Relationship Punch
Say the difficult thing kindly

Dating

The friends who see it before you do

External observers predict relationship outcomes better than the people in them, which is worth knowing and difficult to use.

Father expressing frustration while teenage son shows defiance, seated outdoors.
Father expressing frustration while teenage son shows defiance, seated outdoors. · Photo via Pexels
Health information notice. General information — not a substitute for professional advice. Read the full disclaimer.

One of the more consistent findings in relationship research is uncomfortable: friends and family are better at forecasting whether a relationship will last than the couple.

The finding

Studies asking friends, family and the partners themselves to predict relationship outcomes have generally found that observers' predictions correlate with actual outcomes at least as well as the participants', and in several studies better.

Roommates and close friends have performed particularly well in this research.

The effect is not enormous and it is consistent, and it holds for a straightforward reason: observers are not motivated to reach a particular conclusion.

Why the people inside are worse

Positive illusions. People in relationships hold idealised views of their partners, which is associated with satisfaction and stability and which distorts prediction.

This is a documented and generally benign phenomenon. It does not help with forecasting.

Investment. Having committed to something changes how evidence about it is evaluated.

Access asymmetry. Partners see much more of each other, and observers see the relationship rather than the person.

Which turns out to matter. How two people behave together — the interruptions, the tone, who defers — is visible from outside and difficult to observe from inside.

The narrative. People in relationships have an account of the relationship that organises the evidence, and disconfirming details are absorbed into it.

The practical problem

Knowing that friends are better predictors does not make their views easy to use.

Friends generally do not say anything. The costs of criticising a friend's partner are high — the relationship may continue, the criticism may be remembered, and the friendship may not survive it.

So concerns are softened into hints, or expressed only after the relationship ends, at which point everyone reports having known.

Which means the information exists and is not transmitted, and both parties are behaving rationally.

Extracting it

Some approaches that work better than asking whether they like your partner.

Ask for the specific rather than the general. "What do you notice about how we are together" produces more than "what do you think of him".

Give explicit permission, once. "I'd genuinely rather know now than in two years, and I won't hold it against you."

Whether this is honoured depends on whether it is true, and friends can generally tell.

Ask a friend who is not close to your partner. Less invested in the ongoing social arrangement.

Notice non-answers. Friends who avoid the subject, respond to questions about your partner with comments about you, or become notably positive about something unrelated are frequently communicating something.

Ask afterwards, about previous relationships. Which is not too late to be useful.

Friends will speak freely about a relationship that has ended, and the pattern across several is informative about how you choose.

Receiving it

The difficult part, and the reason the information usually does not arrive.

The instinctive response to criticism of a partner is defence, and one defensive response is generally enough to close the channel permanently.

What is available instead is to receive the observation, ask for specifics, and not respond immediately.

Also worth holding: friends can be wrong. They see selected moments, they may have their own reasons, and the loss of your attention to a new relationship is itself a factor.

A single concern from one friend is weak evidence. The same observation from several people who do not know each other is much stronger, and that pattern is worth taking seriously.

The isolation signal

One specific application worth stating.

Where a relationship is accompanied by reduced contact with friends and family, and where concerns raised by them are attributed to jealousy or interference, that combination appears repeatedly in accounts of controlling relationships.

The mechanism is direct: removing outside observers removes the people best placed to see the pattern.

Which makes maintaining independent friendships a protective measure as well as a good in itself.

General information about relationships, not psychological advice. If a partner discourages contact with your friends or family, contact a domestic abuse service or emergency services in your area.

friendsjudgementoutside viewprediction
Daniel Osei
Conflict & Communication, Relationship Punch

Daniel is a mediator who works mostly with separating couples, and is unromantic about what actually reduces harm.

More from Daniel →

Also by Daniel Osei

Endings & Repair

Starting again, with what you know now

A short account of what the evidence across this whole subject actually supports, for anyone beginning again.

Martina Vogel··3 min read

Endings & Repair

What people say they learned

Accounts from people looking back on relationships that ended converge on a small number of observations.

Daniel Osei··3 min read

Endings & Repair

Recovering, and what actually helps

The research on adjustment after separation is unglamorous and reasonably clear about what makes a difference.

Imogen Hart··3 min read