more messages than the men.

• The two most attractive women received 83% of all messages.

• The two most attractive women probably would have received several thousand more if their inboxes hadn’t have reached maximum capacity.

• It took 2 months, 13 days for the most popular woman’s inbox to fill up. At the current rate it would take the most popular man 2.3 years to fill up his.

Apart from seeing the difference in message volume, this experiment also allowed me to see the content of messages received and sent by men and women. My impression, after reading several hundred in the women’s inboxes, is that most men compliment the attractive women a lot, they make reference to something in the woman’s profile (you would not believe how many times men mentioned the party tricks and ‘Arrow’ the cheetah from the generic profile I wrote), or they ask a general question about travel or something equally boring.

They are rarely, if ever, imaginative and I sympathise with any woman who has struggled to find any diamonds amongst the rough myriad of messages she is bombarded with each and every day. Then again, what can a man say that hasn’t been said before?

He has to make a good impression and show he’s attractive without coming off as a creep, without looking needy and without saying the same thing as every other chump.

What is the perfect message a man could send to a woman to maximise his chance of blowing her away and creating interest?

The Perfect Message

If this experiment has shown anything, it’s that men online face extreme competition with other men in getting noticed by women, especially the very attractive women.

So, what’s the best possible message a man could send to stand out from the crowd and wow the woman? I decided to write what I thought would be a very good first message and send it to the most attractive woman on OKCupid I could find (after a 3-minute browse).

The message needed to:

• Demonstrate creativity, intelligence and a great sense of humour
• Be totally different to anything she may have received before
• Be obviously unique and not a cut-and-paste job
• Show that I’ve read her profile and absorbed facts about her
• Not be needy!

I ended up writing a very long message that weaved lots of facts from the recipient’s profile into a faux-news script, as if news readers were talking about her live on television. As she read the message, she’d notice more and more clever references to her hobbies, dress sense and so on.

Here’s what I sent.

The Perfect First Online Dating Message

***

And here, happily, is her reply.

Perfect Message Reply

In Summary

The fact that the first stage of online dating is so heavily stacked in women’s favour doesn’t necessarily mean that it’s any easier for them, compared to men, to reach the end goal of pure love or perfect sex. They may have the pick of the bunch to begin with, especially if they happen to be really attractive, but they can still only date one man at a time—they must still filter the largely undifferentiated onslaught of male attention into yes and no piles. Then the yes pile has to be sorted through in much the same way as anyone else does it—by talking, bonding, finding common interests, realising there’s been a big mistake, or a wonderful discovery.

In the end men and women probably do have it about equal, it’s just a bit different for each.

Oh, and if you’re a man, it’s in your best interest to make sure your messages are really well-considered, creatively-constructed and demonstrative of your intelligence, humour and lack of neediness.

Easy. Right?

***

Like this article? Please leave a comment or, better yet, subscribe to my RSS feed to be kept informed of my newest articles and projects. Cheers!

***

Notes and references

I think this experiment roughly demonstrates the differences in the volume of messages women receive, especially attractive women, compared to men. However, it was by no means scientific. For it to have been, it would have needed much more than 10 profiles. You could also argue that it tested the same thing for both sexes (looks), whereas in reality, women mostly judge men on criteria other than how they look. Therefore, perhaps a fairer experiment would be to create a profile for men that advertises the traits in men that women pay most attention to. These would be, according to the studies I’ve read, their job, income and social status.

It should also be noted that once the accounts were relocated in England, I stopped logging into them for the following several months. This means that, over time, other users may have noticed that the accounts were lying a bit dormant and subsequently been put off the idea of sending a message. However, because all accounts would have shown the same “Last Online” date, I don’t think this could have skewed the final message totals in any meaningful way. The women still beat the men.

Finally, I know that certain aspects of this experiment are a bit shady and underhand, but apart from not having any messages replied to, the general OKCupid user base shouldn’t have noticed any discrepancies in their online experience.

That’s it. For now.

References
1. Journal of Communication, Vol. 59, No. 2. (2009), pp. 367–386,
doi:10.1111/j.1460–2466.2009.01420.x

2. Pers Soc Psychol Bull, Vol. 34, No. 8. (2008)

73 Comments

  1. Nick
    July 9, 2012

    Interesting work, although there’s a serious flaw in placing the pairs in different cities. Having used OkCupid in both the Midwest and on the east coast for years, I can tell you that there is a great deal of regional difference in folks’ response rates.

    I don’t know if I missed it in my reading, but I didn’t see which pairs were placed where. New York is by far the most active city on OkC barring maybe San Francisco, so it would have a sort of intense multiplier effect on the message rates of anyone placed there.

    • JonGoddy
      July 9, 2012

      I hear what you’re saying. This is why I moved all accounts to cities in the UK, to help lower the chance that the original population sizes of the US cities would skew the response rates.

      • Name *
        July 9, 2012

        I think it’s less the population sizes than the cultures. Phoenix is a big city, but there’s a lot less OkCupid action there than, say, Austin.

        • JonGoddy
          July 9, 2012

          Yeah, but presumably London is different in culture to Phoenix, so if the same pair received the same relative volume of messages in both cities, their popularity could be reasonably assumed. In other words, moving countries and cities would prevent issues arising from the culture differences between the initial 5 US cities.

          • Simon Jester
            July 31, 2012

            There’s a big discrepancy between the sizes of the English cities (yes, the UK cities are all in England) — Bristol has an urban population of around 600,000, while London has an urban population of over 8,000,000. (Metropolitan area figures are approximately 1,000,000 and 14,000,000, respectively.)

            It might have been better to pick six UK cities that were closer in size to each other.

  2. David Olsen
    July 9, 2012

    Interesting experiment, though I’d like to see which profile ended up in which city as I think that would make a significant difference in the number of responses. Even assuming cultural homogeneity (i.e. the users of each city have the same taste in appearance), the population would be enough to affect the results. For instance, New York is significantly bigger than Miami, thus I’d expect the New York profiles to get far more hits as more people are looking at them (even though the pool itself is bigger; that’s why this is hard). There isn’t a good way to mitigate the difference, but a fast way is to divide the hits by the percentage of the city’s population to the largest city in the sample. Ideally, you’d have these profiles all in the same city, but that increases the chances subjects will notice their identical profiles, contaminating the experiment.

  3. John
    July 11, 2012

    I will sum it up. Online dating is waste of time for men and most women.

    To put it another way, some sap(s) on Match.com just paid 120 dollars for 4 months of rejection and didn’t even get a beer out of it. Thank you, good night.

  4. Liz
    July 12, 2012

    Lovely findings!! Just a comment on improving the sampling (I am a scientist, nerdy much). To account for the differences in cities, you might want to rotate all the profiles through all the cities and compare. I know there is a chance someone might remember that the profile had the same description as another profile, but if you give about a month between, chances are that the pool of daters would have forgotten the specifics. Plus most profiles are fairly generic in nature anyway! You could the take out the effect of different cities and normalise them to some baseline :)

    • JonGoddy
      July 12, 2012

      That’s true — unfortunately, for now my sneaky experiments in online dating are on hold, at least until my next project (the biggest data analysis I’ve done so far) is complete. I moved all the pairs from US cities to UK cities to try to lower the chance of the US cities skewing the results, which it should have done a bit. Although, admittedly, not as much as making sure all profiles spent time in the same cities over the course of the experiment.

  5. As someone who has just signed up to online dating (in the UK), this was a very interesting article. Although it did make me a little sad. I am a female and I have not been flooded with messages. Looking at the photos that you chose to use, I wonder where I would be placed on that sliding scale of attractiveness? Also would using a photo of a non white person significantly alter your results?

    I am actually writing about my own social experiment comparing the results of two different websites. Check out the free ebook at the link below. Would love to know your thoughts.

    Thanks

    • Jay
      August 1, 2012

      Amusing.

      I don’t suppose anyone will find these results surprising. The classic stereotype is that women want men to initiate contact with them and not vice versa, and men accept this role. Your experiment appears to confirm the stereotype.

      It’s interesting to note that the existence of on-line services like this makes such experiments much more practical. In theory, in pre-internet days you could have taken a group of people and asked them to count how many personal approaches or phone calls from potential dating partners they got. But it’s much more ambiguous: If a guy meets a girl in school, he may approach her with the hope of getting a date, but with the option that if she shows no interest he can say he was “just chatting” to save his ego. Or for that matter he may really be just chatting when she thinks he’s hitting on her. Etc.

      All sorts of potential follow-up experiments come to mind. Like, what if you post a picture of a very pretty girl, but with a profile that makes her sound boring, stupid, whatever, versus a not-so-pretty girl with a more interesting profile. Or an actively disturbing profile, like mentioning that she’s just started dating since getting out of prison or rehab. Etc.

  6. AnonymousDog
    July 30, 2012

    I found it interesting that you started out by asking ” Is online dating a different experience for men than it is for women?”, but your conclusion was “In the end, men and women probably do have it about equal, it’s just a bit different for each.”

    Did the question change part way through? I suspect that you knew at the front end, that it’s a different experience for men than it is for women, but for some reason you felt you had to soft pedal exactly how different your results showed it to be.

  7. Jay
    July 31, 2012

    I find it surprising that you didn’t mention the most glaring revelation of your experiment:

    The most attractive of both sets did not receive the most attention– the second most attractive of both sets did.

  8. Carlos Nunez
    July 31, 2012

    This is a great study. Thanks a lot for doing what I thought of doing to much greater detail. I wonder how many profiles on all popular sites are decoys. :)

    OKcupid has done a hell of a lot of experiments and analysis with the massive amount of data they have on tap. Their blog always has an interesting read, albeit depressing ones for some.

    All I know is that I’ve been doing this online dating thing intermittently for a while now (several years) and, as a pretty awesome guy, I have to do the fishing to fill my inbox. Profile and pictures are everything!

  9. Shelly
    August 6, 2012

    An even more insightful “study” would be to do the same thing but use people of similar attractiveness but different ages. I am in my 40’s and I receive virtually no messages. I did the Match thing for six months and sent over 200 messages, all of which were “custom” to the guy I was contacting. I can’t stand cut and paste emails, not to mention they’re obviously cut and paste. I got fewer than 10 responses. In six months. What I have taken away from this whole experience is if you’re female and your age starts with a number equal to or higher than 4 (I’m 45) it is not going to be a great experience. And if, like me, you’re tall (5’11″) it’s going to be even worse.

  10. alexsuspended1
    August 13, 2012

    All very true, however if you’d put that much effort into a message to me (the newsroom bit) I’d think you were a try hard with far too much time on your hands…!

  11. G$
    January 8, 2013

    hmmm. well I’m getting far more messages than any of your men did and I don’t feel like I’m any better looking or have a social status that bares any resemblance to a knight in shining armor although I do have the musician thing on my side… which probably turns more off than on. ( I mean we all know about musicians, right?)

    creating an interesting profile that includes humor and thoughtfulness (know what women like guys, jeez!) creates an atmosphere that allows women to feel comforted that you’re not a creep and boom! here they come. however I’ve also received some pretty disturbing messages from some women who clearly have no control over their “desires”. (“Do you kiss your mother with that mouth?”)

    great experiment. I know my recent ex is simply gorgeous, very photogenic and I frothed at the mouth considering who may be sending her naughty messages but in that instant I also realized something of what you “discovered”… even with all those messages she’s got it worse than I ever will.

Share your thoughts