How Romantic Attraction and Compatibility Really Work

In the 2013 film About Time, a young man and woman lock eyes for the first time and the universe seems to click into place. It is the oldest promise of the romantic comedy: find the right face, feel the bolt of recognition, and you will know. Psychologist Paul Eastwick of the University of California, Davis has spent two decades testing whether that promise holds up outside the movie theater. His answer, delivered with the patience of someone who has run the numbers, is that it mostly does not (00:00).

Eastwick's interest in the subject began with heartbreak. As a teenager he read a Time magazine cover story called "Infidelity: It May Be in Our Genes," published not long after a girlfriend left him for his best friend. The article described evolutionary psychology's view of mating as a kind of ruthless marketplace, men and women forever scanning for an upgrade (04:43). A friend later gave him blunt advice: hit the gym, buy better clothes, raise your "mate value," a number from one to ten that supposedly determines how well you do in the dating market. Years later Eastwick met a woman named Lauren who told him he checked every box on her list, intelligent, accomplished, compatible interests, and then chose someone else anyway (11:11). The scoreboard said he should have won. He did not. That contradiction became the engine of his research career.

Eastwick calls the marketplace story the "EvoScript": the idea that everyone has a mate value, that men and women want fundamentally different things, and that people pick a lane between short-term and long-term strategies (15:12). Popularized by podcasters and self-help gurus, the script feels intuitive. It is also, according to Eastwick's data, badly incomplete. One clue is what happens to agreement about attractiveness as people spend time together. Strangers looking at photographs tend to agree closely on who is a six and who is an eight. But the longer two people actually know each other, the more that consensus falls apart, some people start seeing a three where others see a nine (18:23).

"Relationships are gonna work best when two people find each other who think that each other is a 10." — Paul Eastwick, UC Davis psychologist (18:50)

That line matters because it inverts the marketplace logic entirely. A relationship does not need the approval of twenty thousand strangers rating a photograph. It needs one person to see something nobody else quite sees. Eastwick's speed-dating studies back this up in an unglamorous way: when researchers ask people what they want in a partner, intelligence, humor, adventurousness, those stated preferences barely predict who they like once they actually meet someone face to face. Introduce a person with three different but equally appealing qualities, attractive, sociable, kind, and the attraction is just as strong (21:56).

"The things that we say we're looking for don't match up all that well with the traits that actually appeal to us once we meet folks face to face." — Paul Eastwick (22:25)

The much-repeated claim that men chase looks and women chase status also shrinks under scrutiny. In live speed-dating interactions, a trait like ambition draws equal interest from men and women, the gender gap shows up in surveys, not in behavior (27:04). Dating apps, Eastwick argues, make the gap look bigger than it is. In speed dating, men say yes to a second date about 50% of the time and women say yes about 35% of the time, a real but modest difference (31:35). Online, men's yes rate stays near 50%, but women's collapses to about 5%, a swing that has nothing to do with biology and everything to do with scrolling through strangers' photos without the social cues of an actual conversation (31:55). Eastwick thinks this distortion feeds the online "manosphere": men who feel rejected 95% of the time start blaming women or feminism instead of the format of the app itself (32:50).

Perhaps the most humbling finding involves psychologist Samantha Joel, Eastwick's collaborator, who tried to see whether self-reported data, traits, goals, preferences, fed into a machine learning model could predict who would hit it off at a speed-dating event. The model's accuracy: 0% (36:30).

"We have been completely unable to explain ahead of time why two people end up being compatible." — Paul Eastwick (36:45)

Compatibility Is Built, Not Found

If compatibility cannot be predicted from a questionnaire, what is it? Eastwick's answer is closer to gardening than shopping. Two people grow together, developing private rituals, in-jokes, and shared stories, and it is that accumulated history, not a matching algorithm, that creates closeness (41:15). He offers his own long-distance relationship with his partner Allison as an example: unable to see each other often, they started cooking the same dinner over video calls several nights a week, he taught her about wine, she taught him to cook, and the ritual became scaffolding that still holds the relationship up years later (51:30). He also credits a college friendship, with a woman named Crystal who kept inviting him into a wider circle, for introducing him to the partner he eventually dated, a reminder that the number of friends of the opposite gender someone has predicts their odds of forming a relationship later (47:03). His practical advice follows from this: do not rule people out because they fail to match a mental checklist, since every new person is a chance to discover a preference that was never on the list in the first place (47:29).

The Mirror We Mistake for a Window

Compatibility is only half the story. The other half is how accurately partners read each other once they are together, and here University of Maryland psychologist Ed Lemay has sobering news. Lemay studies social projection, the tendency to assume other people, especially romantic partners, are more like ourselves than they really are (57:53).

"Social projection is a bias where we assume other people are more like us than they really are." — Ed Lemay, University of Maryland psychologist (58:03)

The bias cuts both ways. Assuming a partner shares your warmth can build closeness. Assuming a partner shares your irritability or wandering eye can manufacture conflict that was never really there (58:26). Lemay walked through several listener stories to show how this plays out. One listener, Kelly, felt hurt when her husband did not open the car door the way a friend's husband had, reading his omission as a verdict on his care for her, when for him the gesture might carry no meaning at all (63:11). Another listener, Morgan, twenty-three and certain she does not want children, assumed her boyfriend's ambivalence masked hidden resentment, when research suggests people often project their present-day feelings about a relationship onto its imagined future, with no guarantee the forecast is accurate (67:33).

Projection reaches beyond the living room. Lemay's research on "primal world beliefs," general convictions about whether the world is good, safe, or enticing, found that people who see the world as more enticing also rate their partners as more supportive, even compared to what objective observers and the partners themselves reported (72:40). The same mechanism shows up in politics: elected officials project their own views onto constituents at roughly the same rate ordinary citizens project onto their political party, a finding Lemay says helps explain why polarization can feel starker than it actually is (75:30). It even extends to machines. Lonely people are more likely to anthropomorphize pets and AI chatbots, treating them as companions who share their values, a pattern Lemay links to a basic human need for connection rather than any confusion about what AI actually is (79:11).

What can be done about a bias this pervasive? Lemay is cautious about promising a cure. But listener Linda, married forty-three years and navigating her husband's post-retirement PTSD, offered a method that Lemay says tracks real evidence: she and her husband began naming their assumptions out loud before treating them as fact.

"When we're projecting, we do a lot of asking each other, 'The story I'm telling myself is,' you know, and we check with each other." — Linda, listener (93:22)

Lemay calls this kind of open checking-in a genuine corrective: it will not erase projection, which he considers too strong a bias to fully extinguish, but it narrows the space where misreading can grow unchecked (94:32). Paired with deliberate reassurance and a habit of noticing a partner's actual expressions of care rather than dismissing them, the practice turns a blind spot into a conversation. Between Eastwick's gardens and Lemay's mirrors, the episode arrives at a similar place from two directions: the partner worth keeping is rarely the one who matches a list. It is the one with whom the story, however imperfect the telling, keeps getting rewritten together.