Peak-End Rule
We don't remember experiences as a whole. As Daniel Kahneman and colleagues showed, we judge them mostly by the most intense moment and the ending, even when those don't represent the rest.
Your memory is a lazy reviewer. When you look back on an experience, it doesn't replay the whole thing; it mostly keeps the best (or worst) moment and how it ended. That's the Peak-End Rule. A few examples:
VacationYour week away starts with a few uneventful days. On day four you have a fantastic time at the beach, and on the last night you treat yourself to a delicious dinner at a fancy restaurant. Months later, the beach day and the dinner are what you remember, not the quiet days.
MovieA two-hour film has some dull stretches, one incredible action scene in the middle, and an emotional, heartwarming ending. Ask yourself about it later and you'll recall the action scene and the ending far more than the slow parts.
Dental appointmentAn hour-long visit is mostly a routine checkup, apart from a brief jolt of pain when the dentist cleans a sensitive tooth. It ends with the dentist praising your oral hygiene. What sticks is the pain and the praise.
This is why the most intense or emotional moments, and the endings, tend to dominate how we remember events, people and situations, and why they quietly shape the decisions we make next.
The peak-end rule is one of the better-supported ideas in this area, but the evidence comes with an important catch: it works best for short, simple experiences studied in the lab.
Lab episodes versus real lifeThe classic studies used brief, controlled experiences such as film clips, cold-water immersion or a medical procedure. Real experiences like holidays are long and made of many different moments. When Simon Kemp, Christopher Burt and Laura Furneaux (2008) followed students on vacation, the peak-end rule was not an especially strong predictor of how happy they later remembered being. A 2019 study by Strijbosch and colleagues of complex, multi-part experiences went further: the average emotional tone across the whole episode predicted memories better than the peak and the end did, and the authors concluded the rule can't simply be generalized to real-world experiences.
Memories get rebuiltKemp's vacationers couldn't recall their day-to-day mood swings at all, which suggests that remembered happiness is largely reconstructed after the fact. If memory is rewritten that freely, a tidy two-point formula is at best a partial description.
Peaks and ends aren't the only things that countA 2022 meta-analysis by Balca Alaybek and colleagues in Organizational Behavior and Human Decision Processes, pooling 174 effect sizes, found strong overall support for the rule and for duration neglect. It also found the peak-end effect was about as large as the effect of the plain average of the whole experience. Averages matter too; peaks and ends are not a substitute for them.
Defining the "peak" is slipperyIn a short lab task the most intense moment is easy to spot. In a week-long trip or a year-long project, people may disagree about which moment was the peak, and whether the "end" is the last hour, the last day or the trip home. That flexibility makes the rule harder to test, and easier to fit after the fact.
The fair summary: for short, well-defined episodes, peaks and endings carry outsized weight in memory, and duration barely registers. For long, varied experiences, the rule is a useful nudge (end well, and make a high point) rather than a law.
Resources
The Peak-End Rule, introduced by Nobel laureate Daniel Kahneman together with Barbara Fredrickson and others, is a cognitive bias in how we remember and judge experiences (Fredrickson & Kahneman, 1993). Rather than averaging over everything that happened, people judge an experience largely by its most intense moment (the peak) and its final moments (the end). The rule connects to several psychological ideas:
Affective forecastingThe rule shapes how people predict their emotional reactions to future events. When anticipating how satisfied or happy an experience will make them, people may overweight peaks and endings (Gilbert & Wilson, 2007). Forecasting errors of this kind also help explain the hedonic treadmill, where the thrill we expect from a new thing fades faster than we predicted.
Duration neglectThis is the Peak-End Rule's close cousin, since both describe a bias toward specific moments over the experience as a whole. Duration neglect is the finding that how long an experience lasts has little or no effect on how we evaluate it overall (Kahneman, 2000).
Memory biasesThe Peak-End Rule can be seen as a memory bias, because it affects how we encode, store and retrieve our experiences. Other memory biases, such as the recency and primacy effects, likewise shape how we remember and judge events (Baddeley, 1999).
Decision-makingBecause it colours our evaluations of past experiences, the rule also steers future choices. It has been applied in consumer behaviour, healthcare and public policy to understand how people decide based on memories of what came before (Redelmeier & Kahneman, 1996). Like Kahneman's work on the framing effect, it shows that our choices follow how an experience is represented, not simply what happened.
In short, we score experiences mostly by their high points and their endings. That has real consequences for affective forecasting, memory and decision-making, and understanding it opens up ways to design experiences and interventions that improve well-being and choices.
References
- Baddeley, A. (1999). Essentials of Human Memory. Hove, UK: Psychology Press.
- Fredrickson, B. L., & Kahneman, D. (1993). Duration neglect in retrospective evaluations of affective episodes. Journal of Personality and Social Psychology, 65(1), 45-55.
- Gilbert, D. T., & Wilson, T. D. (2007). Prospection: Experiencing the future. Science, 317(5843), 1351-1354.
- Kahneman, D. (2000). Evaluation by moments: Past and future. In D. Kahneman & A. Tversky (Eds.), Choices, Values, and Frames (pp. 693-708). New York, NY: Cambridge University Press.
- Redelmeier, D. A., & Kahneman, D. (1996). Patients' memories of painful medical treatments: Real-time and retrospective evaluations of two minimally invasive procedures. Pain, 66(1), 3-8.