The neuroscience of chemistry.
Why you cannot predict it from a photo.

What chemistry actually is

The experience of chemistry — the immediate sense of ease and resonance with another person, the feeling that conversation flows without effort, the impression that someone understands you before you have finished explaining yourself — has a measurable neurological basis. It is not magic, and it is not arbitrary. It is the product of specific forms of alignment between two people's emotional and physiological states.

Neuroscience research on interpersonal connection identifies several components that contribute to the chemistry experience. Emotional valence alignment — whether two people tend to experience stimuli in similarly positive or negative terms — creates a foundation of shared perspective. Arousal compatibility — whether the energy levels and engagement styles of two people complement rather than clash — determines whether interaction feels energising or exhausting. And dominance dynamics — the implicit negotiation of who leads and who follows in social interaction — shapes whether connection feels comfortable or strained.

These are the three dimensions of the VAD (Valence-Arousal-Dominance) emotional model that underpins Attune's matching system. They are measurable from involuntary emotional response. And they are entirely absent from any photograph.

Why photos predict attraction but not chemistry

Physical attractiveness is consistently predictive of the likelihood that someone will be approached or swiped on. It is considerably less predictive of whether genuine connection will occur once two people interact. This gap — between who seems attractive in a photo and who produces chemistry in person — is one of the most reliably documented findings in relationship science.

In research by Paul Eastwick and Lucy Hunt at the University of Texas, couples who had known each other before dating rated their partners' attractiveness and chemistry-potential very differently from strangers who assessed the same people from photos. The in-person assessments showed much higher variance: people who were unremarkable in photos were often rated as highly desirable by people who had actually interacted with them, and vice versa. Chemistry, in other words, is individually specific in ways that physical appearance is not.

This individual specificity is what makes photo-based matching so poor at predicting it. What produces chemistry between two specific people depends on the interaction of their emotional patterns — patterns that are uniquely personal and that emerge only in genuine interaction, not in static self-presentation.

What makes chemistry measurable

The signals that produce the experience of chemistry are involuntary. When you genuinely find something funny, the facial muscle activation pattern is different from performed laughter. When you are genuinely engaged, the micro-expression patterns of attention and interest differ from polite nodding. When something genuinely moves you, the physiological response precedes conscious awareness by 200-400 milliseconds.

These involuntary signals — captured by the 44 facial Action Units of Paul Ekman's Facial Action Coding System — represent the raw material of emotional profile construction. They cannot be faked, because they occur faster than conscious control allows. They are individually specific, because everyone's genuine emotional response pattern is their own. And they are predictively powerful, because they represent exactly the emotional information that produces the chemistry experience.

Attune's EchoDepth engine analyses these signals during a three-minute video session. The resulting emotional profile — a continuous-valued representation across Valence, Arousal, and Dominance — is matched against compatible profiles to identify people whose emotional patterns are likely to produce the kind of resonance that manifests as chemistry in person.

This is not a guarantee of chemistry — no matching system can guarantee subjective experience. It is a substantially more predictive basis for matching than a photograph, because it is measuring the right thing.