Decades of relationship research have identified one pattern that reliably distinguishes satisfied couples from dissatisfied ones. It is simpler than most people expect.
The pattern is not how couples argue, how often they have sex, whether they share interests, or how compatible they were at the start. It is what happens in the small, unremarkable moments when one partner reaches out and the other either responds or does not. Researchers call these bids for connection, and the accumulated record of how they are answered predicts relationship outcomes better than almost anything else that has been measured.
What a bid actually is
A bid is any small attempt to gain attention, affection, or shared experience. It is rarely framed as a request. “Look at this,” said about something on a phone screen. A comment about the weather. A hand on a shoulder while passing behind a chair. A sigh. Most bids are so minor that neither partner would describe them as significant, which is precisely why the response to them is so revealing — nobody is performing.
The response falls into one of three categories. Turning toward means engaging, however briefly: looking up, answering, acknowledging. Turning away means missing it, usually through distraction rather than intent. Turning against means responding with irritation or dismissal.
In observational research on newlyweds, couples who were still together years later had turned toward each other’s bids the large majority of the time. Couples who had divorced had done so far less often — roughly a third of the time in the most-cited figure. Neither group was arguing during those observations. They were making small talk.
The version of this finding that generalises best
The bid research comes largely from one research tradition. A related finding, from independent work, is arguably more robust and gets far less attention: what matters most may be how partners respond to each other’s good news.
Research on what psychologists call capitalisation — sharing positive events with someone — has found that a partner’s response to good news predicts relationship wellbeing, in some studies more strongly than their support during difficulty does. Enthusiastic, engaged responses that invite elaboration are associated with higher satisfaction and intimacy. Muted, distracted, or deflating responses are associated with the opposite, even when the partner intends no harm.
This is counterintuitive, because the cultural script for a good relationship emphasises being there when things go wrong. The research suggests being there when things go right is at least as diagnostic, and it comes up considerably more often.
The pattern is not how couples fight. It is what happens in the ordinary moments when one of them reaches out.
— WellnessLife editorial
Where the famous numbers overreach
This field has a specific credibility problem worth naming, because the same research is usually presented with a certainty it has not earned.
The widely quoted claims that researchers can predict divorce with over 90% accuracy do not mean what they appear to. In several of these studies, the predictive model was built and tested on the same sample — the variables were selected because they separated the couples in that dataset, then the model was scored on its ability to separate those same couples. A formal critique published in 2001 made this point directly: without cross-validation on an independent sample, accuracy figures of that kind are inflated by construction and say little about predicting outcomes for anyone new.
The sample limitations compound it. Much of the foundational work involved small numbers of predominantly white, middle-class, heterosexual American couples. The 5:1 ratio of positive to negative interactions during conflict, one of the most repeated figures in popular relationship advice, derives from this body of work and functions better as a rough heuristic than as a validated threshold.
What survives the critique is the direction rather than the precision. Responsiveness matters, it shows up in small moments, and it is measurable — those claims are supported by multiple independent research traditions. The specific percentages are not the part to rely on.
The negative side of the same pattern
The corresponding research on destructive behaviour identified four: criticism aimed at character rather than behaviour, defensiveness, stonewalling — withdrawing and shutting down — and contempt, which is described as the most corrosive of the four.
Contempt is worth singling out because it is the clearest inversion of responsiveness. Where turning toward a bid communicates that the other person’s attention is worth having, contempt communicates that they are beneath consideration. Eye-rolling, mockery, and sarcasm delivered from a position of superiority all fall into this category, and they are among the more consistently identified markers of relationships in trouble across studies.
- Bids for connection are small, frequent, and mostly unremarkable — which is why responses to them are informative
- Couples who stayed together turned toward each other’s bids far more often than those who separated
- Independent research finds that responses to a partner’s good news predict satisfaction as strongly as support during difficulty
- Enthusiastic responses that invite elaboration outperform brief, distracted acknowledgement
- The widely quoted divorce-prediction accuracy figures were not cross-validated on independent samples
- The 5:1 positive-to-negative ratio is a heuristic from small samples rather than a tested threshold
- Contempt is the most consistently identified destructive pattern, and is the direct opposite of responsiveness
What to actually do with this
The practical implication is unusually concrete, which is rare in relationship research.
Notice bids. Most turning away is not rejection; it is a phone, a deadline, or a train of thought. The intervention is attention, not effort.
Respond to good news properly. When someone tells you something went well, the response that helps is interested and curious — asking about it, letting them tell it. “That’s great” while looking at a screen is technically positive and functions as turning away.
Watch for contempt specifically, in yourself more than in your partner. Sarcasm and eye-rolling are easy to defend as humour and are the most reliably damaging thing on the list.
The practical takeaway
The most useful finding in relationship science is not about conflict resolution, compatibility, or communication techniques. It is that relationships are largely built and eroded in moments too small to notice at the time — and that the response to a partner’s minor attempt at connection, particularly when they are pleased about something, carries more information than any single argument does.
The precise numbers attached to this research are shakier than their popularity suggests, and it is worth knowing that. The underlying pattern has held up across independent lines of investigation, and it points somewhere unglamorous and immediately actionable: look up when someone speaks to you.
Sources
- Gable SL, Reis HT, Impett EA, Asher ER. “What do you do when things go right? The intrapersonal and interpersonal benefits of sharing positive events.” Journal of Personality and Social Psychology, 2004.
- Heyman RE, Smith Slep AM. “The hazards of predicting divorce without crossvalidation.” Journal of Marriage and Family, 2001.
- Gottman JM, Levenson RW. “The timing of divorce: predicting when a couple will divorce over a 14-year period.” Journal of Marriage and Family, 2000.
- Reis HT, Clark MS. “Responsiveness,” in The Oxford Handbook of Close Relationships, 2013.
- Gottman JM, Silver N. The Seven Principles for Making Marriage Work — source of the bids and turning-toward framework.
