You leave a good conversation half-convinced you were being tolerated. The other person leaves thinking exactly the same thing. Psychologists have a name for it.

You had a good chat. Fifteen minutes, easy enough, a couple of proper laughs. Then you walked away and started editing: the bit where you talked too long, the joke that got a courtesy laugh, the way you said goodbye. By the time you got home you had settled on a verdict. They were being polite.

They were not being polite. They went home and did the same thing about you.

What is the liking gap?

The liking gap is the consistent difference between how much someone actually liked talking to you and how much you think they did. After a conversation with someone new, people reliably underestimate how positively their partner viewed them and how much the partner wants to talk again.

It is not modesty and it is not low self-esteem. It shows up in people who are perfectly confident in other areas of their life. It is a specific, repeatable error in one specific judgement: reading how a conversation landed for the other person.

Where the research comes from

The term comes from a 2018 paper in Psychological Science by Erica Boothby, Gus Cooney, Gillian Sandstrom and Margaret Clark, titled "The Liking Gap in Conversations: Do People Like Us More Than We Think?".

The researchers paired strangers for conversations, then asked each person two things separately: how much did you like your partner, and how much do you think your partner liked you. The second number came out lower than the first, over and over, in both directions at once. Both people walked away underrating themselves.

They then tested whether familiarity dissolves it, following first-year university students who were living alongside each other. It did not dissolve quickly. The gap was still measurable long after these people had stopped being strangers to one another.

Later work by Wolf, Nafe and Boothby found the same pattern in children, appearing at around age five. Whatever this is, we acquire it early and we do not grow out of it on our own.

Why it happens

Three things stack up, and none of them are about you being bad at talking.

Your inner monologue is only audible to you. You have a running commentary on your own performance and complete silence on theirs. So you compare your worst internal moments against their calm exterior, which is not a fair fight.
Politeness reads as pity. People are warm in conversation as a default. When you are already braced for rejection, ordinary warmth looks like someone being kind about a bad situation.
Nobody says it out loud. The one thing that would settle the question, someone telling you plainly that they enjoyed it, almost never happens, because they are busy assuming you were the one being generous.

The result is a closed loop. Two people leave a conversation both wanting a second one, both convinced the other does not.

What it actually costs you

The liking gap would be a harmless quirk if it stayed in your head. It does not, because you act on it.

You do not send the message. The person you got on with at the conference, the one you swapped numbers with, becomes a contact you never text, because texting first would confirm you were the keener one. This is most of the reason connections made at events quietly disappear afterwards.

You do not go back. A club, a class, a running group where the first session went fine gets written off as not for you, on evidence that was never real.

You do not start the next one. Each conversation you score as merely tolerated becomes proof that these things do not work for you, and the sample builds. This is how a bad estimate hardens into a settled belief about yourself, and why social confidence looks so much more like a skill than a trait.

What to do with this

Knowing about the liking gap does not delete it. You will still leave conversations feeling like you overstayed. What changes is what you do with that feeling.

Treat the post-match analysis as unreliable evidence. Not wrong, necessarily. Just not admissible. Your read on how it went is the one part of the event you are demonstrably bad at judging.
Assume symmetry. Whatever you are worrying about, there is a decent chance they are worrying about the mirror image of it. That is the actual finding.
Act on the conversation, not the verdict. If it was enjoyable while it was happening, that is the data. Send the message on that basis.
Say the thing. Telling someone you enjoyed talking to them costs nothing and closes their gap instantly. It is the rare social move with almost no downside.

Where this meets the harder problem

The liking gap explains why we retreat after a conversation. It has a sibling that explains why so many conversations never start: the person who speaks first carries the entire risk of the approach, and the person who stays quiet carries none. We have written about that asymmetry separately, in why the world is full of lonely people afraid to make the first move.

Both problems have the same root. We are working from private information about ourselves and public information about everyone else, and the two are not comparable. Structure helps where willpower does not. A room where openness is signalled rather than guessed at removes most of the guesswork the liking gap feeds on.

That is what we built FirstMove for. If you would like to test the theory somewhere real, download FirstMove and use it at the next event you go to alone.