What You Remember Is the Peak and the End

Two people run the same brutal half-marathon on the same hot morning. Same course, same hills, finish times within a minute of each other. One limps through the final stretch cramping and dizzy, hating every step, desperate for it to be over. The other catches a second wind in the last mile, passes a few people, and floats across feeling strong. Ask them a month later whether they would do it again. The first says never. The second is already eyeing the next race. They went through almost the identical event. They carry almost opposite memories of it, and it is the memory, not the event, that decides what each does next.

That gap is one of the sturdier findings in the psychology of memory. We do not store an experience as an average of every moment it held. We store a compressed version built mostly from two snapshots: the emotional peak, the single most intense moment, and the ending, how the thing felt as it closed. Daniel Kahneman named the two-part shortcut the peak-end rule, and its stranger companion, the fact that how long the experience lasted barely registers, duration neglect.

The ice water that people chose to relive

The cleanest demonstration is almost uncomfortable to describe. In a study led by Kahneman with Barbara Fredrickson, Charles Schreiber and Donald Redelmeier, people sat through two versions of a cold-water ordeal, one with each hand. The short version was a hand in water kept at 14 degrees Celsius for 60 seconds, which is genuinely painful. The long version was the same 60 seconds at 14 degrees, then 30 extra seconds while the water was warmed very slightly, to around 15 degrees. Still cold, still unpleasant, but distinctly less awful than before.

Look at what the long version actually is. It is the entire short version, every second of it, plus half a minute of additional discomfort. By any honest accounting it contains more total pain. Then people were told they would repeat one of the two trials a third time and asked to choose which. A significant majority chose the long one. They volunteered for more suffering because it ended on a gentler note, and that gentler ending was what their memory had filed under "not as bad."

The sentence separates two things we usually treat as one: the experience you have, moment by moment, and the memory you keep of it. They can disagree, and when the time comes to decide whether to do something again, the memory wins. You do not choose the future based on what happened. You choose it based on the story of what happened.

The peak and the last few minutes

Cold water in a lab is one thing. Kahneman and Redelmeier went looking for the same pattern where the stakes were real, and found it in a hospital. In 1996 they recorded pain in real time during two genuinely unpleasant procedures, colonoscopy in 154 patients and lithotripsy in 133, having patients report how much it hurt at regular intervals throughout. Afterward, each gave a single overall judgment of how bad the whole thing had been.

The remembered total did not track the sum of the suffering. It tracked two features of the real-time record: the peak, the worst moment, and the pain during the final minutes. Both mattered, and the study does not rank one above the other. What barely mattered was duration. Some procedures ran far longer than others, meaning far more total discomfort, and yet the long ones were not remembered as meaningfully worse. Kahneman and colleagues found that the memory of an ordeal is built from its most intense moment and its ending, and is remarkably indifferent to how long the ordeal actually went on.

That indifference runs against intuition. It feels obvious that a longer bad experience should be remembered as worse. It usually is not. A rough forty-minute procedure that eased toward the end can be recalled more kindly than a sharp ten-minute one that stopped at its worst.

What happens when you engineer the ending

If memory is built from the peak and the end, you could in principle change a memory by changing only the end, leaving everything before it untouched. In 2003, Redelmeier, Katz and Kahneman tested exactly that in a randomized trial of 682 colonoscopy patients. Half received the standard procedure. For the other half, the doctor added a short, deliberately mild extension at the very end, leaving the instrument still for a moment so the final sensation was gentle rather than sharp.

Everything up to that point was the same. The extension only added time and mild sensation; it removed none of the pain that came before. And yet the group with the gentler ending remembered the final moments as milder, 1.7 versus 2.5 on a ten-point scale, and rated the entire experience as less unpleasant, 4.4 versus 4.9 on a ten-centimeter scale. Those are average ratings on the scales, not percentages. A softer last minute reached back and recolored the whole thing.

The effect even reached into behavior, though gently. Over years of follow-up, the rate of returning for a repeat colonoscopy averaged 50.4 percent across both groups, with the gentler-ending group slightly more likely to come back. The nudge was modest, but the direction is the point. The added stretch worked because it was less painful, not because pain was piled on; a harsher ending would have done the reverse. It is the softness of the close, not the extra time, that rewrote the memory and tilted the choice to repeat.

A shortcut, not a law of everything

Before we carry this to goals, one honest boundary. The peak-end rule and duration neglect are strong, well-replicated findings for remembered pain and aversive episodes, not a universal law of how the mind remembers everything. The effects weaken, and can reverse, when duration is made salient, over very long real-life stretches, and for some pleasant experiences rather than painful ones. So treat it as a powerful tendency, not a guarantee. For emotionally charged episodes, the ones that leave a mark, how it peaked and how it ended do an outsized share of the remembering. A goal pursued over months is exactly that kind of episode.

Your goal is an experience you will remember

Here is where this stops being trivia. A goal is not a spreadsheet of progress. It is a long, emotionally charged experience, and the day you finish it, the whole thing collapses into a memory built from the same two snapshots: the hardest moment, and how it felt at the end.

We have written before about the messy middle, the sag of motivation that hits partway through, and that is usually where the peak of difficulty lives. You cannot fully control that part; hard goals are hard. What you have far more say over is the ending, which does at least half the work of the story you carry forward.

This matters because the memory of the last goal is the recruiting pitch for the next one. Kahneman's participants chose to relive the ice water based on how it was filed, not on how it went, and you do the same with goals. If a stretch ends in grinding exhaustion, a limp over the line, and an immediate slump into the next demand, that memory gets saved, and it argues against ever starting anything ambitious again. If the same stretch ends with a moment of acknowledgment, a real look at how far you came, the identical months get filed as worth repeating. Same work, opposite conclusion about who you are and what to do next.

This is a different point from journaling in general, which we take apart in journaling for goals. Day-to-day journaling shapes the experience while you are inside it; this is about the ending specifically, the deliberate closing act. That act is not a footnote to the goal. It is a large fraction of what the goal will have meant.

How to end a goal on purpose

The practical move is to refuse to let a goal simply stop. Most goals do not end so much as trail off. The last task gets checked, attention is already elsewhere, and the whole thing dissolves without a close. That is an ending by default, and defaults tend to be flat or sour.

Build a real ending instead. When you finish, stop and take stock: what actually changed, what the hardest moment taught you, what you would tell yourself back at the start. Name the win before the next thing swallows it. The aim is not to lie about how hard it was, the peak was real, but to make the close a considered, generous moment rather than an exhausted shrug. That is the whole distance between the 14-degree ending and the 15-degree one, and over a goal that took months, it is the difference between a memory that recruits you and one that warns you off.

Future You is built around that closing act. When you complete a goal, the app does not just mark it done; it walks you through a completion reflection, then surfaces a unique reflection archetype that captures the shape of your particular journey, so the ending is deliberate and favorable instead of a fade-out. That is engineering the end on purpose, the way the gentler colonoscopy did, so the memory you carry into the next goal makes you more likely to begin it. See how the pieces fit together on our features page. Future You is free on iOS and Android.