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

Why does it matter? Because the dating scripts most people are running were written on bad data

The evolutionary psychology narrative — men want youth and beauty, women want resources and status, compatibility can be calculated in advance — shaped a generation of dating culture and continues to fuel online toxicity. Dr. Paul Eastwick has spent two decades testing those claims against actual behavior. The results land differently than almost everyone expects.

• Men and women show identical preference strength for ambition and attractiveness when evaluating real people — the gender gap exists in surveys, not in behavior. • Both sexes are drawn to younger partners in practice; women actively deny this preference on paper, then contradict it on dates. • An "83% compatibility match" from any algorithm predicts attraction no better than a coin flip. • Men — not women — are more likely to say "I love you" first, push for exclusivity sooner, and think about their exes after a breakup.

Men and women want the same things — the gender gap in mate preferences disappears the moment you watch what people actually do

Men and women show identical preference strength for ambition and physical attractiveness in real-world interactions — the widely-cited sex differences are a measurement artifact. That's Dr. Paul Eastwick's conclusion after 20 years of studies across ongoing relationships and more than 40 countries.

The problem is the experimental paradigm. Decades of evolutionary psychology had people rate hypothetical traits in hypothetical partners on scales. Men said they cared more about attractiveness; women said they cared more about earning potential. The gap looked robust. But rating a checklist in the abstract may capture cultural scripts about what you're supposed to want rather than actual desire.

Eastwick ran speed dating studies to test the difference. Men met real women varying in ambition — some headed to law and medicine, others less professionally driven. The men liked the ambitious women more. Small but real. Then the flip: women evaluated men of varying ambition. "When we flipped it and we looked at what the women were drawn to, not what they said, but what they were drawn to, they also tended to like the ambitious men a little bit. And the magnitude of that preference was identical."

"It's been 20 years of this," Eastwick said. "We've looked at ongoing relationships. We've looked at 40-something countries throughout the world. That narrative plays out every time. There's no gender differences in the extent to which these traits appeal to men and women when they're evaluating real people they've actually met."

One caveat that matters: online is different. The survey-derived gaps persist when people evaluate profiles and hypothetical partners. Face-to-face contact "dramatically reduces the power of the gender differences." The upshot: stop filtering potential partners through absorbed scripts about what your gender is supposed to want, and trust the subjective signal you get from actual in-person time instead.

Women prefer younger partners too — they just won't admit it until they're actually on the date

Four thousand people. That's the sample behind one of Eastwick's more counterintuitive findings: in practice, women are drawn to younger partners at nearly the same rate as men.

Eastwick partnered with a professional matchmaking service where men were, on average, four years older than the women in the pool. Dates were arranged across the full age spread — sometimes the woman was much younger, sometimes the same age, sometimes slightly older than the man. Standard evolutionary predictions would show men strongly preferring younger women and women strongly preferring older men.

"The younger folks appeal to the men more," Eastwick confirmed. "And by the way, it's not a huge effect. It's not like the gross stereotype that's out there." Then the parallel finding: "Women are doing the same thing. They're a little bit more interested in the younger guys. They don't say that on paper. In fact, sometimes they're like, 'Don't set me up with the younger guys.' And then they do and they say, 'Huh, that was interesting. I enjoyed that date. I would like to see him again.'"

The population-level age gap — men pairing with women roughly four years younger — is real and visible in the data. But Eastwick suspects it forms later in the relationship arc, perhaps in who ultimately commits rather than who generates chemistry at first contact. "Whatever is creating this age difference, it's at least not happening on date one."

If you're using any service with age filters, this is worth sitting with: your stated preferences may not predict your actual reaction. Don't filter out mismatches that contradict what you think you want.

An 83% compatibility match predicts attraction no better than a coin flip

An 83% compatibility match predicts attraction no better than a coin flip. That's Eastwick's finding from studies that tried to do exactly what dating apps promise — assess shared traits in advance and use the overlap to predict who would click.

The key distinction is between actual similarity, measured objectively before two people have met, and perceived similarity: the warm, free-floating sense that you have a lot in common with someone. Perceived similarity strongly predicts attraction and relationship satisfaction. Actual similarity, assessed in advance, does almost nothing.

"You're an 83% match on all the things I could assess. You two should like each other," Eastwick described the algorithmic premise. "When we've done studies like that, you basically get a coin flip every time."

The reason is the causal direction. When people are attracted to someone, they generate perceived similarity through motivated reasoning — they find the shared values, focus on them, and convince themselves those are the most important things in the world. "What people think they want doesn't match up with what they actually end up liking once they meet somebody face to face." The people in the happiest relationships are doing exactly this self-serving accounting. "It's like stupid human tricks," Eastwick said, "but it kind of works."

Algorithms can't manufacture perceived similarity in advance because it's a product of attraction, not its cause. The trait-matching that feels like rigorous filtering moves the odds from 50/50 to maybe 53/47 — and may be excluding people who would have generated real chemistry in person.

Dating apps are kleptocracies — the scarcity singles feel online is a medium artifact, not a social reality

Researchers who've analyzed swipe and message data have called it one of the most unequal markets in the world. Eastwick's word for it is sharper: "It's basically a kleptocracy." A tiny fraction of highly-rated profiles captures nearly all the right swipes, while everyone else experiences something close to that classroom exercise where students walk around with low numbers on their foreheads — nobody approaches them, and they start to panic.

This is not how real-world acquaintance works. When two people independently rate whether someone is attractive after a brief interaction, they agree only about two-thirds of the time. That built-in 33% disagreement already begins to redistribute attraction away from a single consensus hierarchy. As time accumulates, the disagreement grows — idiosyncratic impressions form, and the person everyone rates a "six" becomes a "nine" to the one person who spent twenty minutes talking about something nobody else was there to hear.

"The more that people spend time together getting to know each other," Eastwick says, "it reduces some of those market forces that give the desirable people all the advantages."

Apps eliminate that elapsed time entirely. With only a photo and a bio, consensus reigns immediately. The chronic scarcity — everyone competing for the same small pool of profiles — is a feature of the medium, not a reflection of actual social dynamics, which are far less stratified.

The structural antidote is repeated in-person group exposure: intramural sports, improv classes, hiking clubs, recurring dinners. Any context where time accumulates, opt-out pressure is low, and idiosyncratic attraction has room to diverge from the crowd.

The typical first impression in a lasting relationship is middling — attraction grows from there, not from a lightning-bolt moment

The spark model is wrong on the data. When Eastwick looks backward at successful long-term relationships, the starting point is almost always underwhelming. "The typical first impression is middling. That's how we feel at first. Middling. Just kind of... I don't know, middle of the sky. That seemed all right." Not bad. Not over the top. Fine.

What actually happens is a slow accumulation. Each interaction adds a piece: he was pretty funny this time. She was a good listener. He did something kind for someone else that I happened to witness. The window of uncertainty about how you feel gradually narrows. "It takes a little while," Eastwick said, "but attraction can form when two people spend that time together sort of pulling unique things out of each other."

Huberman offered a vivid example: a colleague who became smitten after watching a woman in a neighboring lab aliquot antibodies with unusual speed and ease — tubes between her fingers, talking while working — a sight no one else had ever framed as attractive. They married and had children. Nothing in their external metrics would have predicted it.

"The life of the thing," Eastwick said, "is the little stories and moments that two people are sharing — and that's something that people can be doing more with."

On dates, generating one good micro-story or exchange matters more than landing the right answers to every trait question. A second date after a "fine" first impression is more predictive of long-term attraction than the size of the initial spark.

Men are consistently more eager for commitment than women — and it's because most of them have no one else

Men say "I love you" first. They push for exclusivity sooner, want to escalate the relationship faster, and after a breakup they're the ones still thinking about their exes. "Who's more likely to want to break up?" Eastwick asks. "It's women." The effect is medium-sized and consistent across every stage of the relationship arc — from early dating all the way through dissolution.

The commitment-phobic-man narrative is, by the data, exactly inverted.

The explanation Eastwick finds most compelling: for women, social support flows from multiple sources — friendships, family, colleagues. "For men, it's largely their romantic partner. That's where they're getting most of their support, intimacy needs met." When men enter a relationship, they're often placing most of their intimacy load onto one person. Which explains both the eagerness going in and the devastation going out.

Eastwick is precise about what social support actually requires. It's not about literally using your network — it's about having the sense that people are there. "It's like a bank account you never have to dip into. It just gives you the sense to dip into it." A lot of men don't feel like they have that account. Activity-based friendships stay surface-level. Siblings drift. The romantic relationship becomes structurally load-bearing in a way that makes it fragile.

Building a real support network through teams, clubs, and recurring group structures isn't aspirational self-improvement. It's relationship resilience infrastructure — and Eastwick calls male social isolation one of the challenges of modern masculinity that genuinely worries him.

Happy couples automatically see alternatives as 'weak sauce' — when that protection quietly erodes, infidelity risk follows

Being attracted to someone who isn't your partner is not the problem. People in happy relationships automatically downgrade their perception of alternative partners — whoever gets introduced as a possible competitor ends up perceived as less desirable than any external metric would suggest. "Any alternative partner that you can throw at them," Eastwick said, "they will tend to think that that alternative partner is pretty weak sauce."

Attraction to others can even briefly rebound onto the relationship. Studies where participants fantasized about someone else — then about their partner — show elevated desire for the partner afterward. The flash of outside attraction doesn't threaten; it can redirect.

The actual early warning sign is specific and behavioral: not attraction to an alternative, but sustained private engagement with one. Repeated contact, secretive conversations, a specific person whose presence has somehow stopped triggering the automatic downgrading. "What will start to happen is that the protective layer that people typically put around their partners, it will start to erode, and that's when people are at a greater risk of infidelity."

"The simple fact that we can be attracted to other people — that is not a problem for the average relationship. It's the repeated follow-through on that attraction that becomes a problem."

Don't police attraction. Watch whether there's one specific person your protective mechanism seems to have quietly stopped working on.

Feeling your partner is a good lover is one of the strongest predictors of relationship satisfaction — and desire can be rekindled

Don't treat declining physical intimacy as a verdict.

"Feeling like this person is a good lover or likely to be a good lover is a very, very good sign for how positively you feel about the relationship in general and whether you want the relationship to continue," Eastwick said. The effect holds whether people are describing established long-term relationships or just early attraction — the subjective sense of physical chemistry consistently tracks overall satisfaction.

What concerns him is the fatalism. When desire fades, the instinct is to read it as structural evidence that something fundamental is gone. Eastwick pushes back directly: "There are ways of recultivating sexual feelings about somebody that actually are sexual feelings about somebody. It's not like a switch where it's just on or it's off." Fading passion doesn't mean absent. It may mean it surfaces in specific circumstances now rather than on a reliable daily schedule — which is a different thing entirely.

"I think the key thing for me is not to engage in fatalism about the sexual desire component. When the passion fades in a relationship, that doesn't mean that it's gone forever."

Treat physical intimacy as an active maintenance target, not a passive indicator. The data support investing in rekindling over accepting decline as verdict.

The whole framework points the same direction: real people, repeated contact, room for time to do its work

What's coming into focus — and what Eastwick sees in his own classroom — is a generation that arrived skeptical the apps have been working, and is looking for permission to do something different. The research doesn't offer a better platform. It offers evidence that the conditions for attraction to develop (time, proximity, idiosyncratic exposure, the slow collapse of uncertainty into a stable positive impression) were never replaceable — just temporarily obscured by a medium that made the worst conditions feel like the default.

The algorithm was always a detour.

---

Topics: attraction, relationships, dating apps, evolutionary psychology, compatibility, gender differences, mate selection, social support, commitment, physical intimacy, attachment theory, similarity, speed dating research