Thursday, 4pm. The plan is at seven. You have been low-key dreading it since Tuesday. Your prediction is precise: you will be tired, the conversation will be effortful, you will be checking the time by nine, and you will need Friday to recover. You consider cancelling. You don't, because you cancelled last time.

Then 11pm. You are walking home and you feel, of all things, good. Lighter. Slightly surprised. You liked the people. You said things you had not planned to say. Nobody drained anything.

If this loop is familiar, you have run into one of the most replicated findings in personality research, and it has a name. Psychologists call it an affective forecasting error. Introverts, it turns out, are unusually bad at predicting how socialising will feel, and they are wrong in one specific direction.

The study that measured the gap

In 2013, John Zelenski and colleagues at Carleton University in Ottawa published a series of studies in the Journal of Personality and Social Psychology with a blunt title: trait introverts underpredict the hedonic benefits of acting extraverted. The design was simple. They asked people to imagine behaving in an outgoing, talkative way across a range of situations and to forecast how they would feel. Then, in later studies, they had people actually do it and reported how they felt.

Introverts forecast less pleasure and more discomfort from acting extraverted than extraverts did. Nothing surprising there. The interesting part came when the behaviour was actually performed: introverts enjoyed acting outgoing far more than they had predicted, and the negative feelings they braced for largely did not arrive. Zelenski's team found introverts were especially prone to overestimating the negative side, the awkwardness and self-consciousness, rather than simply underrating the positive.

In other words, the dread is real, and the thing being dreaded mostly is not.

The one-week experiment

A forecasting study tells you people are wrong about the evening. A different question is whether behaving more sociably for a sustained period actually helps. Seth Margolis and Sonja Lyubomirsky at the University of California, Riverside, tested this in a 2020 paper in the Journal of Experimental Psychology: General, summarised here by the British Psychological Society.

They recruited 131 people and ran them through two weeks. For one week, participants were asked to be as talkative, assertive and spontaneous as they could manage. For the other, as deliberate, quiet and reserved as possible. Half did the outgoing week first. Everyone listed five concrete behaviour changes and got reminders.

During the extraverted week, people reported more positive emotion, more sense of connection and more flow than at baseline. During the introverted week, those measures fell below baseline. And, importantly for anyone reading this as an introvert, baseline personality did not change the result. Introverts got the lift too.

Two honest caveats. An earlier study by a different team had found that strongly introverted people got less benefit and more fatigue from acting extraverted, so the picture is not unanimous. And nobody yet knows which of the three behaviours, talking more, asserting more, or planning less, is doing the work. This is one well-designed experiment, not a settled law.

Why the forecast goes wrong

Zelenski's group offered an explanation that matches what we hear from people constantly. When you imagine a social event in advance, you imagine the hard parts in high resolution: the moment of arrival, the pause after you say something, the small talk with someone you don't know. You do not imagine, because you cannot, the specific good moments, which depend on what other people say and do and have not happened yet. The forecast is built from the parts you can picture, and the parts you can picture are the uncomfortable ones.

There is a second mechanism, which we wrote about in why loneliness makes you say no to the invitation you need. If you have been isolated for a while, your threat detection runs hot, and the forecast gets even more pessimistic. Introversion and loneliness are different things, but they stack. A lonely introvert is forecasting with two thumbs on the scale.

And a third: the memory of the last social event is not the event. It is the overthinking that came after it, and that replay tends to be harsher than the evening was. So the next forecast is trained on a distorted record of the last one.

What we see on our platform

On YaraCircle, a large share of new users describe themselves as introverted, shy, or "bad at this." What we observe about their behaviour is a near-perfect replay of the Zelenski finding.

The dread shows up before the first conversation, as long gaps between matching and sending a first message. But when we look at what happens after a conversation actually takes place, the people who said they were introverted are not less likely to continue it, request a friendship, or come back the next day than people who described themselves as outgoing. In some cohorts they are slightly more likely, we think because once the hard part is over, they are more selective and more invested.

The forecast and the outcome diverge exactly where the research says they should. The gap is at the threshold, not in the room.

The social battery is real, but it is not what you think

None of this means introversion is a mistake to be corrected. The social battery is a real thing, and the research on acting extraverted does not deny it. What it questions is the way the battery gets used as a forecasting tool.

The battery is a good measure of what happened. After four hours in a loud room, an introvert is genuinely depleted, and the extravert genuinely is not. That is a real trait difference and it is fine.

The battery is a terrible measure of what will happen. It gets consulted at 4pm on Thursday, when it is reporting on Tuesday's meeting and a bad night's sleep, and it gets treated as a prediction of the evening. It is not one. The research says the prediction it produces is reliably too dark.

So the practical rule is: trust the battery for recovery, not for forecasting. Use it to decide how long to stay and what to do afterward. Do not use it to decide whether to go.

How to use this without faking extroversion

The advice that follows from this research is not "act like an extrovert." Margolis and Lyubomirsky's participants were asked to be more talkative and spontaneous for one week, not to change who they were. Here is the version that works for people who would rather not.

Pre-commit before the forecast runs. The dread is strongest in the window between agreeing to a plan and the plan itself. Say yes quickly, put it in the calendar, and treat the Thursday-afternoon forecast as weather, not information. It is telling you about your state, not about the evening.

Choose the shape of the event, not whether to go. Introverts do worse in loud, large, unstructured settings and better in small, structured ones. A walk with one person. A game with three. A dinner where the seating means you talk to two people, not twelve. The research is about being more sociable, not about tolerating the wrong format.

Run the experiment on yourself, once. For two weeks, before any social plan, write down in one line how you expect to feel afterward, on a scale from one to ten. Then, afterward, write down how you actually felt. Most introverts who do this find a gap of two or three points in the same direction every time. Once you have seen your own data, the forecast loses its authority.

Give the first ten minutes away. Arrival is where nearly all the dreaded feelings are concentrated. Decide in advance that the first ten minutes will be awkward and do not count them. Judge the evening from minute eleven.

Make the smallest move first. For the online version of this, the equivalent of arrival is the first message, and the reach-out gap research says the recipient will be far more pleased than you expect. Send something short. The forecast is wrong about that too.