People chose more pain because it ended better
Daniel Kahneman ran an experiment where the rational choice and the actual choice pointed in opposite directions.
In 1993, Daniel Kahneman, Barbara Fredrickson, Donald Redelmeier, and Charles Schreiber ran a small, deliberately strange experiment. Subjects put one hand in 14°C water — cold enough to hurt — for 60 seconds. Later, they put the other hand in the same water for the same 60 seconds, except this time the water was left in for an extra 30 seconds while it slowly warmed to 15°C.
The second trial contained strictly more suffering. Same painful minute, plus an additional half-minute that, while milder, was still unpleasant. By any measure of total discomfort, it was the worse experience. Asked which trial they'd prefer to repeat, most subjects chose the longer one anyway.
Kahneman's explanation was that people don't evaluate experiences by totaling them up. They remember a small number of snapshots — usually the most intense moment and the final moment — and build their verdict from those two points, more or less ignoring everything else, including how long the whole thing lasted. He and Redelmeier found the same pattern later studying real colonoscopy patients: how they rated the procedure afterward tracked the pain at its peak and at its end, almost independent of the procedure's actual duration.
The uncomfortable implication isn't really about cold water or colonoscopies. It's that your memory of anything — a trip, a relationship, a year — is being written by an editor who cares mostly about the worst moment and the last one, and is quietly discarding the rest of the manuscript.