Last month, Will_Newsome started a thread about OKCupid, one of the major players among online dating sites--especially for the young-and-nerdy set, given their mathematical approach to matching. He opened it up for individual profile evaluation, which occurred, but so did a lot of fruitful meta-discussion about attraction in general and online dating mechanisms in particular. This post is a summary of the parts of that thread which specifically address the practical aspect of good profile editing and critique. (It also incorporates some ideas I had previously but hadn't collected yet.) A little of it is specific to OKCupid, but most of it can be applied to any dating site, and some to dating in general. I've cited points which came from single comments (i.e. not suggested by several people); if I missed one of yours, please comment with a link and I'll add the reference.
On OKTrends
"Wait a minute," I hear experienced OKCers cry. "Why reinvent the wheel of profile analysis? OKCupid already has a blog for just that, and it's called OKTrends."
OKTrends has its merits, but it also has one major flaw. Wei_Dai summed it up well by observing that OKTrends does not make "any effort to distinguish between correlation and causation," citing this post as an example. The reason for that is obvious: the first purpose of OKTrends is to bring traffic to OKCupid. It does this with entertaining content about racy subjects, and rigorous analysis comes (optimistically) second. Of course, datadataeverywhere added, that's exactly the Mythbusters formula. They're both junk food science, but it's also the only look at their data we're going to get, so I'll link a few relevant OKTrends posts in the appropriate sections.
How to Write a Good Profile
Okay, you've created your account and answered a few questions. Now it's time to summarize your whole personality, your appeal, and your worldview in ten little text boxes. Where to begin?
The obvious answer is to reply to the ten profile prompts with your answers to them. Don't fall for it! What you write in your profile, along with your picture, will be the whole sense of yourself you convey to other people. Do your favorite media selections and the fact that you need oxygen, water, food, shelter, and two other obvious things to live constitute 20% of your identity?
Concrete Advice #1: Don't just follow the prompts. Think about what you want to say in your profile, and then fit that into the answers.
Or don't even find a way to fit it into the answers. I've seen excellent profiles which literally ignored the questions and just said what they had to say. But fear not, I won't leave you entirely promptless. There are two goals in writing a good profile:
- Honesty (so as to find people who will actually like you)
- Attractiveness (so that they will realize, upon reading your profile, that they might like you)
We'll address these one at a time, beginning with honesty.
There's a distinction in anthropology between "ancestral traits," whose genes go back so far that they are common among a huge variety of species, and "derived traits," which evolved recently enough to be an informative descriptor of a group. Pentadactyly is an ancestral trait, and is not specific enough to tell a human from a newt; opposable thumbs are a derived trait, and indicate that you're probably (although not necessarily) looking at a primate. You can speak similarly of traits which are memetic rather than genetic; ancestral traits are shared by almost everyone in the culture, and derived traits by smaller subgroups.
Ancestral: "I like listening to music and hanging out with my friends."
Derived: "I like taking photographs and playing board games."
Concrete Advice #2: Write about your derived traits, not your ancestral ones.
Notice that it's not about specificity. The second set of interests isn't very much more specific than the first one. They're just less common interests. Therefore, they do a better job of identifying where you fit in personspace, and in fewer words. For the convenience of newcomers to online dating, here's a quick laundry list of cliches which are so common as to tell the reader nothing about you:
Concrete Advice #3: Omit all of these: "it's hard to summarize myself" "what should I say here" "I'm contradictory" "I'm nice" "I'm shy until you get to know me" "the first thing people notice is my eyes" "I need [obvious literal things] to live" "if it were private I wouldn't write it here" "you can ask me anything" and explicit suggestions that the reader should date you, even tongue-in-cheek
That said, it is hard to summarize yourself. It's hard to recognize the parts of yourself which matter, and even harder to remember them later when you're staring at a form on a webpage. Furthermore, self-identity is susceptible to environmental pressure, and it's easy to just write up the stereotype of the group you feel you belong to. If you'll pardon me quoting myself:
The first few versions of my profile were geared to show off how geeky and smart I was. This connected me to people who spent a lot of time playing tabletop roleplaying games, reading fantasy novels, and making pop culture references to approved geeky television shows, none of which are things which interest me particularly.
Eventually I realized that I am not actually just popped out of the stereotypical modern geek mold, and it was lazy, inaccurate, and ineffective to act like I was. Since then I've started doing the much harder thing of trying to pin down my specific traits and tastes, instead of taking the party line or applying a genre label that lets people assume the details. In that way, OKC has actually been a big force in driving me to understand who I am, what I want, and what really matters to me.
Concrete Advice #4: Learn what you actually care about. Get into the habit of noticing things in your day-to-day life which excite you, please you, infuriate you, or make you think. That's what belongs in an honest description of you.
That's tough, but it's easier than it sounds. Remember that the reason you're being honest is that you want to attract someone who will actually like you, not just the person you claim to be. Don't worry at this stage about appearing "interesting" enough, or whether the generic average airhead represented by OKTrends would like you. Interpolate put it perfectly:
No one you want to meet would find you boring.
Keep that in mind when you're wondering how to balance the honesty and attractiveness goals. Yvain wondered why some users openly express non-mainstream views about transhumanism in a dating profile; this may be honest, but to a lot of people it won't be attractive. Apprentice was surprised by the number of LWers who talked about outdoorsy interests, which can intimidate geeky homebody types. In both cases, whether the interest warrants a mention depends on how significant that interest is to your personality and lifestyle.
Concrete Advice #5: The more you mention something, the more important it will seem to be to you.
rhollerith_dot_com came at the same point from a different angle, with the specific advice not to go into too much detail about work. What field you're in is interesting; what project has been taking up your work hours lately probably isn't. Unless your job is particularly cool or a big part of your identity, it doesn't deserve more than a sentence or two. The same goes for academic fields and most hobbies. If it would only generate conversation with someone who shares your job, major, or hobby, leave it out (unless those are the only people you're looking for). More generally, keep track of how much you mention a given topic in your profile. Count instances, if you have to. When you sort the list by quantity, what matters most to you should be on top. Right below that on the frequency list ...
Concrete Advice #6: Write about the traits or interests that you want a potential partner to share.
Describing what you want in a partner is about as hard as describing yourself, and for the same reasons, but you can approach it the same way (by paying attention and thinking about it in real-life contexts, not just when working on your profile). There are two reasons to make a point of including those things: It will appeal to people who share those traits with you, which is by definition your target audience; and OKCupid connects people in part based on shared interests listed in their profiles, even the ones that the user didn't choose to highlight. More to the point, the adorable but nonsentient cartoon matching robot does that. Which means:
Concrete Advice #7: Do not mention your dislikes in your profile unless they are otherwise important.
As far as I can tell, once OKC has decided you like something, there's no way to explicitly tell it you don't. Even removing it from your profile doesn't kick in immediately. If someone searches for, say, "scientology," and you put in your profile that "scientology is crap," you will come up on the search. This is not what either of you is trying to accomplish. Besides, that doesn't describe you. If you're an active organizer of major scientology protests and are looking for someone to do that with you, okay, put it in. Short of that, don't give yourself keywords you don't want.
One last thing about searchability before we move on.
Concrete Advice #8: Fill out any applicable sidebar information.
Alicorn's example was religion: If you like the idea of being found by an atheist looking for another atheist, make sure OKCupid knows that you are one. I would go a step further and recommend filling in as much as you can. Single completed fields, or single omitted fields, will look more significant than they probably are--but do leave out any where all possible responses would be misleading. (I've left the "children" field blank, for example, because I don't want them now but might some day, so neither "wants" nor "doesn't want" is correct.) If you want to expound on any of your answers, of course, you can do it in the profile body, as long as it maintains an acceptable importance/frequency ratio and doesn't make your profile unreadably long.
Concrete Advice #9: Write between 50 and 350 words in most of the fields.
I got these numbers by measuring answers which make my eyes glaze over (on the long end) or which made me think "that's it?" (on the short end). This isn't a hard-and-fast rule. The self-summary is justified in being a little bit longer; the six things are justified in being shorter. Your favorites section should be one of your shorter answers, unless media and food happen to be really important to you (in which case, write about why, don't just list them).
Last but not least, here is the most-discussed and hopefully most obvious thing you can do to improve your profile.
Concrete Advice #10: Upload at least one clear, flattering, decent resolution photo of yourself. No excuses.
I'm just going to hand it over to mattnewport for a sec, responding to comments about not being "photogenic."
... the word 'photogenic' should be like a red flag to a rationalist bull ... people who are 'not photogenic' are not made of some different type of material that reacts differently to light than photogenic people.
He goes on to point out that OKTrends did not one but two posts on what makes a good (read: message-attracting) profile picture. The first one is about content (poses, props, situations), and the second one is mostly about camera choice and timing. If you can read those and then turn around and take a good photo of yourself, great. If not, and especially if you're frustrated by the task, enlist the help of an actual photographer. You may know one. One of your friends may know one. A local skilled amateur may be willing to trade prints for practice. Whoever they are, find them. If you claim to be trying to prepare a good profile, and you don't have a picture on it that you're proud of, you're fooling yourself. (Hypocrisy alert: I haven't yet done this. But I just talked myself into it, so I will.)
Yvain defends, quite fairly, that all of his photos are of him out doing interesting things which don't lend themselves to clean sparkling images: backpacking, scuba diving, and so forth. He's right to want to keep those to show off his activities; however, four different people commented that his pictures could be improved. I think it's clear that he would be well-served by adding one more, whose sole purpose is to flatter him physically.
How to Make It a Better Profile
Congratulations! You've written a competent profile. But the only person who's seen it yet is the least objective person in the world with regard to your attractiveness. Time to get a second opinion. The purpose of the profile critique is to verify that you've met your two goals in profile writing: honesty (have you actually depicted your personality?) and attractiveness (does the profile encourage messages?).
The best people to judge your profile's honesty are those who know you well. They're the only ones who can tell whether the words you chose give an impression of you which matches the impression you give in reality. Unfortunately, this means they also have preconceptions about you. Better would be a critique from someone who formed their in-person impression only after reading your profile, but if your profile is working that well it's probably fine. In any case, ask your honesty evaluators if there's anything in your profile which surprises them, or anything they're surprised you omitted.
There are two schools of thought on whom you should ask to judge your profile's attractiveness. One is to ask the sort of person you're trying to attract: members of your preferred gender, and probably of your own culture. They can tell you whether your profile is attractive to them and whether they'd message you based on it ... or at least, whether they think they would. The other school of thought is that the right people to ask are those who share your gender/culture preference, and have been successful attracting such partners. They can tell you what has empirically worked for them and compare notes. Both have potential biases, but anything both types of critic agree on is probably correct. (I didn't see any gay users pipe up in this part of the conversation, but I'd love to know how the overlap between the two sets affects their feedback.)
Of course, a once-over by a relative stranger (e.g. another LWer) can be useful as well. They can tell you what assumptions they make about you, knowing little more than what you've chosen to write. Have your critic read the profile line by line and write down their impressions as they have them; when they finish, they can add the overall gist they got from reading. The idea is to give you a fuller picture of the reader's immediate responses--ideas which could stick in the subconscious even if they're forgotten consciously by the end. These are the details that they're filling in between the lines, and that's what you want to be sure is accurate. In particular, this is good for ensuring that your frequency of mentions actually matches your degree of interest; whpearson noticed such a discrepancy in mine, which I corrected.
It should go without saying that any profile editor should also be encouraged to report problems with the language or flow. Get rid of typos, clean up the grammar. Check for subtler things as well, like unusual words repeated close together, or using the same sentence structure over and over. If a joke isn't funny or a reference doesn't make sense, replace or omit it. All of these errors are distractions from what you're trying to communicate, and produce fleeting impressions of confusion or irritation which are then associated with your profile. Other than that, write in a style which is natural to you. That style is a fair part of your self-description.
Finally, review your profile from time to time. Every few months is a good minimum, give or take any life-altering events. The purpose of this is to ensure that your profile changes as you change, to stay up-to-date on the honesty goal. For the same reason, cycle in a new picture periodically, especially when your appearance has changed. If you really want to be thorough, re-answer old match questions from time to time as well. They're the biggest part of how OKCupid connects you to other people, and updating them keeps it current on your tastes and values. That this requires continuing to think about and adjust your tastes and values as time passes is just a perk.
One of the best examples is, of course, PUA communities.
On the main forum I used to post on, here are some of the norms:
In the Techniques forum, you aren't supposed to post any technique until you've tried it a couple times yourself. Furthermore, you aren't even supposed to ask if something might work, instead you are told "go out and try it, then come tell us if it worked." Even talking about an idea that hasn't been tried can privilege the hypothesis too much. Of course, the fact that you've tried something doesn't prove that it works (maybe something else you were doing caused the result instead), or that it generalizes to other people and situations, but it makes the hypothesis worth talking about.
Users are discouraged from posting on subjects they aren't experienced about; doing so is called "Keyboard Jockeying" (aka "KJing"). There is some tolerance for speculation as long as you are clear that you are speculating (some guys will preface ideas with "I'm gonna KJ a little here..."), but too much KJing, even admitted KJing, is frowned upon.
Building up "street cred" is important. Users are encouraged to describe their experience, skill levels, and results. One way to do some is to relate anecdotes from your own experience, or to write "Field Reports," stories of real life interactions. The purpose of field reports is to get feedback, and also to display experience, so other guys can contextualize why you make the claims you make. FRs are of course biased, hyped, selective, and sometimes false, but even they are much better than knowing nothing about the people making claims. At least, you have an idea of the upper bound of how good they are.
It's considered acceptable to challenge people for not knowing what they are talking about. Especially boastful or implausible FRs do get challenged, especially when they are by people trying to sell products. Making bold claims without street cred is frowned on. All the time you will see guys coming on to the forum, writing about how awesome they are, and expounding on some new theories. The more skill they claim without detail, and the more controversy they make, the more likely they are to get a response like, "Dude, what you say sounds cool, but who the hell are you? Post some FRs so that people can know where you're coming from and get back to us."
The "street cred" norms do lead to a lot of ego wars among at least moderately skilled guys, but it helps prevent completely unexperienced guys from thinking they are qualified to argue with guys they should be shutting up and listening to. Could the less experienced guy be right, and the more experienced guy be wrong? Yes, of course. But since knowledge is correlated with success, the advice of the more successful guys gets higher priors.
The weakness of the community is that it's vulnerable to all sorts of biases: availability heuristic, confirmation bias, hasty generalization, bandwagon effects, etc... But it's no more vulnerable to these things than any other practical community that can't do scientific testing. At least with PUAs, you have a better idea of why people think what they think, and it's acceptable and encouraged to challenge people and say, "what's your field experience?". Even though the people who are accepted as having "street cred" have biases, these guys are generally successful, and they are less wrong than the keyboard jockeys on average about what works.
If I discovered some things that contradicted conventional PUA teachings, and talked about it on a PUA forum, I believe I could get people to rethink things if I did so through the appropriate channels. I would write a few field reports, answer some newbie questions, and do so in a way that reveals my skill and experience levels. Once I did so, I would have proved myself epistemically trustworthy to PUAs, and they would be open to unorthodox ideas as long as I backed them up with field experience, and explained them in the context of existing PUA theories. Sometime I will actually perform this experiment, but I am still gathering field experience.
Relative to a scientific community that can perform controlled empirical trials with large samples sizes, the epistemic standards of PUAs are weak. But the epistemic standards of PUAs are much, much better than any other form of dating advice, which is why the community is the gold standard.
Lots of mainstream dating advice "experts" have only have their own experiences to go on. Instead of giving any charismatic person who claims to be a successful dater a book deal and TV time, imagine if you took thousands of successful daters and aggregated their experiences (some PUA forums have tens of thousands of members). Even though this aggregation is biased by ideological bandwagon effects, and by the fact that the community self-selects for successful daters who agree with it, it's still way more epistemically trustworthy (about what works, not necessarily about what is true) than most mainstream dating "experts."
An illustrative anecdote is that "Mystery" admits that he is actually quite bad at attracting "7s". Sure, that's hyperbole for the purpose of showing off, but there is an element of truth behind it. If he wasn't able to take on board knowledge from people other than himself and extracting the general insights then his advice would be qu... (read more)