A Brazilian study published in Evolution and Human Behavior finds that people’s attraction to antisocial traits largely mirrors their own personalities. Those with higher levels of Dark Triad traits rate such traits as more desirable in a potential partner, while a profile combining dark characteristics with physical assault remains highly unattractive. The researchers also found that adding fame, wealth, or good looks did not improve the appeal of a violent criminal profile. The study tested two ideas about why someone might be drawn to such partners: assortative mating, or homogamy, and the so-called desirability by-product hypothesis.
In evolutionary terms, lasting relationships are typically built on prosocial qualities such as kindness, trustworthiness and cooperation. Yet some people are drawn to antisocial personalities, and in extreme cases to individuals who have committed serious crimes, a sexual attraction known as hybristophilia. The new work delves into whether these attractions have a basis in mating patterns or are simply by-products of other desirable traits.
The idea for this study came from observing the attraction to famous criminals, such as Ted Bundy, who famously received many love letters during his imprisonment. This observation led me to question whether there might be an evolutionary basis for being attracted to individuals with socially undesirable traits, given that these traits could bring disadvantages to those who form relationships with them, study lead author Katharina P. da Silva told PsyPost.
The researchers focus on the Dark Triad of personality—narcissism, Machiavellianism, and psychopathy—and test whether people’s own traits influence whom they find desirable. Narcissism involves grandiosity and a need for admiration; Machiavellianism is marked by manipulation and deceit; psychopathy features callousness and impulsivity. The traits are spread along a spectrum in the general population, not confined to clinical disorders or criminality.
Early evolutionary studies suggested these traits might offer advantages in dating, with some evidence that higher Dark Triad scores relate to preferences for short-term relationships. A sequence of studies since has debated whether dark traits are inherently appealing or merely co-occur with more attractive characteristics. The current work describes two competing explanations: the desirability by-product hypothesis and homogamy, the tendency to pair with similar others.
The researchers recruited 400 Brazilian adults aged 18 to 60, with an average age of about 28. The sample comprised 193 women, 176 men, and 31 participants who identified in non-binary terms or in other gender identities. Because the latter groups were small, they were excluded from statistical models testing gender differences, leaving 369 participants for those analyses.
Participants completed online questionnaires to measure their own personalities and behavioural tendencies. The Dirty Dozen 12-item scale assessed Dark Triad traits and was adapted to rate how desirable each trait would be in a hypothetical partner. They also completed two risk-taking measures: a sexual sensation-seeking scale and a sensation-seeking portion of an impulsivity questionnaire.
In the experimental portion, participants read one of six brief, gender-neutral vignettes describing a hypothetical person and then rated how willing they would be to enter a casual short-term relationship or a stable long-term relationship with that person, on a seven-point scale.
The six vignettes began with a prosocial control profile—an intelligent, kind, and helpful individual who values teamwork and considers the impact of actions on others. The second vignette described a Dark Triad profile with no criminal history, portraying calculated manipulation for personal gain. The third used the same dark profile but added that the person had committed physical assault.
The remaining three vignettes tested the desirability-by-product idea by pairing the dark profile with a highly desirable attribute. In the fourth scenario, the person gained fame after the assault and built a large social media following. In the fifth, the person was described as wealthy and of high social status, able to post bail after the arrest. In the sixth, the person was extremely physically attractive, drawing attention for looks after the assault arrest.
Analyses showed clear evidence of homogamy: participants who scored higher on Dark Triad traits tended to rate those same traits as more desirable in a partner. The correlations were about 0.59 for Machiavellianism, 0.66 for psychopathy, and 0.69 for narcissism, on a scale where 1 equals a perfect match. Even when gender, sexual sensation-seeking, and thrill-seeking were controlled for, a person’s own Machiavellianism, psychopathy, and narcissism remained the strongest predictors of preference for those traits in a mate.
While sexual sensation-seeking and thrill-seeking showed weak links to dark-trait preferences in simple comparisons, their influence disappeared once participants’ own Dark Triad scores were accounted for. The researchers also compared each participant’s preferences with randomly paired self-ratings and found that preferences aligned more closely with their own scores than chance.
The findings align with previous PsyPost coverage of real-world couples, which has suggested that partners tend to share similar levels of Machiavellianism, narcissism and psychopathy, and that people with dark personalities feel greater relationship satisfaction when their partners share similar levels of certain traits.
“It is worth noting that, to the best of our knowledge, this is the first evidence of assortative mating for antisocial traits in a Brazilian sample,” da Silva noted. “We hope these findings encourage further investigation into the underlying factors driving such attraction in Latin American populations.”
In the vignette results, the prosocial profile was rated most desirable, with an average willingness to date of 5.92 on the 1-to-7 scale. By contrast, the Dark Triad profile without criminal history fell to 2.40, and the profile including physical assault averaged just 1.52. The addition of fame, wealth or looks did not improve desirability: the famous criminal averaged 1.46, the wealthy 1.49, and the attractive 1.60, with none differing statistically from the plain criminal profile.
“The main takeaway is that, despite evidence showing some attraction to antisocial traits and criminal individuals, prosociality still remains a far more desirable trait in the context of mate choice,” da Silva explained. “However, this finding suggests that the effect of antisocial traits combined specifically with the crime of physical assault may have acted as a dealbreaker, that is, a negative trait in a potential partner that carries more weight in attraction than any positive characteristics presented by them.”
The study also noted that participants were marginally more willing to engage in short-term than long-term relationships overall, with men slightly more willing than women for casual dating. Crucially, the dark and criminal profiles did not gain any advantage in the casual context, with the gap to the prosocial profile remaining similar across dating contexts.
Researchers highlighted several caveats: most participants rated themselves and their ideal partners low on the Dark Triad scale, creating a floor effect that could inflate apparent similarity between self-ratings and preferences. The authors also acknowledged limitations of the 12-item scale’s internal consistency, the lack of attention-check questions, and the use of written vignettes rather than photographs or real interactions. They noted that the crime choice—physical assault—might especially deter attraction, and that the study could not test how non-violent offenses or different forms of criminal behaviour might be received when paired with wealth, fame, or looks.
Looking ahead, the researchers hope to test whether mating preferences shift in social environments where aggressive behaviour might be perceived as protective or economically advantageous, suggesting targeted sampling in such contexts. “Future research could benefit from employing targeted sampling strategies (e.g., in high-crime areas or specific subcultures) to test whether the observed mechanisms are amplified in ecologies where the cost-benefit ratio of criminality is altered,” da Silva said.
The study was conducted by Katharina P. da Silva, Bruno H. Amaral, Marco Antonio C. Varella, and Jaroslava V. Valentova.
