Loading
Ask anyone what makes a relationship work and you will hear about the other person. Kind. Funny. Ambitious. The right type.
Somebody checked whether any of that predicts anything.
Forty-three long-running studies of couples, from twenty-nine different laboratories, pooled together. More than eleven thousand couples.
Then a machine learning method, so the answer had to predict couples it had never seen.
It is easy to explain a relationship after you know how it went. Predicting one you have not met is a much harder test.
That is the test these findings passed.
Which predicted relationship quality best?
Believing your partner is committed. Feeling appreciated. Being satisfied with sex. Believing your partner is happy with the relationship. And how much conflict there is.
Every one is about how you see your own relationship. Not one is a fact about who your partner is.
That is the opposite of how people choose, and of how dating advice is written.
Which of these predicted relationship quality well?
Tap an item, then a bucket
Separately from the relationship, some things about each person predicted quality: how satisfied they were with life generally, how much negative feeling they carried, and depression.
You bring yourself into it.
A relationship is not sealed off from the rest of a life. Someone struggling alone tends to struggle in it.
That is not a fault. It is worth knowing when things feel harder than they should.
All of these together predicted under a fifth of where relationships ended up by the end of each study.
The best measures we have, across eleven thousand couples, leave most of it unexplained.
It does not mean relationships are random or that nothing can be done.
It means self-report questionnaires cannot see most of what happens between two people, and honest research says so.
This is what goes together, measured well. It is not a set of levers.
Nobody tested what happens if you deliberately try to feel more appreciated.
The strongest signals are about perception. Whether you feel your partner is in it, and whether you feel noticed.
Those are things two people can actually talk about.
Mostly Western samples, and mostly heterosexual couples. That is worth saying, because it is the limit of what these numbers can speak for.
What holds elsewhere is a question the data does not answer.
What predicts a happy relationship is mostly how you see it, not who they are. Most of it stays unexplained. And the strongest measurable things are the ones you can say out loud.