Do Women Like Assholes?

Cross-posted from Putanumonit where a lively discussion is already going.

Thank you to Ben Pace for helping migrate this post to LessWrong.


Insofar as Putanumonit promotes a normative stance, it boils down to the following:

  1. Be savvy about statistics and research

  2. Be nice and cooperative

  3. Saying “sex is cool, but have you considered…” is cool, but have you considered having sex?

These are not unrelated. Being savvy with math can help your romantic life. Being nice and cooperative can really help your romantic life. At least, that’s what I believe based on my experience.

Some readers enjoy my posts and profitably employ my philosophy and then write me lovely messages about it. They are the reason why I get up in the morning and put so much time and effort into this blog.

But some readers don’t enjoy my posts. They tell me that I’m a fool or a liar, that women date jerks and disdain nice guys, that the gender wars are real and must be fought ruthlessly, that all this talk about win-win romance and compatible goals is a blue pill conspiracy to oppress men.

I’ve been mostly ignoring and mocking these latter readers. But recently, they started posting links to research papers supposedly proving their point. And so, in the name of stance #1, I got up in the morning and put way too much time and effort into my own research project to investigate: do assholes really do better romantically, or is there hope for men and women to get along after all?

Literature Review

The literature sucks. That’s it, that’s the review.

Granted, this question is hard to measure empirically. It’s hard to define who is an asshole, let alone to identify them, let alone to measure how well they do with women in the long run — I struggled with those issues in my own research. But that’s far from the only problem.

The study I was sent most often is The Dark Triad personality: Attractiveness to women by Carter et al (2013). The dark triad is the combination of narcissism (entitlement, grandiose self-image), psychopathy (callousness, lack of empathy), and Machiavellianism (insincere manipulation). The attractiveness was measured by asking women in an online questionnaire to read descriptions of men and say how attractive they find them. The women are 128 college undergrads in psychology.

The result of the study was a positive but statistically insignificant boost to the attractiveness of the dark triad descriptions. The DT guys were rated significantly higher on extraversion (which is attractive) and significantly lower on neuroticism (which isn’t). This would seem to imply that the dark triad isn’t attractive in itself, but only in what it signals about extraversion and neuroticism. Yet, somehow, the authors threw those three into a structural equation model (while conveniently ignoring other confounders like agreeableness) and squeezed out the requisite p-value to get published:

SEMs are a legitimate tool of social science research, but they’re impossible to replicate without access to the data and are rife with opportunities for multiplicity and p-hacking. I don’t know if this study shows anything at all about dark triad and attractiveness. Even if it does, I’m not sure for whom it shows this effect:

Here are a couple of other studies I looked at:

Basically, all the studies in this field use 19-year-old girls on college campuses. Not only are they WEIRD, which is a problem for a lot of psychology research, but 19-year-old women in college are in an extreme and unusual mating situation.

With this, I stopped searching for more papers based on hungover college students with one exception I’ll get to later. I instead listened to six hours of dating podcasts with Geoffrey Miller, David Buss, and Tucker Max. Miller wrote books about how sexual selection shaped our evolution. Buss wrote books about how evolution shapes our sexual selections. Max wrote books about being an asshole and getting laid. If anyone would know whether women prefer jerks and why, it’s those guys.

Everyone who believes that women like jerks is convinced that they know why, but of course they each have their own story. I’ve compiled a laundry list of hypotheses on the subject, based on the literature, the experts, and people I know.

Why may women prefer assholes?

Hypothesis 1 — Signal of extraversion and assertiveness

Women strongly prefer men who are extraverted and assertive to those who are socially passive. It could be that social dominance leads to social and professional success for the man making them a desirable partner, or that outgoing and decisive men simply make better lovers.

1a Being an asshole is, in fact, positively correlated with assertiveness and extraversion and is thus a signal of those traits.

1b Being an asshole isn’t correlated but is mistaken for assertiveness and extraversion. For example, someone may strive to be the center of attention because they are socially skilled and popular or because they are narcissistic, and it’s hard to immediately tell which is which.

1c Being an asshole is a signal of high status or skills, because a loser could not get away with being a jerk. A weak and unpopular man would get laughed at for narcissistic delusions or beaten up for acting like a psychopath. Thus, exhibiting dark triad traits is a signal that one is not a loser.

1d A corollary to the “asshole as signal” theories is that women will fall for assholes less as they grow in experience and wisdom. This is the main reason why studying this on 19-year-olds may be useless: women at 19 don’t have the experience to read men’s status and personality well. Moreover, men sharing a college campus at 19 are very undifferentiated, unlike later in life when women can look at stronger signals like career success.

When I was 19 I was nice and considerate and didn’t get laid a lot with 19-year-olds. Now that I’m 33 I’m trying to be nice and considerate and I’m happily married and having threesomes with smart and lovely women my age. A few of the women I asked admitted to falling for jerks who mistreated them while in college, and how they learned from that experience to recognize assholes and avoid them.

Hypothesis 2 — Short-term mating strategy

Assholes aggressively seek out short-term mating: more casual sex, less lasting relationships. They are more successful at it mostly because of their single-minded pursuit of it. The downsides of dating assholes only emerge in the long term, when the Machiavellian’s lies can’t be sustained or the narcissist’s volatile self-esteem swings from peak to nadir.

2a Women don’t like assholes but sleep with them because of selection effects — only psychopaths approach women aggressively at bars and clubs and it works because of the law of large numbers. This would be kinda funny if true because the sort of guy who posts links to research papers on blogs is almost certainly the sort of guy who will never do well in a bar or nightclub no matter what personality he adopts.

I’m the sort of guy who writes research posts, and none of the women I ever dated or slept with were met in a bar or nightclub. I mostly get dates through friends or my online presence, two areas that I built up through years of long-term-oriented effort. In the club nothing matters beyond the next 30 seconds.

2b Some women just want a guy for short-term mating and are choosing the assholes consciously because they know these men will not want to hang around. What kind of women?

One trope that comes up often is women who have bad relationships with their fathers date jerks. My evo psych take on it is that in the absence of a role model for good fatherhood, women take the good genes in the good genes-good father tradeoff. Tucker Max’s take is that “some girls need to work through the trauma of their daddy issues on some asshole’s dick, and there’s nothing wrong with that”. Either way, I regret not asking about it on the survey.

Hypothesis 3 — Being an asshole is just better

The final option is that being an asshole is not a signal or a correlate of anything, it just works better for romantic and sexual success.

3a Assholes successfully manipulate women into sleeping with them and staying with them with their dark skills.

3b No manipulation needed — women just consciously prefer to date jerks and be mistreated.

Hypothesis 4 — Women actually don’t prefer assholes

But some people think that they do because:

4a — They’re misogynists and want an excuse to be mean to women.

4b — Instead of simply being nice they’re being Nice Guys (TM) who objectify women and treat relationships as transactional.

4c — They confuse being high status among men (which is obviously attractive) to being high status relative to your partner. The latter would imply that belittling (negging) and undermining your partner to lower their status would be a successful strategy.

4d — They’re neophyte PUAs who measure success by getting numbers at bars, and scared women readily give a fake number to pushy psychopaths.

4e — They assume that guys with different norms around flirting (e.g., working-class people, or the French) are assholes, when in fact they’re just more direct (which women like).

4f — They derive the causation backwards, judging men who talk about their own romantic success to be assholes because they talk about it (to less successful men).

The hypotheses in this group are outside the scope of this research, but they’re worth mentioning. Even if women don’t prefer assholes at all, there are many reasons why this trope could flourish.

Study Setup

Participants

My survey on personality and relationships received 1,220 responses. Thank you to everyone who filled it out, and huge thanks to everyone who shared, retweeted, reddited, and told their mom. Thanks for nothing to the 8 people whose responses I threw out for being nonsensical and fucking up the attention check questions. This is a huge sample, larger than in any academic paper I looked at, and quite varied. I’m really grateful.

The median age is 29, with 90% of respondents between the ages of 21-45. We’re talking about adults who are looking to date, not college freshmen looking for course credit.

801 of the respondents are male and straight. 256 are female and either straight or bi, i.e. the mating target for straight men. Given the core question, the bulk of my study focuses on these two groups and I will mostly use men and women as shorthand to refer to them. I’ll discuss some findings that relate to everyone else separately.

Personality Variables

The survey estimated 6 personality traits using 4 questions each (you can review all the questions on the survey itself). Narcissism, psychopathy, and Machiavellianism with questions taken from this paper, agreeableness and extraversion with questions from here, assertiveness from here.

The first three are collectively referred to as the Dark Triad. By subtracting Dark Triad from agreeableness I get a measure of niceness. Henceforth, nice guys are those high in agreeableness and low on the DT traits, while assholes are the opposite.

Assertiveness is often considered a sub-trait of extraversion, and the two showed up very similarly on the survey. They correlate highly with each other and have the same correlations with other traits. Given this, I sometimes combined both into a single measure I called social dominance for lack of a better term. Dominant people are decisive, talkative, like attention. Passive people are the opposite.

I also asked people to rate themselves on physical attractiveness (hotness), attractive talents, and popularity. My responders are slightly hotter than average according to themselves, and hotter the more cisgendered they are without any difference between men and women.

Relationships Variables

I asked people for their lifetime number of sexual partners, current relationship status, and percent of their adult life that they’ve been in a relationship. I also asked what they’re looking for, which I operationalized as a numeric scale for short-term orientation: 3 for those looking for sex, 2 for casual dating, 1 for serious relationships, and 0 for those not looking for any more partners (14% of this latter group are single).

Here is the correlation matrix for all the raw variables measured, it does not look particularly different when broken down by gender.

We see some interesting things right away. Narcissism is correlated with attractive traits, but so is agreeableness. As people get older they become less narcissistic and more assertive. Extraversion is great for both friends and lovers. Of course, many of these traits confound each other so we’ll use regressions and controls to tease out the effects.

The direct measure of short-term mating success is the number of lifetime partners, but we’d expect that to correlate with age in a non-linear fashion. To control for age I pulled data from the giant National Survey of Family Growth to derive the average number of partners for each age bracket. This is shown in the black line in the chart below (with the dots being my actual respondents), going up from 1 to 9 partners over people’s teens and twenties and topping out at 12 partners. Note the log scale of the Y-axis, modified to include those reporting 0 partners.

On the NSFG men report a lot more partners than women (15 vs. 8 by age 40), as common wisdom would suggest. In my survey women actually reported more partners (12 vs. 10), especially bi women. Gay men reported slightly fewer partners (but they are 6 years younger on average than straight men in my sample), lesbians the least, queers the most (despite lower self-rated hotness). By and large, my data seems at least as trustworthy as the NSFG.

My ultimate metric for short-term mating success is log(N partners + 1 /​ expected N partners for age). The log scale makes intuitive sense since finding your first partner is about as hard as the next two, or going from N to 2N. Using a log scale prevents it from being overly skewed by outliers reporting hundreds of partners. So a virgin at 32 (expected N is 10) scores −3.3, while someone with 99 partners at that age scores +3.3 on short-term mating.

For long-term success, I wanted to combine the questions on current relationship status and overall percent of time in relationships. Looking at a bunch of data like this, it seems that married people should expect 20 more years of marriage and single people should expect to stay unmarried for another 12. I decided to err on the conservative side and just add 15 years of “being in a serious relationship” to those currently in one for the purpose of calculating % of time romantically engaged. So a 33-year who spent half the time since age 18 in a relationship (7.5 years) but is now in a serious one will have that metric upgraded to 75%, since I assume their next 15 will be in a relationship as well.

Regressions and reporting

We’re 2,400 words in and I haven’t told you what the mainline finding is or mentioned p-values. That’s because p-values are a perversion of science, and reporting headline results out of context is a perversion of science reporting.

Instead, I’ll post a lot of regression tables (which you can derive the p-value from if you’re kinky), a lot of colorful charts (all clickable for a larger version), and precise results like the 20% nicest men are slightly likelier to be virgins (13.5% of them) than the 20% least nice ones (11%). My goal is to showcase the data first, not to argue a particular narrative.

Results for Straight Men

Here’s the regression of short-term mating success on all the personality and attractiveness variables. All the variables except for age have been normalized to have the same sample standard deviation so that their coefficients can be compared directly.

Age-Adjusted Number of Sexual Partners (on a log scale) R2 = 0.24
VariableCoefficient (SD)
Narcissism-.09 (.05)
Psychopathy.11 (.06)
Machiavellianism.21 (.06)
Agreeableness-.03 (.07)
Assertiveness.15 (.06)
Extraversion.23 (.07)
Physical attractiveness (self-rated).23 (.05)
Attractive talents.14 (.06)
Popularity.27 (.06)

No big surprises here: men who are popular, good looking, and extraverted have more sexual partners. On the nice-asshole axis, assholes do have more partners mostly due to Machiavellianism. Let’s dig into this.

Hypothesis 1 – Asshole as signal

Narcissism and agreeableness are the strongest predictors of social dominance (sum of assertiveness and extraversion), accounting for 25% of the variance in it. You can see on the chart that the bright red dots (for socially dominant people) are concentrated towards the top right corner of those high in both agreeableness and narcissism.

These two traits are also correlated with popularity, but once we control for social dominance the effect of narcissism is cut in half while the effect of agreeableness remains.

Popularity R2 = 0.30
VariableCoefficient (SD)
Narcissism .11 (.04)
Psychopathy-.08 (.04)
Machiavellianism 0
Agreeableness .22 (.04)
Social Dominance .36 (.04)

So narcissism is close to assertiveness and extraversion and is some signal of popularity (i.e. social status). Narcissism is also the only personality trait that positively predicts short-term orientation, i.e. reporting that you’re looking for sex or something casual and not a serious relationship (.13 coefficient with .04 SD). And yet, narcissists are not getting laid.

This matches the story I told in Go Fuck Someone. Narcissists want to be fuckable more than want to fuck. They put all their effort into preserving their image and status, while getting intimate involves vulnerability and making room for the other person’s story. Narcissism is also the only personality trait that strongly predicts caring about one’s partner being good looking — a trait that’s more important for making impression on observers than for building relationships.

Agreeableness (measured as empathy, willingness to help, putting others at ease) is an even stronger predictor of social dominance and popularity, while Machiavellianism has no correlation with them and psychopaths are unpopular introverts. The latter are the two asshole traits that actually contribute to getting laid. So insofar as being an asshole helps, it is not through signaling status or extraversion.

Hypothesis 2 – Assholes (and some women) just want to bang and ghost

As mentioned, narcissism is the only trait that predicts short-term orientation for single men. For women, short-term orientation is basically predicted only by age — older women want more serious relationships.

However, women are less short-term oriented in general. Despite being slightly younger in my sample, 65% of single women report looking for a serious relationship (55% of men) and only 7% are looking for mere sex (12% of men). As I mentioned when discussing gender ratios, this is not a huge difference but it’s important on the margin. For every two men looking for a one night stand (and those are likely the guys driving the number-of-partners metric), there is just one woman seeking the same.

62% of women who look for serious relationships answered that it’s very (5/​5) important that their partners share their relationship goals. 45% of men don’t actually share those relationship goals, but would still like to bang those women.

Here are the four questions I used to assess Machiavellianism, which is a predictor of short-term mating success:

I have used deceit or lied to get my way

I tend to exploit others towards my own end

I have used flattery to get my way

I tend to manipulate others to get my way

Women don’t report seeking out assholes in any way — “nice and considerate” was a close second to “shares my goals” among the traits that are important to women in a partner, ahead of “happy and confident”, “physically attractive”, and “assertive and dominant”. This rules out hypothesis 3b (if women liked jerks, why would they lie about it?) and leads us to:

Hypothesis 3a — Fuckbois

My data, as well as the entirety of this horrible subreddit, seem to point to some number of Machiavellian dudes successfully manipulating women to get laid. For example by lying about their relationship intentions. Machiavellianism (along with psychopathy) is in fact negatively correlated with caring about your partner sharing your relationship goals — they only care about getting what they want themselves.

However, successful manipulation is not the only possibility. Machiavellianism and sexuality: on the moderating role of biological sex by McHoskey (2001) looks at the relationship between, well, Machiavellianism and sexuality. Machiavellianism correlates with psychopathy and extraversion (replicated in my data), sexual success, and also with promiscuity, curiosity, and excitement about sex. Machiavellians are also more likely to feign love, get someone drunk, and coerce someone into sex.

So there are three reasons why Machiavellians could be having more sexual partners:

  1. Coercion and manipulation.

  2. Correlation with extraversion, which gets you laid.

  3. Promiscuity and seeking out sex — if you seek you shall find.

If the first reason was the main one, it’s likely that Machiavellianism would correlate in particular with the number of partners but not the longevity of relationships. Once the lies come to light the Machiavellian fuckboi would have to move on to their next victim. We should see this in a negative relationship with long-term relationship success.

Serious Relationship Success R2 = 0.31
VariableCoefficient (SD)
Narcissism0
Psychopathy.03 (.06)
Machiavellianism.10 (.06)
Agreeableness.06 (.06)
Assertiveness.11 (.06)
Extraversion.02 (.07)
Physical attractiveness (self-rated).18 (.05)
Attractive talents.14 (.06)
Popularity.09 (.06)
Age.05 (.005)

In fact, Machiavellianism has a weak but positive impact on serious relationship success. This still holds if we look at both components of long-term success separately, the percent of adult life spent in serious relationships and being in one right now. This could be an artifact of noise, but it’s likely that there’s at least some weak effect there which provides some evidence against the idea that the success of Machiavellians is entirely due to nefarious tactics.

The data also goes against the “signaling extraversion” hypothesis, since neither including nor removing extraversion from the regression has any effect on the coefficient of Machiavellianism. We are left with the story that Machiavellians are simply more promiscuous.

Machiavellians in my sample don’t show any unusual preferences for casual sex over serious relationships, although that’s not quite the same as promiscuity and excitement. They could just be more relationship-seeking overall, or they get a woman drunk for a one-night stand and then catch feelings by accident and end up a decade later married with three kids and a golden retriever while also cheating on the side. Many such cases, as they say.

Other than that, what’s the secret to finding a serious relationship? Be hot, be funny, be assertive, be patient.

30% of men below age 30 report never having been in a serious relationship, but only 2 out of the 128 men over the age of 40 report the same. A lot of my readers are right at the cusp of that age transition — I hope you don’t stop reading Putanumonit once you find girlfriends and wives!

Summary of Results for Straight Men

  1. Looks, popularity, and social dominance (assertiveness + extraversion) get you laid, with neither factor dominating the others.

  2. Machiavellianism predicts sexual and romantic success. It’s unclear if this is due to successful manipulation or simply seeking out sex and romance more.

  3. Narcissists want casual sex with hot partners and predictably fail to obtain it.

  4. Agreeableness beats psychopathy for both friends and romantic relationships.

  5. Women don’t seem to consciously seek out assholes.

  6. Insofar as assholes are successful, it has little to do with status and their success doesn’t diminish with age.

  7. There’s a huge variance in the number of sexual partners for men of all ages, but almost all men end up in romantic relationships in their thirties.

Other Results

Below is a grab bag of other results that showed up in the data. Some of them fit what I would have predicted and some were surprising, but take them all with a pinch of salt since they were not the original object of the study.

We seek partners like us

Attractive people care more about their partner’s attractiveness, nice people care about their partner being nice, assertive people care more about their partner being happy and confident (although they don’t care about their partner being assertive). All of those relationships are significant and hold for both men and women. This should serve as a word of caution for those looking to be assholes as a romantic strategy — you may end up dating assholes yourself.

You can imagine virtuous and vicious cycles as a result of this. I was always nice and considerate, and it didn’t work until I figured out how to filter for women who are themselves lovely and kind. Now my partners and I can all be nice to each other and enjoy life. If you start off being a jerk you attract jerks, and this further justifies being mean and perpetuating the cycle.

Attractiveness matters for women only in the short term

A woman’s self-rated attractiveness predicts her number of sexual partners, but not her success at being in serious romantic relationships. The latter is correlated with assertiveness and agreeableness, and of course with age. This matches the preferences reported by men: guys who look for casual sex care more about a partner’s looks than those who look for serious relationships.

Narcissism also correlates with women’s short-term mating success but not serious relationships. I talked about it when discussing women’s mating markets. Hot young women have their choice of short-term partners, and they don’t pay much of a penalty for narcissism or disagreeable political stances like #KillAllMen. But they can remain in the mindset that a relationship is something they deserve for who they are instead of something they have to build and compromise for. If that’s you:

Perhaps you are going on dates with lovely people but the dates aren’t going exactly according to the script you envisioned. Or the people who flirt and match with you are not quite what someone with your degrees and BMI and yoga skill deserves. In this case you should go back to self-development: fix your narcissism and figure out what value you actually provide to a romantic partner besides imagining that you raise their status through mere association.

How to tell if you’re in the latter category? If you get a lot of “I can’t believe a great guy/​gal like you can’t find a girlfriend/​boyfriend” from your friends, that’s a sign. Your friends saying that is not a compliment, it’s a mockery of your misguided self-focus.

The opposite is true for gay men

The only trait that contributes to short-term success for the 122 gay and bisexual men in my sample is agreeableness (.59 coefficient with .19 SD, p=.003 without correcting for multiplicity). The only trait that correlates with long-term success aside from age is hotness (.09 coefficient with .03 SD, p=.006). I have no theories about this result or much confidence in it despite the statistical significance.

Personality and gender

Cis men are more psychopathic, disagreeable, and assertive. Women (queer and cis) are more narcissistic. Queer men (N=16) are meek sweethearts. This seems mostly in line with prevailing stereotypes.

True self-confidence comes with age

Personality predictors of age R2 = 0.05
VariableCoefficient (SD)
Narcissism-1.96 (.35)
Psychopathy-.38 (.37)
Machiavellianism .13 (.37)
Agreeableness-.35 (.36)
Assertiveness1.44 (.35)
Extraversion .46 (.40)

People become less narcissistic and more assertive with age. This result is statistically significant although the effect is quite weak — people who are 1 SD more assertive are only 1.44 years older on average. Older people and people who date younger partners are also significantly less likely to report wanting a partner who is dominant and assertive, with no other major changes in partner preference.

Good looks don’t affect popularity

Popularity with friends is driven by the same traits for both men and women: extraversion, agreeableness, and attractive talents. Quite remarkably, physical attractiveness has almost no impact for either gender, and neither does the dark triad.

I find the lack of relationship between looks and popularity surprising. Looks are strongly correlated with attractive talents (humor, art, athletics) and if we don’t control for those talents then the relationship between looks and popularity shows up, although still much weaker than either extraversion or agreeableness. Perhaps people prefer to hang out with friends of similar physical attractiveness, rather than those who overshadow them in the beauty department.

Sex is Other People

As I said before, my goal with this post was to showcase a lot of information and let the readers draw their own conclusions. Before you do, remember that statistical significance doesn’t imply a huge effect size, that my measures are messy and limited, and that some of the positive results are likely artifacts of noise. My sample also has various selection biases, although if you’re the sort of person who reads Putanumonit you’re probably dating the sort of people who fill out Putanumonit surveys and so these results are very relevant to your own life.

But with all those caveats, I think there’s a major theme that emerges: mating success is about focusing on other people, not yourself. Assertiveness, extraversion, humor — engaging with others leads to romantic success for both genders. Caring about others also helps men make friends and helps women find partners. Of the dark triad traits, the one that is focused on engaging with others even if in a nefarious way (Machiavellianism) is helpful, while caring about oneself instead of others (narcissism, psychopathy) are neutral or negative. Physical attractiveness is important, but it’s far from being an overwhelming factor.

This is good news. Assertiveness and extraversion don’t show up on your forehead, they are demonstrated in your behavior which you have control over. It’s hard to change fixed characteristics about yourself such as beauty or status. It’s easier to practice engaging with people.

People who have met me since I came to New York in my mid twenties find it hard to believe that for long stretches of my youth I didn’t have social confidence or many friends, but it’s true. I had to change social scenes several times and learn to thread the line between assertiveness and disagreeableness. I became less self-absorbed, more curious about others. This all massively helped my dating life. I also got older, of course, which helps.

Mating success isn’t guaranteed, and some people have a much stronger starting point than others. But it always starts with going out and talking to people.

P.S.

I’m not sure if I’m going to publish the raw data, but I can prepare a sanitized version to share upon request if you write to let me know what you want it for. If you’re a researcher and think that this data or analysis could be used for a published paper I may be interested in collaborating.