In the hopes of finding the “choreography of human attraction,” scientists measured the physiology of people as they embarked on a first date at a music festival. To their surprise, visible cues such as facial expressions or eye contact weren’t able to predict whether a couple was bound to hit it off. Instead, mutual attraction between a couple was predicted based on whether their heart rate and levels of sweating synchronized with each other. 

The study, which has not been peer-reviewed yet, can be found on the preprint server bioRxiv.

Psychologists from Leiden University in the Netherlands gathered 140 heterosexual people and asked them to go on a first date in a small cabin at the Lowlands music festival. Just to add to their nerves even further, the couples were monitored using eye-tracking glasses, that could track their gaze, while their heart rate and sweat levels were measured with electrodes.  

While sat in their dating cabins, a screen was opened to reveal their partner and gauge a first impression. They were asked to rate their date, then asked to rate how much they liked their partner, how much they thought their partner liked them, how similar they thought the partner was in terms of personality, and how much of a connection they had. Finally, participants were also asked to predict whether they thought their partner wanted to exchange emails and go for another date, after which they were asked to rate their partner once again. 

Machine-learning techniques were then used to detect whether behavioral and physiological patterns could predict participants' mutual attraction level. This was unable to predict a successful match if solely looking at expressions (such as smiling, laughing, head nodding, hand gestures, face touching) and gaze fixations (like looking at the partner's body, eyes, etc). However, when looking at their levels of sweating and heartbeats, it could make a relatively accurate guess. 

The researchers also noticed that females were more physically expressive during the date, particularly when it involved mimicking their partner’s expressions and eye gaze. Around 53 percent of the men fancied another date, compared to just 34 percent of the women. 

One other key takeaway was that people are not accurate at predicting their partner’s romantic intentions, with only half of the participants correctly guessing how much their partner liked them.

“We further show that participants who perceived their partner as highly attractive predicted that their partner liked them more than participants who rated their partner as less attractive," the researchers write. "Yet, in reality, there was no correlation between partners’ perceived attraction and partners’ actual liking scores. 

"At the same time, this result contradicts the notion that people excel in their ‘mindreading’ capacities," they add. "Instead, we found that similarly to economic predictions, people are not very good 'emotional statisticians'."

[H/T: New Scientist]