"You're 94% compatible!" Love calculators adore percentages. But can science actually measure real love compatibility? Psychologists have been studying couples for decades, and their findings upend quite a few common beliefs.
Opposites attract… mostly in films
It's one of the most widespread ideas: a shy man and an outgoing woman, a planner and a dreamer, and sparks fly. In the 1950s, sociologist Robert Winch even put forward a theory of "complementary needs". The studies that followed did little to support it.
Similarity, on the other hand, is one of the most robust findings in relationship psychology. In the 1960s and 1970s, psychologist Donn Byrne showed that we are more attracted to strangers who share our opinions. And in real life, partners resemble each other far more than chance would predict in age, education, social background, values, and political or religious views. Sociologists call this assortative mating, or homogamy.
When it comes to personality, though, similarity matters much less than you might think. Two people with different temperaments can make a very happy couple.
The similarity we think we see
A large review published in 2008 by Matthew Montoya and colleagues, covering more than 300 studies, adds an interesting nuance. Actual similarity matters mostly at the start, when we know little about the other person. In established relationships, what counts is perceived similarity: the feeling of being on the same wavelength, even when it isn't entirely true.
In other words, compatibility is also a story partners tell about themselves.
What we think we want… and who we actually like
Ask someone to describe their ideal partner and they'll happily list humour, ambition, looks, culture. In 2008, psychologists Paul Eastwick and Eli Finkel compared these stated preferences with what actually happened at speed-dating events. The result: the criteria people announced beforehand were very poor predictors of who they actually liked face to face.
Checklists are a poor guide. We find out what we like by meeting people, not by filling in a form.
What really predicts a happy couple
In 2020, an international team led by researcher Samantha Joel pooled 43 longitudinal studies covering more than 11,000 couples and let algorithms search for the best predictors of relationship satisfaction.
Each person's own traits (being anxious, being generally satisfied with life, being comfortable with closeness…) play a role, but a modest one. And the partner's traits add almost nothing. What matters most are things that arise within the relationship itself:
- the sense that the other person is committed to the couple;
- the appreciation partners feel for each other;
- sexual satisfaction;
- the impression that the partner is happy in the relationship;
- a low level of conflict.
The authors' conclusion: compatibility isn't found ready-made, it is built. The "right" partner is less an ideal person than a relationship that works.
How you argue matters more than how often
American psychologist John Gottman observed hundreds of couples discussing their disagreements. Every couple argues; what sets lasting couples apart is how they do it. He identified four especially destructive behaviours, which he called the "four horsemen":
- criticism (attacking the person rather than the problem);
- contempt (sarcasm, mockery, eye-rolling), the most toxic of the four;
- defensiveness (playing the victim, counter-attacking);
- stonewalling (shutting down, refusing to engage).
Gottman also found that in stable couples, positive exchanges (humour, affection, listening) far outweigh negative ones during conflict, by a ratio of roughly five to one.
What about attachment?
Since the work of Cindy Hazan and Phillip Shaver in 1987, researchers have applied attachment theory, which grew out of observing infants, to adult love. People with a secure attachment style are comfortable with both intimacy and independence; others are more anxious (afraid of being abandoned) or avoidant (uneasy with closeness). These styles shape how people experience relationships, but they are not destiny: they can change, especially with a reassuring partner.
So, what are love calculators worth?
No calculation based on names, birth dates or star signs predicts whether a couple will succeed. That is exactly what Lovemeter honestly says: its score is calculated letter by letter, using a fixed method, and the test is meant above all as a game. Its five dimensions (chemistry, communication, passion, trust, long term) have the merit of pointing at what really matters, and of getting the conversation started.
In short
- Couples mostly resemble each other in age, education, background and values; opposites rarely attract.
- The criteria we list are poor predictors of who we're actually drawn to.
- A couple's happiness depends mostly on the relationship itself: commitment, appreciation, intimacy, handling conflict.
- Love calculators are a game. The real test is time.
