Am I boring to talk to?

Probably you can't tell. That is the finding, and it isn't the comforting version — it's the literal one. Of all the people who were present at your last conversation, the one worst placed to judge how it went is you.

The liking gap

When researchers put strangers in a room, had them talk, and then asked each of them separately how much they liked the other person and how much they thought the other person liked them, the two numbers didn't match. People rated their partner higher than they believed their partner had rated them. Both of them did. They cannot both have been right.

That gap showed up in every setting it was looked for: strangers in a laboratory, adults at public workshops, and first-year students rating the flatmates they had just moved in with. The students were asked five times across an academic year, and the gap was still there months later. It is not first-meeting nerves that wear off by the second week.

The part that makes it hard to argue with

You could reasonably object that people are just being modest, or that they secretly know and are hedging. So the researchers took the recordings of those same conversations and showed them to trained observers who had no stake in any of it.

The observers could tell. Watching the tape, they predicted how much the two people had actually liked each other, and they were right. The liking was visible. It was in the room, on the recording, legible to a stranger with no context.

So this is not a case of people failing to send a signal. The signal was sent, and it was picked up by everyone except the person it was aimed at. What the study does not tell us is what the people in the room were doing instead of reading it — our guess would be the running commentary on their own performance, but that is a guess, and the finding stands without it.

What this does not say

It does not say you are not boring. It cannot: it is a pattern across groups of people, not a verdict on you or on any particular conversation. Some conversations genuinely go badly, and the person who thought that one went badly was right.

It also only covers conversations that happened. Everybody in these studies was already talking to somebody. If what you are reading as a verdict is silence — an opener nobody answered, on a site where people filter a roster by gender, age and country before they read a word — then none of this applies, because there was no conversation to misjudge. That is a different problem with a much duller explanation, and it is not one you should be reading a page about your own company to solve.

It is also worth knowing what didn't matter. Self-esteem made no difference to the size of the gap. Neither did rejection sensitivity, and neither did narcissism. Shyer people did appear to have a wider gap, which is the finding you would most want to be true if you are reading this page — and it rests on a comparison between groups of ten, so treat it as a hint and not much more. We would rather tell you that than let you lean on it.

What to actually do with this

Nothing, during the conversation. There is no technique here.

The use of it is afterwards. If you replay conversations and reach a verdict — that was painful, they were being polite, I won't do that again — you now know something about the instrument producing that verdict: it reads low, it reads low for most people, and it reads low in a way outside observers watching the same conversation do not reproduce. That is not a reason to believe the opposite. It is a reason to stop treating the post-mortem as information and act on something else instead.

The most common thing that verdict causes is not starting the next one — which, as it happens, is also the thing that guarantees it never gets corrected.

And if what you are actually replaying is the ending — the moment you left, or they did, and what it must have meant — that has been measured too, and it almost certainly meant less than you think.