Picture two patients on a table, each having a colonoscopy in the early 1990s, before gentler sedation was common. One procedure is short. The other lasts three times as long, with more minutes of pain. Afterward you ask each of them how bad it was. The patient who suffered longer says it wasn't so bad, and a few years later they're the one more willing to come back.

That sounds like a mistake, but it's how memory works. The pattern was reported by Donald Redelmeier and Daniel Kahneman in 1996. The patient who suffered longer had a quieter last few minutes. Their memory kept the ending and dropped the length.

What the peak-end rule actually says

The peak-end rule is a finding about retrospective judgment. When you look back on an experience, you don't replay it and add it up. You grab two snapshots: the most intense moment (the peak, good or bad) and the final stretch (the end). Those two points stand in for the whole thing.

In Redelmeier and Kahneman's study, patients rated their pain every minute during the procedure. Afterward they gave an overall rating. The overall rating tracked the average of the worst moment and the last moment. It had little to do with how long the procedure ran or how much total pain was recorded.

The pattern has a name: duration neglect. Your memory is poor at recording how long something lasted. It records how it felt at the extremes and how it closed.

The cold-water experiment

Kahneman and colleagues ran a cleaner version in 1993. Participants put a hand in painfully cold water for 60 seconds. In a second trial they held the hand in the same cold water for 60 seconds, then kept it there for another 30 seconds while the water warmed slightly. It was still unpleasant, but a little less so.

Objectively, the second trial was worse. It was longer and held strictly more discomfort. Yet when participants were asked which trial to repeat, a majority chose the longer one. They weren't choosing less pain. They were choosing the memory with the kinder ending.

A follow-up with colonoscopy patients in 2003, by Redelmeier, Katz and Kahneman, found the same thing in a clinic. Patients whose procedure was extended by a short, milder stretch at the end rated the experience more favorably than those who got the standard version.

To be fair, the picture isn't perfectly tidy. Later work suggests duration matters more in some settings, such as when people are asked to think carefully about length or when the experience is a long one. Peak-end is a strong tendency, not a law. But it shows up often enough to change how you think about your own past.

Why memory works this way

Your brain isn't a camera. It compresses. Storing every minute of every day would be impractical, and as another piece of the memory literature argues, it wouldn't even be useful. So the system keeps what's most informative.

Intense moments get extra encoding because emotional arousal engages the amygdala, which signals the hippocampus that this one matters. Endings get extra weight for a simpler reason: they are the most recently stored, and they are what you were thinking about when the experience handed you its conclusion. Ending also works as a verdict. If it ended well, the story gets filed as "fine."

This is efficient. It also means the person you were during the experience (the one who lived through every minute) and the person remembering it disagree. Kahneman called these the experiencing self and the remembering self. When you decide whether to take another vacation, repeat a hard workout, or call back a friend after an argument, the remembering self is the one voting.

Where this quietly runs your life

Once you see it, you notice it everywhere.

A relationship can be good for years and still be remembered through one searing fight and a bad final month. A job can be mostly satisfying and get summarized by the last chaotic quarter. A holiday is judged by the best afternoon and the miserable airport at the end. The reverse also holds: a mediocre week gets rewritten as lovely because the last evening was warm.

This isn't a reason to distrust all your memories. It's a reason to know which parts of the record are reliable. The peaks and the ending are vivid and easy to retrieve, and the middle is thin. If you're deciding something important, such as whether to leave a job or forgive someone, ask yourself what the long, unremarkable middle looked like. That's the part your memory is least likely to hand over.

What it means for learning

The same rule applies to studying. Your sense of how a session went, and so whether you want to do another, is built from the hardest moment and the last one. If a session ends on a card you blanked on, your remembering self files the whole thing under "that was awful." Do that several days in a row and you stop wanting to open the deck. The material didn't get harder. The memory of studying got worse.

So the way a session ends affects whether you return to it. That matters more than it sounds, because spaced repetition only works if you keep showing up.

This doesn't mean you should skip hard material. Difficulty is useful for learning, and retrieving something you nearly forgot builds stronger memory than retrieving something easy. The point is to place the difficulty in the middle and put something kinder at the end.

Your next moves

  • End every study session on a win. In the last two minutes, review a few cards you know well, or re-answer something you got wrong earlier in the session now that it's fresher. Finish on a success, not on the card that stumped you.
  • Put your hardest work in the middle, not the end. If you can choose the order, start with something moderately hard, get to the toughest material at the midpoint, and leave the last stretch light.
  • Write a one-line verdict right after any experience you'll want to judge later. After a trip, a job, a tense conversation, jot a sentence immediately, plus one thing from the unremarkable middle. Your memory will overwrite it with the peak and the ending, and the note keeps a record of what it was actually like.
  • Before a big decision, interview the middle. If you're about to quit, break up, or commit to something because of how it ended, list three ordinary days from the stretch before the ending. See whether your verdict survives them.
  • Plan the last ten minutes of things you want to repeat. For a hard workout, a long meeting, or a difficult conversation with a child, build in a gentle finish: a cool-down, a thank-you, a small treat. It costs little and changes what gets stored.

A last thought on endings

There's something tender in the peak-end rule. It says the way you close a thing matters more than you'd expect. A kind last word can recolor an hour, and a calm last minute can change how a whole day is remembered. You can't fix a bad middle after the fact, but you often get to choose the ending.

This is the thinking behind Recall, a flashcard app that schedules reviews with FSRS, so each session stays short and ends before it turns into a slog. It works fully offline, and you can bring your Anki and Quizlet decks with you. If you'd like studying to be something you're glad to come back to, you can take a look at Recall.