The first three dates, and what they can tell you

People treat the first few meetings as an assessment exercise. The evidence on how well that assessment works is not encouraging, and knowing that improves the process.

What early judgement gets right

Research on thin slices — judgements made from very brief exposure — finds people are reasonably accurate about some things quickly.

Extraversion is judged relatively accurately. So, to a lesser extent, are conscientiousness and general warmth in interaction.

People also form rapid impressions of attractiveness and of trustworthiness, and those impressions are consistent across observers, which is not the same as being correct.

What it gets wrong

Almost everything that matters over years.

Emotional stability, openness and how someone behaves under stress are judged poorly from brief interaction.

How someone treats people when things go badly, how they handle disagreement, whether they follow through, and how they behave when not being observed by a person they want to impress — none of this is available in a first meeting, and all of it determines whether a relationship works.

There is also a structural problem. Both people are performing. Early dates sample behaviour under conditions of maximum effort, which is precisely the least representative sample available.

The compatibility prediction problem

Worth stating because the dating industry rests on the opposite claim.

Large studies attempting to predict relationship satisfaction from pre-relationship characteristics of both people have found this to be very difficult.

One well-known analysis combining data from many couples found that individual traits predicted a small amount of variance, that the combination of two people's traits added almost nothing beyond that, and that most of the predictable variance came from how people already felt about the relationship.

Which is a substantial result. It suggests that matching on characteristics — the premise of most algorithms — has limited value, and that relationships are made more than found.

What early meetings are actually good for

Reframed accordingly.

Whether you want to see them again. A low bar, and it is genuinely the question a first meeting answers.

Whether the conversation moves. Not whether it was impressive, but whether it went somewhere neither person planned.

Whether you were performing the whole time. Sustained performance is a reasonable early signal that something is off, either about the match or about how you are approaching it.

Basic conduct. How they treat staff. Whether they arrive when they said. Whether they ask anything about you.

These are low-information signals individually and they are the ones that are actually observable.

The things worth noticing early

Not because they are predictive of compatibility but because they are difficult to change.

How they talk about previous partners. A pattern in which every former partner was unreasonable is worth registering.

Whether disagreement is possible. A small difference of opinion in a first conversation is informative, and its complete absence over several meetings is also informative.

Whether they can be uncertain about anything.

Practical incompatibilities on things neither person will move on — children, location, religion. These are worth raising earlier than feels comfortable, because they do not resolve with attachment and the cost of discovering them at eighteen months is high.

The pace question

Early intensity is a poor predictor of anything.

Rapid escalation of feeling in the first weeks is common, pleasant, and unrelated to whether the relationship lasts. It reflects novelty and uncertainty, both of which are temporary by definition.

Which argues for a specific discipline: making no large decisions on the strength of the first two months, and treating the intensity as enjoyable rather than as evidence.

The reasonable approach

Meet, cheaply and without ceremony. Decide only whether to meet again.

Repeat until you have enough ordinary, unperformed time together to have seen how the other person actually behaves — which takes months rather than dates, and cannot be shortened by better questions.

Most of the anxiety around early dating comes from treating a low-information encounter as a high-stakes evaluation. It is not one, and it cannot be made into one.