Don’t Be Ugly By Accident!

August 10th, 2010 by Christian Rudder

If you're anything like me, you usually think of your pics in terms of content: Here's me smiling. Here's me looking tough. Here's me in Hawaii with that wacky turtle. And so on. Today, however, we'll analyze photography from a numerical angle—we'll discuss flash, focus, and aperture instead. We feel like people don't really think about these things when they choose a profile photo, and yet, as we shall see, their misuse can seriously mess you up.

As always, our data comes from dating site OkCupid, one of the largest, and most interesting, datasets on the web. This article aggregates 11.4 million opinions on what makes a great photo.

. . .

Our experiment:

  1. We collected 552,000 example user pictures.
  2. We paired them up and asked people to make snap judgments, like so:
  3. We collated these millions of judgments with the time of day each picture was taken, what the shutter speed was, and so on. Almost all modern cameras embed this stuff in a special header, called EXIF data.
  4. We made graphs.

Here are our findings:

1. Panasonic > Canon > Nikon.

The type and brand of camera you use has a huge effect on how good you look in your pictures. This is a plot of the most popular makes:

As you can see, the general pattern is that more complex cameras take better pictures. Interchangable lens cameras (like digital SLRs) make you look more attractive than your basic point and shoot cameras, and those in turn make you look better than your camera phone. I'm not sure what's going on with Kodak all the way to the right there. They might want to consider making sharing more difficult.

Beyond the advantages or shortcomings of any specific brand, the more-complex-is-better trend bears out at all ages:

And we also found similar numbers looking only at people who uploaded all three types of photos. Putting such a triplet together dramatically illustrates the difference:

oh, also—iPhone users have more sex.

File this under "icebreakers, MacWorld '11". Finally, statistical proof that iPhone users aren't just getting fucked by Apple:

The chart pretty much speaks for itself; I'll just say that the numbers for all three brands are for 30 year-olds, so it's not a matter of older, more experienced people preferring one phone to another. We found this data as part of our general camera-efficacy analysis: we crossed all kinds of user behaviors with the camera models and found we had data on the number of sexual partners for 9,785 people with smart phones. We dropped what we found into Excel, and voila. Here's the plot by age:

Just so you know, the names and the actual photos are removed when we do this kind of research; we just see the stats in aggregate. Everything is anonymized. Now let's leave brands and gadgets aside and look at how purely photographic phenomena can affect your precious face.

2. The flash adds 7 years.

This is another simple finding that needs little explanation.

Soft light can hide wrinkles, blemishes, devil eyes. The hard light of a flash often brings them out. As I illustrate with the dotted lines below, you can calculate the equivalent "aging" effects of a flash by counting years horizontally between the 'flash' and 'no flash' lines. For example, a 28 year-old who used a flash is as attractive as a 35 year-old who didn't. Trace the dotted lines to see what I'm talking about. Don't piss off Ming.

One thing we observed is that most flash exposures—even from SLR's—appeared to be direct flash. That's where the flash was fired directly at the subject, producing harsh shadows. If you have access to a flash that can bounce off the ceiling or walls, that could work much better.

3. Blot out all other reality.

We found that the best pictures have a very shallow depth of field, meaning that the subject is in crisp focus while the rest of the picture is blurry, like this:

I'll spare you my explanation of the optics behind this and instead let a graphic from the 10,000 word wikipedia page fill you in:

Thanks, hivemind, you genius! Basically, you get this sharp/blurry effect from having a wide-open aperture: low f numbers on your camera, like f/1.8, f/2.2, etc. For two pictures taken at the same distance, the lower f number will give you a shallower depth of field.

The widget below plots the aggregate attractiveness, by f number, of our user photos in a little color-coded array, alongside examples of each type of photo, so you can easily see how the depth of field affects things. For obvious reasons, we restricted this analysis to photos by cameras capable of a wide range of apertures.

show women show men

It's my opinion that because the photos with the low f numbers feel more intimate and personal, they get a better viewer response.

4. There are peak times of the day to take a good picture.

Below is a minute-by-minute distribution of when people are taking their pictures. This plot also does a good job of showing off the sheer number of photos we analyzed for this piece:

Of course, the most interesting thing isn't when people are taking their photos, but when they are taking their best photos:

It seems that, broadly speaking, late night and late afternoon are optimal. I can't really say why that is, but I can irresponsibly theorize that photos taken in the former bracket tend to be more provocative, those taken in the latter tend to be pleasantly lit.

As noted, the plotted timestamps are adjusted by time zone and for daylight savings, and when you overlay the path of the sun through the sky during our theoretical "day", you see peaks just after sunrise and just before sunset: evidence of the golden hour.

. . .

In conclusion, the data strongly suggest that if you're single, you (or someone you know) should learn a little bit about photography. Technique can make or break your photograph, and the right decisions can get you more dates.

It's actually not that hard. Use a decent camera. Go easy on the flash. Own the foreground. Take your picture in the afternoon. Then visit the nearest Apple store. Done.

374 Responses to “Don’t Be Ugly By Accident!”

  1. Rich says:

    You should be ashamed of yourself, Christian. This is a horrible failure of an attempt at statistics. You should know better that when there is a correlation, it is very likely to be a matter of common cause, rather than cause and effect. That is very clearly the case here.

    People use their camera phones to take pictures of themselves in their bathroom mirrors. Pictures taken with fancy cameras are likely taken by people with good photography skills, likely on occasions when the subject was well-groomed. Pictures taken with care about depth of field and such things are likewise likely taken by a professional. This does not in any way imply that an average person picking up an expensive camera and taking a picture in their bathroom mirror will produce anything better than their crappy phone camera would.

    Also, you seem to have confused more sex partners with more sex. Those two are not at all related; in fact it could well be the opposite. You’ve only shown that iPhone users are sluttier and/or bigger liars. You’ve also demonstrated once again that Apple fanboys are desperate to rationalize their wasted money.

  2. Tim says:

    Hi guys,

    Would it be possible to have your data for a statistic course?

    Cheers
    t

  3. Cam says:

    More Sex partners = more sex?! hmmmm are you really so sure about this? It would seem that Iphone users are like a bunch of bonobos bouncing around…

  4. Paul says:

    Welll lets look at the obvious, there is such a big hype about iphone that EVERYONE knows of it… But with android, many more havent heard of it and the only people who have are the more technically clever ones as it is better, and generally speaking women wan t the bad asses for sex? So they wont want someone who knows all about technical stuff? That is why you get that.

  5. Anne O'Nymous says:

    It worries me that a match-making site that bases on statistics is making obvious errors in causality, especially regarding complex camera -> good picture. Is that a reasonable assumption? Not really. As many sharp people above have pointed out, complex camera -> more likely to be owned by a photographer -> more likely to be shooting professionally -> good picture.

  6. Jomeo says:

    I think the quality of the camera can effect how it makes u look to some degree & I do believe that standard flashes can put years on u. but I don’t believe one of the optimal times to be photographed is with 1 hour of sunrise.

    The problem being the light is bluer in the morning & redder in the evening due to the red-shift effect. Blue light is harsher & tends to show lines more so it’s good for showing off a good body but bad for face pics. In the evening tho the redder light from the sun is softer & therefore doesn’t show wrinkle & blemishes as sharply.

  7. Truth Sayer says:

    Sooo… iPhone user’s sleep around with more people (more likely to contract STD’s too)… While Android users tend to have less sexual partners (and less likely to have STD’s).

    iPhone bandwagoners seem like things simple and superficial (like the iPhone UI) and stear clear from the depth and intricacy of Android (characteristics of a monogamous relationship).

    Interesting…

  8. me says:

    iPhone users have more sex? That’s not really what the numbers say. They say sexual partners, right? But usually when switching sexual partners, there is some time inbetween where you have much LESS sex than in a steady relationship.
    But of course the other data still supports that the other phones’ cameras are crap.

  9. pjunken says:

    NOKIA gives you an ugly face !!!!
    Sony Ericsson is much better.

  10. Andrew Brooke says:

    I am a photographer and film maker.

    Irrespective of subject (ie beautiful/not so beautiful people) nothing here surprised me.

    What you have here is the rule of thumb guide to taking a picture. Photographers avoid hard light, high sun makes the eye sockets dark and casts hard ugly shadows, unless that is what you want, the trick is move into shadows, use a fill light or diffuse the sunlight but then you have to consider the exposure in relative terms of foreground to subject, in the morning faces are puffy after sleep (allow a couple of hours from slumber before taking a beauty shot), flash needs to be well positioned and manipulated to avoid making reality ugly truths, whichg can be desirable but probably not for a profile picture. Using wide angle lenses on faces, depending on distance is not the most flattering, but the shape of the face will have an influence, it might however be interesting and so, a good photo. The longer lens therefore is the ‘safer’ option to make a more ‘attractive’ photo.

    However there is one set of parameters not discussed and that is the size of the image sensor as this determines the focal length/field of view values, simply: a larger sensor needs not such a wide angle lens to achieve the same field of view as a camera with a small sensor (that’s why they bother with 35mm film / lens value equivalents, though they are not exactly equivalent optically, only in field of view terms). A wide lens (ie low number value for focal length) gives more depth of field and probably distortion through the range of f-stops, whereas a longer (hight number focal length) has inverse properties, less depth of field, less distortion, then there is the distance from subject to lens, the closer the subject (in focus) to the sensor the less depth of field is achieved.

    So: sensor size/film format, distance from subject, lens focal length, f-stop value, light source, light position, light characteristics soft/hard, and camera position are the questions and decisions photographers ask and make very quickly all the time. Its not that hard, its practice.

    Another parameter that also has a big impact is: who is using the camera? Someone who owns a Leica even if its a point n shoot (it might even have a larger sensor) probably has other cameras and is therefore probably a more experienced photographer whereas the the Kodak market is for a far more casual user I assume, owners of DSLR’s are likely to have considered their shot a little more than the Kodak user too.

    There is no right and wrong way, it either looks good or it doesn’t and that is down to the decisions made, so in the end, its the user not the camera.

    You might even like some of my pictures http://www.andrewbrooke.com … taken using anything from point n shoot, to medium format film.

    A.

  11. DrugChicken says:

    Unfortunately, the only thing making that girl attractive in the DSLR shot, versus the camera phone/point & shoot, is the knit cap hiding that unfortunate forehead.

  12. me says:

    What about the non-smartphone-users?
    In “iPhone users have more sex”, the numbers for people belonging to neither of these three groups should also be shown, as a baseline.

  13. manny says:

    dang iphone users are more likely to have STD’s

    and sluts like iphones… not a surprise

  14. paul says:

    this is one of the most RETARDED articles I’ve read in my whole life, while surfing the net. Being a photographer, I couldn’t stop shaking my head. But did smile for a while.

  15. Elizabeth says:

    This is hilarious!!! I’m pretty sure I actually split a gut here. Great research, love the crazy ideas and horrible statistical work which is actually true and provable…just often for other reasons. But seriously- how do I suscribe to your blog??

  16. Inherent Vice says:

    LOL, I love the iPhone/sexual partners discussion.

    So let us make a picture of iPhone users: Statistically while being under age they develop a sexual life that other people get in their twenties. For the next 5 years – while the others develop quite usual – they live monogamic or don´t have any sex at all. In their mid/late twenties they start bitching around (or lie about ist), at the (medial) age of 27 they stop getting more sexual partners and for the next 10 years their sexual lifes becomes as monotonous as the (“nerdy”) Android users in their thirties. In their late 30ies they start changing their sexual partners again (apparently they make out with android users of that age :p )

    Well, it´s absolutely right, that having more sexual partners doesn´t mean having more sex. It absolutely doesn´t mean being good at it, too. But since I´m 30 and used both blackberries and android phones, I´m suddenly very happy having never bought an iPhone

  17. Tim Long says:

    Guys, don’t take all this too seriously. It’s a blog, not a scientific journal. I regard these areticles as slightly tongue-in-cheek but they are interesting and entertaining and I enjoy reading them.

  18. asdf says:

    I´d hit the girls on f/2.2, f/3.5, f/5.6, f/6.3 (<- her absolutely) and f/18. So I prefer pictures with high f numbers? Or do I like pretty brunettes?

  19. Shayne says:

    The correct conclusion is: People who have *friends* with iPhones have more sexual partners. These images aren’t self-portraits.

    Don’t agree? here is a litmus test for any OKCupid user:
    “How many photos do you have of *yourself* on your *own* iPhone?”

    Anyway, Interesting analysis of photos. Thanks & keep the info coming!

  20. Alan Foster says:

    MING THE MERCILESS!!!

  21. Brooks says:

    More sex partners is not necessarily a good thing. At 30, someone could have had a total of 5 sex partners with him he or she shared a LOT of incredible sex, whereas another person could have had a total of 15 sex partners with whom he or she shared 15 shitty, unfulfulling one-night stands…

  22. Brian says:

    lol “statistical”? if that means taking a poll, then the only thing it might prove is that iPhone users are most likely to exaggerate the number of sexual partners they’ve had since the iPhone first came out in 1997. what 40 yr old iPhone user has had 15 different sexual partners in 3 years? brad pitt?

  23. Mat says:

    Most Iphone users class masturbation as sex – Madam palm and her 5 lovely daughters multiplied by 2 = 12 Sexual partners!

  24. ninja says:

    great statistic!

  25. ninja says:

    “Don’t piss off Ming”
    haha : D

  26. Miles says:

    I think my iPhone is broken…

  27. Ron says:

    So you’re saying Bokeh makes a picture look nicer? Novel.

  28. Whitney says:

    I think it’s great that ya’ll have finally made an article about this. As a photographer, there have been times when I’ve wanted to host Profile Picture sessions just so these people have decent pictures to put up that wont make people click right past them. Hopefully now, they might be able to do it themselves!

  29. Andrew says:

    Outstanding!

    Even with all the ways that the cause-effect may be misinterpreted it is very interesting.

    A few observastions and comments

    Panasonic 4/3 winning is very interesting since Panasonic tends to have colours and tones which are not all that realistic. I ultimately chose Canon and Olympus in part because of their faithfullness to colour. Perhaps this is of choosing accuracy over art. Still, if the number one way to share photos is over the web Panasonic appears to be the winner.

    Comparison of photographs at web resolution and then at higher resolution media may yield different results. Canon full-frame vs. Panasonic 4/3 in 19×13 prints may give a different result.

    I find the DOF statistics especially interesting, however, to be more exact you also need to break it out by sensor size. A photograph at f2.8 on an APS-C sensor size gives about the same amount of blur as f4.0 on a full-frame sensor size. This in part explains some of the spread in the data. Of course there is also an additional correlations of lens quality to f-stop and prime vs. zoom. Not all f1.4 lenses are the same and primes tend to have better bokeh than zooms. F1.2 on a FF sensor looks very different than f1.2 on APS-C.

    Thanks for sharing.

  30. Andrew says:

    Am I the only person who wishes that okc didn’t strip the exif of images uploaded to profiles? I mean, obviously if there are gps coordinates those should be stripped for privacy reasons, but is it that horrible to list the date/time/camera?

    Mainly because it’s easy to tell if someone is hiding something if all the pictures are 5 years old.

  31. Mat says:

    It would be interesting to see if the actual position of the sun would yield more pronounced results than just considering timezone adjusted daytime. Keep in mind that 6pm during the winter in a Scandinavian country is a lot different than 6pm during the summer near the equator.

  32. Rod says:

    So I have a Windows Phone. Does that doom me to Celibacy? this article was ridiculous.

  33. Xor says:

    All the little girls with their iPhones are whores ? Big deal, tell us something we don’t know !

  34. Vlad says:

    iPhone users are less discriminating.

  35. Sam says:

    umm every poll I’ve ever seen has women having fewer sex partners than men (heterosexual men lie about the number of people they’ve been with), and that’s in both the median and the mean. And both average way less than 12… What’s going on with the smartphone population?

  36. Anon says:

    Interesting… I know people that use the Leica m9 with a Noctilux lens and people that use the Canon with appropriate lenses and with no doubt they both provide beautiful pictures. However I don’t know anybody that uses the Panasonic… :P

  37. mathguy says:

    I suspect the reason that iphone users have more sexual partners is because of the higher receptiveness of the service, which undoubtedly leads to more 3G and even 4G encounters.

  38. lendra says:

    thks info

  39. MrPentaxian says:

    Bunch of haters who wish they had an iPhone, sounds like. ( I don’t have one, btw) And a few photographers who still can’t get the fact that any given lens’s characteristics do not change relative to sensor size, so depth of field does not change relative to sensor size at any given aperture. Sensor size only determines what portion of a lens’s image circle is used for imaging.

    Oh yes, the iPhone was introduced in 2007, not 1997 as claimed by one of the haters! :-).

  40. Martin says:

    Wow! I’m going to sell my top of the line Pro-Grade Nikon equipment and get a Panasonic so I can make better pictures……….not.

  41. Milkman says:

    All the whining and bashing on here is funny. How old are you all?

    First off there is some truth to taking time to take a better picture. I did online dating for a while just to see what was out there and got “ok” interest from women. As soon as I switched to professional shots the # of winks and emails I got nearly quadrupled. Even though I’m seeing someone I met in person right now, I’m STILL getting emails and winks from women a year later. Take better pictures and you will get noticed more. Period. Whine a bunch and you won’t. It’s your choice to be smart or not… if you’re not smart enough, may your genetics sit stagnant so your great great great great grandchildren are never born and will never intermingle with mine. lol

    Lastly LOL at all the iPhone haters. Calm down, it’s just a phone.

  42. Sjors Provoost says:

    First of all I love the article; I had a great laugh and it’s food for (critical) thought.

    I like the suggestion by Mat to look at the position of the sun in stead of the time. Good luck with the math though; it’s just of matter of combining UTC time and date with latitude and longitude. While you’re at it, you could split the data in sunny and overcast, shade, inside/outside and compas orientation :-)

    There’s a lot of critique about equating the cameraman with the camera. I would be interested to know if that’s the full story. Can you do more research into that?

    The chart with different blurriness uses only “advanced” camera’s, which shows that there is a role for the photographer. Question is how much of variation is caused by the camera and much is caused by photographer.

    You could look at hobby / interest data; do people people consider themselves (good) photographers. If you’re bold: spider their FlickR / Picassa pages for ratings, so you have an objective measure of their photographic skill.

    You also need a way to find out if a photo was taken by the subject or somebody else. Does the Exif info tell you if a self timer was used? Or once again you could crawl their other photos and see if it was the same camera.

    Another angle to look from could be to see what happens when people change camera’s. Especially if they go from high-end to low-end, but that might not happen a lot and it could be because the newe photo is not a self portrait.

  43. Cthulhu says:

    Um, the iPhone came out in 2007 not 1997 (a mere three years ago), Android phones are even more recent. So generally the number of sexual partners (or the rate) was largely set by the time they got their phones.

    It would have nice to seen error bars or standard deviation on those graphs. But the effect sizes do seem quite large.

    As others have, noted, who knows what third variable(s) is/are driving these results but what I find more interesting is the interaction with gender.

    Over all these numbers are quite a bit higher than recent estimates: http://www.kinseyinstitute.org/resources/FAQ.html#number
    but I could see, given the sample, that the numbers might be somewhat higher. Still the female iPhone users are at about 3 times times the national estimate for 30-44 year olds. And all of your data shows females indicating more sexual partners than the males.

    Without digging way deep in the data, it will remain a mystery but certainly makes for an interesting topic of conversation.

    And I like my Android phone!

  44. MagnusMoore says:

    I chalk up the iPhone users having more sex partners to the obvious choice:
    Apple users are whores ;)

    J/k peeps.

    iPhone is a status symbol, so the people that are more likely to own iPhones are more likely to be both shallow and socially active.

  45. oily olly says:

    ok, but owning which camera makes you more attractive to others? I hate the way I look, I want to buy a camera that makes me feel better about myself. and i happen to know i haven’t had ANY sex since i bought my iphone, and i a little pist about that sales pitch!

  46. Tinkererputterer says:

    Hmmm, I’m frugal, my Cellphone was free, a nice one, Verizon, two-way flip Samsung, it’s been serving my purposes well for 2 years – not getting much sex though (I’ll blame it on being an old fossil and working a lot of hours). Young guys who throw $200+ at an IPhone probably also throw money around elsewhere. People who throw money around always attract the ladies. Like throwing birdseed around – Oh look! Here come the chicks! OR The very creative Apple users/IPhone users make up better lies about getting some action! Stats are a funny thing. They can be skewed to accomplish any task. They say that 87% of statistics are made up on the spot, I think it’s closer to 92%. :o)

  47. Eamonn says:

    Guys, can you do some research on baldness? We all know that what women want and what they think they want are often two very different things. I hear almost unanimously that women like guys with shaved heads, but since I shaved my head I’ve found it increasingly harder to get a date. Could it be just because I’m getting older and my pool of potential mates is getting smaller? Or is there some way of analysing photo ratings of follically-challenged guys and finding out what women really think of us?

  48. William says:

    I liked seeing the correlation between camera make/type and perceived attractiveness and also the correlation between f/stop and perceived attractiveness however I noticed that the type of camera that is head and shoulders way better than all the others (Lumix Micro 4/3) comes standard with a f/1.7 lens. Is there a way to normalize the “Photo Attractiveness by Camera Make and Class” for the fact that the photos taken with the Lumix 4/3s were probably all taken at relatively lower f/stop and lower likelihood of using a flash (since lower f/stop allows more light and therefor less need for flash) than other cameras? I’m willing to bet that the gap between the DSLRs and the Lumix 4/3s would be narrowed quite a bit if that was controlled for.

  49. beau says:

    i dont understand… is 12 a lot of sexual partners by 30? thats pitiful.

  50. ren says:

    you have improperly used the word hivemind, but good choice with Ming