Is the Attractiveness of “Dark” Personalities Exaggerated?

Bad boys finish first—at least according to a lot of psychologists, as this continues to be the subject of much attention across many articles and videos.

This is perhaps unsurprising, considering the “bad boy” archetype has been around for centuries, presumably reflecting societal attitudes towards such personalities mainly in literature and film. Most media discussing this usually starts by asking the reader or viewer whether they have fallen for some sort of anti-hero in the past, and if so, reassuring the audience that they are not alone. And according to a recent survey, 43 percent of Britons do in fact find the idea of a “bad boy” attractive, alluding to confidence, passion, and charm.

Of course, there is nothing wrong with discussing such surveys or even talking about anecdotal personal experiences or attractions. However, when media pieces bring up the science behind it, they often refer to the “Dark Triad” personalities and their suggested sexual appeal. This is misleading, as these studies often employ methodologies that are not robust enough to conclusively evaluate the attractiveness of such personality traits.

The concept of the Dark Triad (DT)—narcissism, Machiavellianism, and psychopathy—was introduced by Delroy Paulhus and Kevin Williams in 2002. Since then, the DT has generated a storm of interest in both academia and popular culture, leading to thousands of empirical studies, blog articles, and op-eds on the triad.

While conceptually, the three traits describe different tendencies, evidence shows that all three positively correlate with forcefulness, antisocial behaviours, and most strongly with disagreeableness. Not surprisingly, individuals with high levels of such traits tend to create several problems in the workplace, especially when in leadership positions.

Despite the antagonistic nature of these traits, studies have claimed DT personalities offer evolutionary advantages to certain men. Moreover, it has been reported that men who display DT behaviours are perceived as more attractive, especially for short-term sexual encounters, compared to men with lower levels of DT. Let’s review the existing evidence.

Literature Review

One influential study from 2009 reported that people with DT traits had more sexual partners and a preference for short-term relationships. Participants completed different personality measures assessing their narcissism, Machiavellianism, and psychopathy levels, a scale measuring their attitudes towards casual sex, as well as a single question asking the number of sexual partners they have had in the past year. Even though multiple measurements were used to assess short-term relationship preferences, all of them relied entirely on self-reports. This is an issue, especially when considering DT characteristics correlate with deceit and lying.

Studies have also suggested the attractiveness of DT people over non-DT using digitally manipulated faces suggested to resemble "DT facial prototypes." This started in 2011 after the researcher Nicholas Holtzman collected personality self- and peer-reports as well as images of the faces of participants. Using facial software, the author created the prototypes based on the 10 male and 10 female participants scoring the highest and lowest on each DT trait.

While the computerized work is undoubtedly impressive, perceivers were unable to confidently differentiate between the DT prototype and the non-DT prototype. This suggests that DT traits being linked to noticeable physical-morphological characteristics is unlikely. Nevertheless, multiple studies have adopted those facial morphs to assess perceptual attractiveness for short-term and long-term relationships.

A limitation of both self-report and digitally manipulated face studies is that they do not focus on aspects of behaviour from the perceived target. One way to do this is to adopt hypothetical behavioural scenarios. Studies which have applied this method have relied on fictitiously filled-in questionnaires or vignettes which describe either fictitious or actual persons of the opposite sex who score high or low on a DT measurement.

Dark Triad Essential Reads

The Dark Triad’s Dark Underside Shows Where Things Go Wrong

How to Tell if Your Partner Is Hiding a Dark Side

Five separate studies that used vignettes or dating adverts show that participants preferred the high-level DT person for a short-term relationship (STR), while the low-level DT person was favoured for a long-term relationship (LTR). However, three of these evaluated the attractiveness of the DT using only vignettes or supposedly filled-in questionnaires with no images attached to the fictitious persons.

This means participants most likely used their imagination to picture the physical appearance of each fictitious person. This is particularly problematic as it can generate answers inspired by unrealistic ideals or past romantic experiences. After all, selecting romantic partners based on photos and self-reported behavioural descriptions is mirrored in real-life scenarios (e.g., Tinder, Hinge, Feeld).

The remaining studies used physical images in addition to their behavioural descriptors. Most of these studies, however, did not rate the physical attractiveness of the faces prior to the experiment. This means there is no way of knowing how much physical attractiveness influenced the selection of a hypothetical person—we cannot know whether it is the face, behaviour, or both, that is being rated.

This is especially important because evidence suggests physical attractiveness is the most important variable in STRs. Therefore, it is likely that such a selection is made predominantly by how physically attractive someone is, while the behavioural component is ignored. To potentially control for this, all of the images/hypothetical persons would need to be rated similarly (e.g., all representing medium attractiveness), ideally by independent observers.

To be clear, the authors of the mentioned studies above have all acknowledged these limitations. Most of these methodological issues simply reflect how difficult this concept is to measure quantitatively.

In our recent work, which has not yet been published, we wanted to control for high physical attractiveness while also still including a physical attractiveness stimulus. We adopted a behavioural task too—where participants rate the attractiveness of hypothetical individuals on a 5-point Likert scale for both STR and LTR. We developed and refined descriptions of behaviour based on a DT measurement and attached an image of a face to each “person.”

The faces were taken from a face lab that consists of hundreds of faces whose attractiveness has been rated by multiple independent raters. The faces we selected were all of medium attractiveness based on the ratings. A total of 9 hypothetical persons were introduced in this paradigm—low, medium, and high levels of narcissism, Machiavellianism, and psychopathy. The sex of the hypothetical persons was determined by the sexual orientation each participant reported.

In the study, the composition between face and description of behaviour was randomized; meaning each participant saw a different face for a given DT trait level. This was done to clearly distinguish whether the ratings reflect the face, the behaviour, or both. Multiple mixed effects models were run to evaluate how attractiveness is affected by the level of DT and the respective sexual preferences of participants.

The low level of each Dark Triad (DT) trait was consistently perceived as the most attractive, both in the context of STR and LTR. Notably, our study is the first to show that low-level DT traits (i.e., absence of DT) are most preferred for STRs.

Closing Remarks

Previous studies have claimed that there are people who find DT personalities attractive—that is, when the participants are using their imagination to picture such a person or when they are looking at someone who has a DT personality who is also physically attractive. However, our work suggests that when physical attractiveness is controlled for, a DT personality is not attractive at all. While those with a female preference (mostly men) were less selective than those with a male preference (mostly women), both groups preferred the hypothetical person who displays no DT behaviours over the medium and high-level person, regardless of the relationship type.

This leads to multiple questions: Are people attracted to actual DT people, or are they mostly just attracted to physically attractive people? Going back to the UK survey mentioned above, confidence, passion, and charm are not exclusive to DT people—you can certainly possess all those qualities without scoring high on a DT measurement. On the other hand, it could be that attractiveness promotes DT behaviours in some people, which can explain the link between the two found in previous studies.

Theoretically, it could also be that people find physically attractive DT people more attractive compared to physically attractive non-DT people. However, no studies to our knowledge have shown this, which highlights that we do not know enough to make definitive scientific conclusions.

As researchers continue to explore this topic, it is essential to approach it with caution and resist the temptation to make broad generalizations based on limited and inconsistent findings. Forget the notion of "bad boys finishing first"; when it comes to understanding human behaviour, it's time to ensure that psychologists don’t finish last.