The peak-end rule: two moments become the whole day
Your memory of today will not be an average of it. Under the peak-end rule two moments do nearly all the work, and one of them is the last one.
· 3 min read · The Grateful Team
Eleven hours of an ordinary Tuesday went fine. Then a message at 6pm landed badly, and by the time you were brushing your teeth the whole day had a colour.
Ask how the day was and you do not replay eleven hours. You replay four minutes.
Memory keeps snapshots, not averages
Your memory of an experience is built from a few moments of it, not from an average of the whole thing. In two experiments published in the Journal of Personality and Social Psychology in 1993, Barbara Fredrickson and Daniel Kahneman had people watch pleasant and unpleasant film clips of varying length, rate how they felt moment to moment, then evaluate each clip once it was over.
How long a clip ran barely showed up in the verdict. What people kept tracked a handful of moments instead. The researchers called it duration neglect: a retrospective evaluation behaves like a weighted average of snapshots, as if length did not count.
Later work named the pattern the peak-end rule. It is why a long wait that finishes well can be remembered more kindly than a short one that finishes badly.
The ending is the part still in your hands
The end of a day is the one snapshot you still get to pick. The peak usually is not up for negotiation — the message arrived, the meeting went how it went. The last few minutes before sleep are different, because nothing happens in them except whatever you point your attention at.
Most people spend those minutes scrolling, or running the 6pm message one more time. That makes the loudest moment the closing moment too, and the whole day gets filed under a single heading.
Writing three specific good things is a cheap way to put something else in that slot. Specific, not summary: the chai someone made without being asked, the walk home in good light, the thing you finally sent. The three-good-things exercise that Seligman and colleagues tested in 2005 is exactly this, and it is the loop Grateful' is built around — three lines, then the app stops asking. A paper notebook does the same job. If you would rather be asked a question than face a blank page, start with our prompts for beginners.
This is not the same as pretending
Nothing in the peak-end rule says a bad day was secretly a good one. The 6pm message still happened, and three lines do not unhappen it. What changes is which moments are still on the shelf when your brain reaches for Tuesday next week.
Left alone, the bad moment wins that competition by default — unpleasant events hit harder and stay longer than pleasant ones of the same size. Writing something down at the end is not a correction. It is a second entry in a race that otherwise runs unopposed.
If Monday is the day the pattern bites hardest, Monday quotes is a smaller version of the same move, and reader stories are what this looks like across a real week rather than a study.
Tonight, before you put the phone down, write the three most specific good things you can still picture from today. Not the three best ones. The three you can see.
Research behind this page
- Fredrickson, B. L., & Kahneman, D. (1993). Duration neglect in retrospective evaluations of affective episodes
- Seligman, M. E. P., Steen, T. A., Park, N., & Peterson, C. (2005). Positive psychology progress: empirical validation of interventions
Common questions
- What is the peak-end rule?
- The peak-end rule describes how people judge a past experience by its most intense moment and by how it ended, rather than by its average or its length. A 1993 study by Fredrickson and Kahneman found duration barely affected how an episode was remembered.
- Does writing at night change how you remember the day?
- No study has tested that directly. What the research shows is that the final moments of an experience carry outsized weight in how it gets remembered, so what you pay attention to last is not neutral. Three specific lines put something concrete in that slot.
- What should I write on a bad day?
- Write something small and true rather than something meaningful. A hot cup of tea, one message that landed well, ten quiet minutes. On a bad day the point is noticing that anything survived it, not building a case that the day was good.
One good thing, weekly
Free 7-day gratitude starter pack — one prompt a day, no app required.
Keep reading
Why gratitude journaling works: the one bad review
Nine people liked it, one was lukewarm, and only the lukewarm one stuck. That asymmetry is the whole reason gratitude journaling works.
The week I stopped trying to be profound
Four days of writing things I thought sounded meaningful, then one night of writing that the tea was hot. Only one of those weeks continued.
Three lines on a genuinely bad Tuesday
The day did not improve. The entry was not a correction. It was just the last thing I looked at instead of the message at 6pm.
24 Gratitude Journal Prompts for Beginners
Twenty-four beginner gratitude journal prompts, ordered easiest first, with the two rules that decide whether the habit survives its second week.