And research agrees: studies show that people are uniquely bad at rating their own physical attractiveness.
Well, that’s a relief, that I’m not unusual in being unable to evaluate my own physical attractiveness. (Or am I unusual in being aware that I don’t know how good or bad looking I am?)
One valuable insight from OkCupid is that, “the more men as a group disagree about a woman’s looks, the more they end up liking her”. For what it’s worth, I can add one data point to that as a man. My HotOrNot pictures earned a measly 4.3 rating, which rather hurt my self-image for a while. But in more practical terms, I get asked out multiple times a night at bars. For anyone else who lacks self-esteem about their looks, I’d like to reiterate the usual advice, now backed with OkCupid data: get out there and let other people do the judging of your looks instead of you, because it’s quite likely that someone in a crowd will have eyes for you!
That plus I’d expect a certain amount of sampling bias at HotOrNot. I mean, I could be wrong, but AFAIK it could easily be true that you are in the 43% bracket of HotOrNot (not that I expect their 10-point system actually correlates to this, but anyway...) while still being pretty attractive by real-world mortal human standards.
People’s subjective experience of how attractive someone is is heavily influenced by framing. I can’t find the relevant study but basically people responded with better ratings when someone was surrounded by less attractive people than when someone was surrounded by people who were around the same or more attractive. Conclusion? The same as Mises: preference rankings are ordinal, not cardinal. The frame of hotornot is looking at a very large group, so all but the most attractive in the set will rank slightly worse than they otherwise would have (real life situations are always much smaller sets).
In addition, as the okcupid article indicates, variance matters a lot. 3 people rating you a 9 or 10 and 7 people rating you 1 or 2 means your overall rating will be low, even though a significant fraction of people think you’re the bees knees.
Oh and to quantify: the research I’m familiar with indicates that women should, on average, bump up their estimation of their own attractiveness and men should bump it downward (but a smaller bump than women). But this hides an important dynamic: we don’t care what the average person thinks of us. We care about what people whom we find attractive think. A rating of 8 from someone who we rate an 8 is roughly twelve billion times more important than from someone we rate a 2.
Addtionally, assessments of physical attractiveness are also influenced by assessments of other traits. Suppose you meet someone you think is a 10, but you discover that you cringe every time they open their mouth (to disambiguate, this a reference to the content of their speech, not their dentistry). Not only are you probably not going to want to be with that person, but your physical assessment is going to change. I don’t mean they will suddenly seem ugly, but probably they’ll be a 7 or 8, and you won’t be able to understand how you ever thought they were a 10 in the first place.
Another rating bias: people probably don’t use the rating scale uniformly.
I remember reading somewhere that when women rate men online, instead of “nice, medium, ugly” their rating is more like “nice, ugly, ugly” (the median guy is rated disproportionally low).
If this is true, then this bias could be partially corrected if the web would not display “your rating is 4.3 of 10”, but rather “your rating is higher than 70% of people in the same category”. Or if the site displays global statistics, you can locate yourself in the distribution curve.
You can find that on okCupid’s post about attractiveness. Women rate 80% of guys as worse-looking than medium, whereas male ratings are symmetric and fairly normalized.
I feel a bit ambivalent about that finding (but find the site’s stats really fun to look through).
If male attention is disproportionately directed towards people that the particular guy finds attractive, then it seems possible that women with the same average rating can have different amounts of attention for reasons other than the disagreement. Like, if people who rates someone a 5 gives her 10 attention, a guy rating someone a 4 gives her 5 attention, 3: 3, 2: 1, 1: 0 (entirely made up), then there are different ways of getting a 4. It depends on the ratio of attractiveness rating to attention giving.
A person with an average rating of 4 with people four people giving them a 4-rating gets 20 attention. A person with an average rating of 4 with three 5-ratings and one 1-rating gets 30 attention.
Rather than making up numbers, check out their linear regression model:
msgs = .4m1-.5m2-.1m4+.9m5+k
4s (“eh, cute”) get asked out less. 5s (“hot”) get asked out more. 1s (“weird”) get asked out more, because somebody thought that was a five, and rather than a supermodel who must be inundated with messages, it’s someone quirky whose average rating is only a 3 (and thus approachable).
and rather than a supermodel who must be inundated with messages, it’s someone quirky whose average rating is only a 3 (and thus approachable)
That’s the hypothesis that OkCupid advanced: game-theoretically, it makes sense to go for people you are strongly into who other people aren’t into. But there’s a problem with this hypothesis: it could turn out to be true, but right now, it’s sort of silly.
It’s unnecessary. Look at some normal distributions, and it’s easy to see that having a high variance of attractiveness is sufficient to explain high positive responses (that motivate 5-ratings and messaging) and highly negative responses (that motivate 1-ratings). Let’s say you message a woman with high variance (lots of 5s, lots of 1s) who you consider a five. Maybe that’s because for you, she isn’t actually a 5, she’s a 6! But the scale only goes to 5 inducing a ceiling effect. You are going for her because you are really, really into her (for the same reasons that other guys are really, really not into her), not because you anticipate less competition.
There is no evidence that people are thinking of that sort of game theory. It’s possible, but if men really cared so much about minimizing competition, you’d think they would message women they found 3s and 4s more often.
If you think someone is a 5, even due to high variance traits that other guys hate, you don’t necessarily realize that other guys hate those traits (typical mind fallacy). Instead, you may assume that other guys would be into her just as much as you, which undermines the notion that you are trying to get the women that other guys won’t pursue. This hypothesis gives men too much credit predicting the psychology of other men, and calculating the average appeal of a woman across the whole male population. For instance, I can’t figure out what the guys who give flower-hair-girl a 1-rating are smoking.
Assortive mating. If you message a woman with tattoos and piercings (to use the example in the article), is that because you are thinking “aha! tats and piercings will turn off other guys, so I’ll have her all to myself,” or “wow, I really like that she has tattoos and piercings, and she is probably going to like my tattoos and piercings, too!” This hypothesis isn’t really supported or necessary either, but it helps why men don’t treat all 5s (from their perspective) equally. If someone has tribal markers, it explains why you might both find them a 5 a message them, while you are less likely to message another woman you rate a 5 without tribal markers, and why other guys from other tribes can’t stand women with the affiliations you like.
It’s unnecessary. Look at some normal distributions, and it’s easy to see that having a high variance of attractiveness is sufficient to explain high positive responses (that motivate 5-ratings and messaging) and highly negative responses (that motivate 1-ratings).
Emphasis mine. Is there a difference between this and “quirky”?
I think you were on the right track with the word “quirky.” It was the OkCupid article’s game theoretic hypothesis that I was objected (referenced by avoiding people “inundated with messages” in your comment).
Numbers were made up because people rating someone a 4 don’t give them negative attention (as in intercepting messages), so much as something more like give them less attention than average given their attractiveness level.
It may actually give them negative attention; suppose I don’t message anyone I rate a 4 (I don’t) and by raising their rating I make others less likely to message them (because their average rating is higher). (I thought there was a way to determine another user’s average rating, but I’m not seeing it from a quick check of the site, so this may not be the case.)
To the best of my knowledge, though, the coefficient for m4 and m2 aren’t “relative to m3” but absolute; if someone gets 10 5s, they’re expected to get 9 messages. If they got 9 4s and a 5, they’re expected to get no messages. (Of course, what would be interesting is looking at clusters rather than just linearly regressing the data.)
Well, that’s a relief, that I’m not unusual in being unable to evaluate my own physical attractiveness. (Or am I unusual in being aware that I don’t know how good or bad looking I am?)
One valuable insight from OkCupid is that, “the more men as a group disagree about a woman’s looks, the more they end up liking her”. For what it’s worth, I can add one data point to that as a man. My HotOrNot pictures earned a measly 4.3 rating, which rather hurt my self-image for a while. But in more practical terms, I get asked out multiple times a night at bars. For anyone else who lacks self-esteem about their looks, I’d like to reiterate the usual advice, now backed with OkCupid data: get out there and let other people do the judging of your looks instead of you, because it’s quite likely that someone in a crowd will have eyes for you!
That plus I’d expect a certain amount of sampling bias at HotOrNot. I mean, I could be wrong, but AFAIK it could easily be true that you are in the 43% bracket of HotOrNot (not that I expect their 10-point system actually correlates to this, but anyway...) while still being pretty attractive by real-world mortal human standards.
People’s subjective experience of how attractive someone is is heavily influenced by framing. I can’t find the relevant study but basically people responded with better ratings when someone was surrounded by less attractive people than when someone was surrounded by people who were around the same or more attractive. Conclusion? The same as Mises: preference rankings are ordinal, not cardinal. The frame of hotornot is looking at a very large group, so all but the most attractive in the set will rank slightly worse than they otherwise would have (real life situations are always much smaller sets).
In addition, as the okcupid article indicates, variance matters a lot. 3 people rating you a 9 or 10 and 7 people rating you 1 or 2 means your overall rating will be low, even though a significant fraction of people think you’re the bees knees.
Oh and to quantify: the research I’m familiar with indicates that women should, on average, bump up their estimation of their own attractiveness and men should bump it downward (but a smaller bump than women). But this hides an important dynamic: we don’t care what the average person thinks of us. We care about what people whom we find attractive think. A rating of 8 from someone who we rate an 8 is roughly twelve billion times more important than from someone we rate a 2.
Addtionally, assessments of physical attractiveness are also influenced by assessments of other traits. Suppose you meet someone you think is a 10, but you discover that you cringe every time they open their mouth (to disambiguate, this a reference to the content of their speech, not their dentistry). Not only are you probably not going to want to be with that person, but your physical assessment is going to change. I don’t mean they will suddenly seem ugly, but probably they’ll be a 7 or 8, and you won’t be able to understand how you ever thought they were a 10 in the first place.
Another rating bias: people probably don’t use the rating scale uniformly.
I remember reading somewhere that when women rate men online, instead of “nice, medium, ugly” their rating is more like “nice, ugly, ugly” (the median guy is rated disproportionally low).
If this is true, then this bias could be partially corrected if the web would not display “your rating is 4.3 of 10”, but rather “your rating is higher than 70% of people in the same category”. Or if the site displays global statistics, you can locate yourself in the distribution curve.
You can find that on okCupid’s post about attractiveness. Women rate 80% of guys as worse-looking than medium, whereas male ratings are symmetric and fairly normalized.
I feel a bit ambivalent about that finding (but find the site’s stats really fun to look through).
If male attention is disproportionately directed towards people that the particular guy finds attractive, then it seems possible that women with the same average rating can have different amounts of attention for reasons other than the disagreement. Like, if people who rates someone a 5 gives her 10 attention, a guy rating someone a 4 gives her 5 attention, 3: 3, 2: 1, 1: 0 (entirely made up), then there are different ways of getting a 4. It depends on the ratio of attractiveness rating to attention giving.
A person with an average rating of 4 with people four people giving them a 4-rating gets 20 attention. A person with an average rating of 4 with three 5-ratings and one 1-rating gets 30 attention.
Rather than making up numbers, check out their linear regression model:
msgs = .4m1-.5m2-.1m4+.9m5+k
4s (“eh, cute”) get asked out less. 5s (“hot”) get asked out more. 1s (“weird”) get asked out more, because somebody thought that was a five, and rather than a supermodel who must be inundated with messages, it’s someone quirky whose average rating is only a 3 (and thus approachable).
That’s the hypothesis that OkCupid advanced: game-theoretically, it makes sense to go for people you are strongly into who other people aren’t into. But there’s a problem with this hypothesis: it could turn out to be true, but right now, it’s sort of silly.
It’s unnecessary. Look at some normal distributions, and it’s easy to see that having a high variance of attractiveness is sufficient to explain high positive responses (that motivate 5-ratings and messaging) and highly negative responses (that motivate 1-ratings). Let’s say you message a woman with high variance (lots of 5s, lots of 1s) who you consider a five. Maybe that’s because for you, she isn’t actually a 5, she’s a 6! But the scale only goes to 5 inducing a ceiling effect. You are going for her because you are really, really into her (for the same reasons that other guys are really, really not into her), not because you anticipate less competition.
There is no evidence that people are thinking of that sort of game theory. It’s possible, but if men really cared so much about minimizing competition, you’d think they would message women they found 3s and 4s more often.
If you think someone is a 5, even due to high variance traits that other guys hate, you don’t necessarily realize that other guys hate those traits (typical mind fallacy). Instead, you may assume that other guys would be into her just as much as you, which undermines the notion that you are trying to get the women that other guys won’t pursue. This hypothesis gives men too much credit predicting the psychology of other men, and calculating the average appeal of a woman across the whole male population. For instance, I can’t figure out what the guys who give flower-hair-girl a 1-rating are smoking.
Assortive mating. If you message a woman with tattoos and piercings (to use the example in the article), is that because you are thinking “aha! tats and piercings will turn off other guys, so I’ll have her all to myself,” or “wow, I really like that she has tattoos and piercings, and she is probably going to like my tattoos and piercings, too!” This hypothesis isn’t really supported or necessary either, but it helps why men don’t treat all 5s (from their perspective) equally. If someone has tribal markers, it explains why you might both find them a 5 a message them, while you are less likely to message another woman you rate a 5 without tribal markers, and why other guys from other tribes can’t stand women with the affiliations you like.
Emphasis mine. Is there a difference between this and “quirky”?
I think you were on the right track with the word “quirky.” It was the OkCupid article’s game theoretic hypothesis that I was objected (referenced by avoiding people “inundated with messages” in your comment).
Gotcha. I saw their game theory as justifying why people think quirkiness is (sometimes) attractive, not something people are consciously doing.
Numbers were made up because people rating someone a 4 don’t give them negative attention (as in intercepting messages), so much as something more like give them less attention than average given their attractiveness level.
It may actually give them negative attention; suppose I don’t message anyone I rate a 4 (I don’t) and by raising their rating I make others less likely to message them (because their average rating is higher). (I thought there was a way to determine another user’s average rating, but I’m not seeing it from a quick check of the site, so this may not be the case.)
To the best of my knowledge, though, the coefficient for m4 and m2 aren’t “relative to m3” but absolute; if someone gets 10 5s, they’re expected to get 9 messages. If they got 9 4s and a 5, they’re expected to get no messages. (Of course, what would be interesting is looking at clusters rather than just linearly regressing the data.)
Fair point, that works too.