There’s an experiment by Kahneman that I’ve thought about almost every time I structure a performance since I read it.
Participants were asked to submerge one hand in painfully cold water for sixty seconds. Then they were asked to do it again, but this time for ninety seconds — with the last thirty seconds being slightly less cold, still uncomfortable but noticeably less painful than peak.
Which experience did they prefer to repeat?
Most people chose the ninety-second version. The one that was objectively more total pain. Because it ended less badly.
This is the peak-end rule. We don’t judge experiences by averaging them. We judge them by two moments: the moment of greatest emotional intensity (the peak), and the final moment (the end). Everything in the middle, weighted by its duration, barely registers in memory.
What This Means for a Show
I’ll tell you what I was doing before I understood this.
I was structuring shows the way I’d learned to think about professional presentations. Build momentum. Pace yourself. Save energy for the end. Create a nice arc with highs and lows. Deliver consistent quality throughout. Don’t peak too early. Don’t burn your best material on the opener.
This is perfectly sensible advice in many contexts. A presentation that starts with your strongest arguments and then sags is a weak presentation. A business proposal that front-loads all the compelling material and then delivers fifty slides of supporting detail loses the room.
But magic performance isn’t a business presentation. People don’t experience it in the analytical, evaluative mode of a presentation. They experience it emotionally, narratively, and they remember it according to the peak-end rule.
This means the “save your best for later” advice that applies to presentations actively hurts magic shows. If you hold back your most astonishing material and deliver solid, consistent magic throughout, what the audience remembers is: the best thing that happened, and however it ended. The solid consistent middle? Gone from memory. Averaged away.
The implication is radical: it doesn’t matter how good the middle is. Only two things matter in determining how people remember your show. The single most astonishing moment. And the last thing they experience.
Redesigning Around the Rule
When I first really absorbed this, I sat down with a sheet of paper and graphed the emotional intensity of my current show sequence.
What I found was a relatively flat line with a slight incline toward the end. Consistent quality throughout. No dramatic peaks. A reasonably strong closing but not dramatically stronger than what came before.
By the peak-end rule, that’s a mediocre experience. Not because the material was bad, but because there was no single moment of maximum intensity for memory to anchor on. The experience would average out to “that was good” without any vivid recollection of what was good.
The fix wasn’t to improve the middle. The fix was to create a genuine peak.
Not a point in the show that was better than what came before. A moment of genuine maximum emotional intensity. The thing that, if everything else in the show disappeared from memory, would still make someone say: “there was that moment where…”
And then — separately — to make the ending land with full weight.
Those are different problems. The peak is about intensity. The ending is about resonance. You can have a shattering peak that’s followed by a weak ending and the experience will feel unsatisfying — the peak was wonderful but the landing was soft. You can have a mediocre peak followed by a perfect ending and the experience will feel better than it has any right to.
The Opening Question
This raises an interesting question about openings.
The peak-end rule seems to say the opening doesn’t matter much. It’s not the peak (you’ll build to that) and it’s not the end (obviously). So what do openings do?
I’ve come to think that openings establish the possibility of the peak. They set the audience’s emotional baseline. They determine how much room there is to rise.
If your opening is flat — polite applause, perfunctory setup, nothing that engages emotionally — the audience is starting from a low baseline. Even a strong peak has limited amplitude from that floor.
If your opening creates genuine engagement, curiosity, and warmth — if people are already leaning in, already invested, already feeling something — then the same peak moment registers more intensely because it’s rising from a higher floor.
The opening doesn’t get remembered, but it shapes what gets remembered. That’s a different and important function.
There’s also a practical consideration: the opening determines whether people will be present for the rest of the show. An audience that checks out in the first three minutes because nothing grabbed them is an audience that watches the rest with divided attention. Even your peak will land softer with an audience that isn’t fully there.
Open strong enough to get full presence. Then build to the peak. Then nail the ending.
The Pain Experiment’s Uncomfortable Implication
Here’s the part that made me most uncomfortable when I thought about it.
The cold water experiment showed that people prefer ninety seconds of pain ending slightly less badly to sixty seconds of pain ending at peak intensity — even though sixty seconds of pain is objectively less total suffering.
Applied to magic: people may remember a three-hour show that peaked brilliantly and ended perfectly as better than a tightly crafted forty-five minute show that was consistently strong but ended ordinarily.
Duration, in the peak-end model, is nearly irrelevant. A longer show with the right peak and ending wins over a shorter show with a flat emotional profile, even if the shorter show had better average quality.
I’m not recommending padding shows with mediocre material to get more time for the peak. That’s a misreading. But it does suggest that a single extraordinary moment, placed correctly and followed by an ending that truly lands, will do more for how people remember you than three hours of solid, consistent, impressive work.
The corollary for corporate keynote work — where I often have a specific time slot and specific outcomes the client wants — is that I should identify the one moment I want them to remember and make sure that moment is as extraordinary as I can make it. Everything else is in service of that moment and the landing.
The presentation you gave us. The examples you shared. All of that is available for memory. But what they’ll actually remember? The peak. And how it ended.
Duration Neglect as Practical Wisdom
Kahneman calls the phenomenon “duration neglect” — the puzzling fact that the duration of an experience has almost no effect on how we remember it.
Colonoscopies (which Kahneman also studied in this context — he had eclectic taste in research topics) that ended on a slightly less uncomfortable note were remembered as less bad than shorter ones that ended at peak discomfort. The extra time of the longer procedure didn’t worsen the memory. The ending did all the work.
For a performer, duration neglect is actually liberating. You don’t have to be extraordinary for the full duration of your show. You have to be extraordinary at the right moments — primarily the peak and the ending — and acceptable everywhere else.
The pressure isn’t on maintaining peak quality throughout. The pressure is on identifying, designing, and executing the peak and the ending with everything you have.
This changes what “difficult moments” means in a show. A weak middle section is a missed opportunity but not a memory-damaging event. A weak ending is a catastrophe. A strong peak moment that gets undercut by a stumbling close will leave audiences with a vague feeling that something was off, even if they can’t identify what.
I’ve experienced this from the other side as an audience member. I’ve watched shows with extraordinary moments that ended flatly, and walked away with a complex, slightly unsatisfying feeling. The extraordinary moment was there. I remembered it. But the ending left me with a low final emotional state, and that colored everything.
Close stronger than you’ve ever closed anything.
That’s not a motivational platitude. It’s the cognitive science of memory. The last thing the audience experiences with you is the thing most available to their memory when they try to reconstruct what happened. If it’s powerful, resonant, and complete, they’ll borrow that emotional intensity backward across the whole experience.
If it isn’t — if you drift to a close, wrap up perfunctorily, thank them and step back — the show ends. Not with a landing. Just with a stopping.
And the peak-end rule says: that’s how they’ll remember it.
Not the extraordinary moment you worked so hard to build. The drift at the end.
Peak and end. That’s the whole show, in terms of memory.
Design accordingly.