The verdict forms faster than thought
In a now-famous set of studies from Princeton, the psychologists Janine Willis and Alexander Todorov showed people photographs of strangers' faces for as little as a tenth of a second and asked them to rate qualities like trustworthiness, competence, and likeability. The startling finding was not that people made snap judgments, but that the judgments they made after 100 milliseconds barely changed when they were given more time to look. Extra seconds increased their confidence; they did not much alter the verdict. The first impression was, in effect, already complete before anyone had consciously decided anything.
This is worth sitting with, because it reframes what a first impression actually is. It is not a considered opinion that people gradually assemble. It is a reflex, closer to flinching than to reasoning. By the time you walk into a room and feel someone size you up, the sizing has mostly already happened. What that reflex is responding to is not your character or your history, neither of which is visible. It is responding to your face, your posture, and the set of your expression in that first frozen instant.
Where the snap judgment gets it right
First impressions are not pure noise. Decades of research on "thin slices" of behavior, much of it associated with the psychologists Nalini Ambady and Robert Rosenthal, found that brief observations of a person, sometimes only a few seconds of silent video, could predict certain outcomes with surprising reliability. Strangers watching short clips of teachers could roughly anticipate end-of-semester student ratings. People are genuinely picking up real signal from expression, energy, and the micro-movements of the face, much of which is catalogued in Paul Ekman's Facial Action Coding System, which maps how specific muscles produce specific looks.
The honest accuracy of a first impression tends to be highest for exactly the thing it is built to read: how a person is coming across in this moment. Are they warm or guarded right now? Tense or at ease? Engaged or checked out? These are projection questions, and faces broadcast them loudly. A genuine smile that reaches the eyes reads differently from a polite one, and most people detect the difference without being able to name it. To that narrow extent, first impressions are not a superstition. They are a fast, mostly automatic reading of present-moment signal.
Where it goes badly wrong
The trouble starts when the brain treats a reading of the present moment as a verdict on the whole person. Edward Thorndike named this slippage the halo effect nearly a century ago: when we judge someone positively on one visible trait, we tend to assume a host of unrelated good qualities follow. The person who looks competent is also assumed to be honest, kind, and capable, none of which a photograph can possibly reveal. The mechanism that makes first impressions fast is the same one that makes them overreach.
Two further problems compound it. The first is stickiness: because of confirmation bias, once an impression forms, we unconsciously gather evidence that confirms it and explain away evidence that contradicts it, so a wrong first read can survive contact with the truth for a long time. The second is that the signal is a snapshot. A flat expression after a bad night, a guarded posture in an unfamiliar room, a tired face at the end of a long day, any of these can be misread as a fixed trait rather than a passing state. First impressions are accurate about the frame and frequently wrong about the film.
What this means for your own face
If a verdict is forming in a tenth of a second, the practical question is not whether that is fair. It plainly is not. The useful question is what your own face is putting into that tenth of a second before you have had any chance to speak. Most of us have never actually seen this. We know our face from the mirror, where we are always slightly performing, and from photos, where we have usually composed ourselves. We rarely see how we come across in a resting, unguarded frame, which is precisely the frame strangers judge.
That blind spot is the thing worth closing. You cannot control the reflex other people have, but you can find out what it is responding to, and you can adjust the small, real signals, the set of the jaw, the openness of the brow, the warmth or absence of warmth around the eyes, that tilt the read one way or another. Understanding first impressions is not about gaming people. It is about no longer being the only person in the room who has not seen what your face is saying.