Being wrong about your own relationship

Assessing a relationship you are in is a difficult observational task performed by someone with a stake in the result. The distortions are known.

Positive illusions

People in relationships hold views of their partners that are more favourable than the views held by friends, and more favourable than the partner's view of themselves.

This is well documented and is generally associated with better outcomes — couples with stronger positive illusions report higher satisfaction and are more likely to remain together.

Which is a genuinely odd finding: the accurate view is not the beneficial one.

The proposed mechanism is that idealisation is partly self-fulfilling, with partners moving toward the favourable view held of them.

What it means for self-assessment is that your view of your partner is probably too generous, and that this is normal and mostly fine.

The distortions that are not benign

Effort discounting. You observe all of your own contributions and some of your partner's, which produces a reliable overestimate of your own share.

Both partners' estimates typically sum to well over a hundred per cent.

Attribution asymmetry. The tendency to attribute your own poor behaviour to circumstances and your partner's to character, which is a general property of social perception and operates strongly here.

In distressed couples the pattern is stronger and it is self-sustaining.

Retrospective revision. The current state of the relationship shapes the account of its history.

People in distress recall the past more negatively than they reported it at the time, which is measurable in longitudinal studies where the earlier reports exist.

Which means the sense that it was never good is frequently a symptom rather than a discovery.

Adaptation. Gradual deterioration is difficult to detect from inside, because each state becomes the reference point for the next.

This is why people who leave long relationships frequently describe realising afterwards how bad it had become.

The corrections available

External observers. Discussed elsewhere on this site — friends predict outcomes better, and their view is worth soliciting explicitly.

Written records. Anything written at the time — messages, a diary — is a better source than recollection, and it frequently contradicts the current account in both directions.

Behavioural measures rather than impressions. When did we last do something together that we both enjoyed? How many arguments in the last month? Who initiates conversation?

Countable things resist the distortions that impressions do not.

The hypothetical. Would you begin this relationship now if you were not in it? Not decisive, and it bypasses several of the biases at once.

The asymmetry of error

Worth thinking about.

The two possible errors are staying in a relationship that should end and ending one that should have continued.

The biases described above mostly push toward the first — idealisation, investment, adaptation and retrospective revision all favour continuation.

Which suggests that where the assessment is genuinely balanced, the balance point is probably further toward leaving than it appears.

That is a general tendency rather than advice about any particular relationship, and it is worth weighing against the equally documented tendency to overestimate how bad the aftermath of ending one would be.

The practical version

Assess relationships using countable things, over periods rather than moments, with input from people who are not in them.

And distrust any assessment made during a strong emotional state in either direction, since both the good and bad versions are produced by the same unreliable instrument.

The friend test

One practical method that works around several of these biases at once.

Describe the situation as though it were happening to a friend, in the third person, and ask what you would say to them.

People are consistently more accurate and less forgiving when assessing someone else's relationship, and the shift in perspective is enough to access that.

Writing it down helps, because saying it aloud to yourself preserves the internal framing that the exercise is meant to break.

The measured version

For anyone wanting something harder than impression.

Keep a brief note for a month — one line a day about how the relationship was that day.

A month of such notes is a substantially better instrument than recollection, and it resists the retrospective revision that makes both the good and bad accounts unreliable.