Why People Prefer More Pain

People often choose to repeat more painful experiences because memory of discomfort is shaped less by total duration and more by how the experience peaks and how it ends. In a cold-water hand experiment, participants rated one hand’s discomfort in real time while holding it in ~14°C water for 60 seconds, then compared it with a second trial that added 30 more seconds—during which the water warmed slightly to ~15°C. Even though the longer trial meant 50% more time in discomfort, many people still preferred to repeat it, suggesting that a “less bad” ending can outweigh a longer stretch of pain.

The replication followed a classic setup tied to Daniel Kahneman and Barbara Fredrickson’s 1993 findings. Participants were told the study concerned pain differences between dominant and non-dominant hands, but the real manipulation was the length of exposure and the ending temperature. Both trials began identically, yet the longer one concluded with a small reduction in discomfort—participants described the sensation as “backing off,” with stinging and numbness easing slightly. When asked which trial they would choose to repeat, 7 of 12 people in this small replication favored the longer trial; the original study found an even stronger preference, with nearly 70% choosing the longer experience. The key pattern held: the trial that lasted longer but ended a bit more tolerably was remembered as the better overall experience.

Why does that happen? Kahneman and Fredrickson’s follow-up work points to “duration neglect,” where the experiencing self and the remembering self process time differently. While someone going through unpleasant discomfort may want to stop immediately, retrospective judgments often ignore how long the ordeal lasted. Instead, memory tends to summarize experiences using a few standout moments. Research on pleasant and unpleasant video clips found that length had little impact on retrospective ratings, while emotional intensity and salient moments mattered far more.

That shortcut is closely linked to the representativeness heuristic: people treat the most vivid “peak” moments as representative of the whole event. The transcript connects this to classic Kahneman and Tversky probability errors—like the Linda Problem—where descriptions that “fit” a mental stereotype feel more likely even when logic says otherwise. In the pain experiments, the peak discomfort and the final easing become the mental “photographs” that stand in for the entire episode.

Recency bias further amplifies the effect of endings. More recent moments are easier to recall, so the conclusion of an experience can disproportionately color overall evaluation. Experiments using fictional life stories about “Jen” showed that adding extra years that were pleasant but not as good as the early peak could reduce perceived life quality—because the ending and the most memorable segments dominate the final judgment.

The practical implications are significant. If people remember medical procedures more favorably when discomfort diminishes at the end, they may be more willing to return for follow-up care. A 2003 colonoscopy trial with 682 patients found that adding three minutes where the colonoscope tip stayed in place (uncomfortable but less painful than earlier) made the overall experience about 10% less unpleasant and increased the likelihood of coming back. The same peak-end logic shows up in everyday design choices—like a treat after a checkout experience—because memory, not raw total discomfort, drives what people want to repeat. The takeaway is blunt: optimize for the peak and the end, not just the duration.