The short answer
Replaying a conversation after it's ended — turning over what was said, what a pause meant, what you should have said instead — has a specific name in the research literature: post-event processing, and it describes exactly this pattern: a detailed, often distorted review of a social interaction that continues well after the interaction itself is over, frequently on the drive home or lying in bed that night.
It's not a character flaw or a sign of low self-esteem specifically. It's a documented feature of how a certain kind of mind processes social information, especially under some degree of social anxiety.
The research
'Post-event processing' names the pattern of replaying a social interaction after it's over — a detailed, often distorted review that continues well after the interaction itself is over.
— Clark and Wells, cognitive model of social anxiety, 1995
Why the replay runs longer than the event
Part of what makes post-event processing distinct from ordinary reflection is duration and distortion. Ordinary reflection tends to be brief and roughly accurate — a quick mental check that things went fine. Post-event processing tends to run much longer than the event itself warranted, and the version being replayed skews more negative than what an outside observer of the same conversation would report. The replay isn't a neutral recording being reviewed; it's being actively re-edited each time, usually toward a worse interpretation.
That's a specific and testable claim from the research, not just a description — several studies comparing self-report of a social interaction with an independent observer's rating of the same interaction have found a consistent gap, with the person who was in the conversation rating their own performance as worse than outside observers did.
Why it clusters around specific kinds of moments
Post-event processing isn't evenly distributed across every conversation. It tends to concentrate around interactions with some ambiguity in the outcome — a joke that got a smaller laugh than expected, a silence that lasted a beat too long, a question that could have been read as genuine curiosity or as pointed criticism. Clean, unambiguous interactions (a warm goodbye, clear positive feedback) rarely get the same replay treatment. It's the ambiguous ones that give the mind material to work with.
Why it's not about weakness or low confidence
It's worth separating this pattern from confidence broadly. A person can be genuinely skilled socially and still experience heavy post-event processing — the mechanism is about how ambiguous information gets processed afterward, not about actual social competence in the moment. That distinction matters because a lot of advice aimed at this pattern assumes the fix is building confidence, when the more precise target is the replay mechanism itself.
Want to see this pattern in your own thinking, not just in the research?
What this hub covers
Post-event processing shows up in a lot of specific contexts — after a date, after a disagreement, after a work meeting where the wrong word slipped out — and it often overlaps with a related pattern: the tendency to please or appease in the moment, sometimes called people-pleasing, which can leave more ambiguous material behind to replay later, since agreeing or deferring in the moment often means real feelings never got said and remain unresolved. The linked page below goes deeper into that specific pattern.
What tends to help
Because the replay is a re-editing process, not an accurate playback, arguing with the content of any single replay tends not to work — there's always a slightly worse version available to replace it with. What research on this pattern suggests is more useful is noticing when the replay starts and how long it runs, rather than trying to win the argument happening inside it. Over time, that timing data — not the content of any one replay — tends to be what makes the pattern feel less permanent.