Science of Attraction, Compatibility & Romance | Dr. Paul Eastwick

What It Covers

Dr. Paul Eastwick, psychology professor at UC Davis, presents data from speed dating studies, longitudinal relationship research, and cross-cultural surveys to challenge dominant evolutionary marketplace theories of attraction. His findings reveal that initial impressions are poor predictors of compatibility, gender differences in partner preferences largely disappear in face-to-face settings, and sustained proximity — not trait-matching — drives lasting romantic bonds.

Key Questions Answered

• •Dating App Distortion: Dating apps function as extreme inequality markets, concentrating attention on a tiny fraction of users while real-world acquaintanceship distributes attraction far more evenly. In brief face-to-face encounters, two strangers agree on whether a third person is attractive only about two-thirds of the time — meaning significant individual variation exists from the start. This disagreement is what allows most people to find partners outside the narrow "consensually desirable" tier that apps artificially amplify.
• •Attraction Builds Slowly: Research on how couples actually form shows the typical first impression is middling, not electric. Most relationships develop through repeated low-stakes interactions where each encounter adds a small piece of evidence — humor, attentiveness, a specific behavior — that gradually shifts one person's assessment of another. Waiting for an instant "spark" causes people to prematurely exit situations where genuine attraction would have emerged over days or weeks of natural contact.
• •Gender Differences Collapse in Person: When Eastwick ran speed dating studies measuring actual partner selection behavior — not self-reported preferences — men and women showed statistically identical sensitivity to a potential partner's ambition and earning prospects. The well-known finding that women prioritize financial status more than men only appears on trait-rating surveys. In real interactions across 40-plus countries and ongoing relationship data, that gender gap disappears, suggesting stated preferences and actual behavior diverge substantially.
• •Idiosyncratic Attraction Levels the Field: As people spend more time together, consensus ratings of attractiveness become less predictive of who pairs with whom. While strangers may agree someone is a "five," a person who has shared specific experiences with that individual may genuinely perceive them as a "nine." This divergence from group consensus — not conformity to it — is the mechanism through which most couples actually form, making prolonged shared activity the most reliable path to mutual attraction.
• •Social Embedding Sustains Relationships: Couples who maintain shared friendships and participate in group social activities — double dates, couple friends, mixed-gender gatherings — report higher relationship satisfaction on average. The mechanism is not explicit feedback-seeking but passive validation: experiencing your relationship in real time alongside others provides a sense of social support without inviting potentially corrosive outside assessments. Men especially benefit from this structure, as their primary intimacy and support needs are otherwise concentrated entirely on one romantic partner.
• •Attachment Orientation Is Changeable: Contrary to the popular fixed-style model, longitudinal couple data show that anxious or avoidant attachment patterns shift meaningfully over time within relationships. A person with a highly avoidant history can become measurably more secure after sustained exposure to a responsive partner. This means attachment style functions less like a permanent trait and more like a current state — making partner responsiveness and accumulated shared experience more predictive of long-term security than an individual's attachment history at the relationship's start.
• •Reciprocal Disclosure Accelerates Bonding: The "fast friends" procedure — a structured sequence of progressively personal questions used in 60-to-90-minute interactions — reliably generates genuine closeness between strangers. The key mechanism is reciprocal self-disclosure: when one person shares something they have not told others, and the other person responds in kind, both parties experience a distinct sense of being seen. This effect outperforms trait-showcasing in early dating contexts and is directly actionable by steering conversations toward personal narrative rather than resume-style information exchange.