Nassim Nicholas Taleb wrote about the problem of the turkey. The turkey is fed every day for a thousand days. Each feeding confirms the turkey’s belief that the farmer is friendly. On day one thousand and one — the day before Thanksgiving — the turkey’s model of reality is catastrophically disproved.
A thousand confirming observations. One disconfirming event. And the disconfirming event is the only one that matters.
I have been this turkey. Not at Thanksgiving, obviously. At a corporate gala in Vienna.
The Gala
The event was the most prestigious booking I had received at that point. A financial services company, end-of-year celebration, two hundred guests, formal setting. I was booked for a thirty-minute performance between dinner and dancing. Good stage, good sound, professional lighting. Everything I could ask for.
I had prepared meticulously. My set was tested. Each routine had been performed dozens of times at smaller events. I knew the material. I was nervous — I am always nervous — but I was prepared.
The first fifteen minutes went perfectly. Strong opening, solid middle section, good audience energy. I was hitting every beat. The reactions were exactly what I had practiced for.
Then came the closer.
My closer at the time was a mentalism piece — a routine built around a prediction that I had placed in full view at the beginning of the show. The entire set had been structured to build toward this moment. Thirty minutes of narrative, three routines of escalating impossibility, all converging on this one revelation.
The prediction was wrong.
Not wrong in a recoverable way. Not wrong in a “let me show you something else” way. Wrong in a way that was immediately, visibly, undeniably apparent to two hundred people in a ballroom. The number did not match. The words did not match. The prediction, which was supposed to be the triumphant resolution of the entire evening, was simply and completely incorrect.
I stood on that stage for what felt like thirty seconds — probably three — holding a piece of paper that proved I had failed, in front of an audience that had been enthusiastically on my side until that exact moment.
The Aftermath
I recovered. Not brilliantly, but adequately. I made a self-deprecating comment, pivoted to an improvised closing moment that got a laugh, and walked off the stage to polite applause. The audience was kind. They had enjoyed the first twenty-five minutes. They were willing to forgive the ending.
But I did not forgive myself. Not because of ego — or not only because of ego. Because the failure had exposed something I needed to understand.
I spent the next week dissecting what had gone wrong. Not the method failure itself — that was mechanical, a specific procedural error that I could identify and correct. What I needed to understand was why I had not anticipated this possibility. Why, in all my preparation, in all my practice sessions, in all my rehearsals, had I never considered the scenario in which this particular element went wrong?
The answer was the same answer that Taleb describes in his analysis of financial crises, strategic failures, and catastrophic surprises: I had built my confidence on confirming evidence. I had performed this closer successfully many times. Each success had made me more confident. And that confidence had blinded me to the specific, low-probability, high-consequence failure that was always possible.
The Asymmetry of Information
Here is the principle that Taleb articulated and that I now apply to every aspect of my performing life: there is a fundamental asymmetry between confirmation and disconfirmation. Confirming evidence is abundant and cheap. Disconfirming evidence is rare and expensive. But disconfirming evidence is infinitely more informative.
A successful performance tells you: “The routine worked under these conditions, with this audience, on this occasion.” That is useful but limited. You do not know which conditions were essential. You do not know whether it would work with a different audience. You do not know what the boundaries of the routine’s reliability are.
A failed performance tells you: “The routine breaks under these specific conditions, because of this specific weakness.” That is precise, actionable, and immediately useful. You know exactly what went wrong, which means you know exactly what to fix.
The Vienna gala failure taught me the following specific things:
First, the routine had a single point of failure. One element that, if it failed, caused the entire routine to collapse. There was no redundancy, no backup, no graceful degradation. Success or catastrophe, with nothing in between.
Second, I had never rehearsed the failure state. I had rehearsed the routine many times, but always in the success configuration. I had never once practiced what I would do if the critical element went wrong. When it did go wrong, I had no prepared response. My recovery was improvised in real time, which is why it was adequate rather than brilliant.
Third, my confidence in the routine was based entirely on its track record of success — a track record that, as I discussed in previous posts, was systematically biased by confirmation bias. The routine’s track record did not prove it was reliable. It proved it had worked so far. These are different things, and the difference becomes apparent exactly when it matters most.
What the Black Swan Revealed
The black swan is not just a failure. It is a failure that reveals the structure of your assumptions. Every successful performance reinforces those assumptions. The black swan tears them open and shows you what is inside.
My assumptions, revealed by the Vienna failure, looked like this:
Assumption one: if a routine works reliably in practice, it will work reliably on stage. False. The conditions of practice and the conditions of performance are different in ways I had not fully appreciated. Stress, timing pressure, environmental variables, audience energy — all of these affect execution in ways that practice cannot fully simulate.
Assumption two: the probability of failure is low enough to be ignored. False. Low probability is not zero probability. And in a high-stakes performance context, even a low-probability failure carries enormous consequences. Risk is probability multiplied by impact. The probability was low, but the impact was catastrophic, which meant the risk was high. I had been evaluating only the probability and ignoring the impact.
Assumption three: my preparation was complete because I had practiced the routine many times. False. I had practiced the success state many times. I had never practiced the failure state. My preparation was incomplete in precisely the dimension that mattered most.
These were not obvious assumptions before the failure. They were invisible. They were built into my entire approach to preparation — embedded so deeply that I could not see them. The black swan made them visible.
The Restructuring
After Vienna, I restructured my approach to performance preparation in three ways.
First, I eliminated single points of failure. Every routine now has at least one fallback position — a way to recover gracefully if the primary method fails. This is not a compromise. It is engineering. The best bridges have redundant support systems. The best software has error handling. The best magic routines have contingency plans.
Building contingencies does not mean preparing for failure in a defeatist way. It means acknowledging that the world is unpredictable and that preparation should account for that unpredictability. Adam and I now build contingencies into every Vulpine Creations product from the design stage. It is part of the specification, not an afterthought.
Second, I started rehearsing failure states. Not just the success configuration. I now regularly practice what happens when things go wrong. What do I say if the prediction does not match? What do I do if the spectator makes an unexpected choice? What is my response if a method fails in real time?
This is uncomfortable practice. It feels pessimistic. It goes against the natural performer’s instinct to visualize success. But the value is enormous. When you have rehearsed the failure state, the failure — if it comes — is not a catastrophe. It is a fork in a road you have already mapped.
Third, I started actively seeking black swans. Not in performance — I do not deliberately sabotage my own shows. But in practice and testing, I now create conditions designed to produce failures. I perform for the most skeptical audiences I can find. I test routines under adverse conditions. I ask people to try to catch me, to look for the method, to be hostile. Each successful test under adverse conditions increases my confidence. Each failure under adverse conditions reveals a vulnerability I can fix.
This is essentially the scientific method applied to magic performance. Form a hypothesis (“this routine is reliable”), design an experiment that could disprove it (“perform it under the worst conditions I can create”), and learn from the results (“it survived” or “it broke here, for this reason”).
The Value of Catastrophic Failure
I want to be clear about something: I do not romanticize the Vienna failure. It was painful. It was embarrassing. It cost me sleep and confidence and, probably, a rebooking. I would not choose to experience it again.
But I would not undo it either. Because the information it provided was irreplaceable.
Before Vienna, I had a routine I believed was strong. After Vienna, I had a routine I knew was strong — because I had found its weakness, fixed it, and tested the fix. The difference between believing and knowing is the difference between fifty confirming observations and one disconfirming observation. The confirmations gave me confidence. The disconfirmation gave me understanding.
Gustav Kuhn writes about the value of seeking evidence that contradicts your theories. He describes how a single experiment designed to disprove a hypothesis taught him more than decades of confirming observations. The Vienna gala was my version of that experiment — unplanned, unwelcome, but infinitely informative.
The Turkey’s Lesson
Taleb’s turkey believed in the farmer’s goodness because every day confirmed that belief. The turkey had no way to anticipate the one day that would disprove it. The turkey’s model was perfect right up until the moment it was fatally wrong.
As performers, we are all turkeys to some degree. We build our confidence on streaks of success. We count our white swans. We tell ourselves that the routine is solid because it has always worked before. And we forget that “always worked before” is not the same as “will always work.”
The difference between the turkey and the magician is that the magician can choose to seek the black swan before it finds them. You can create the conditions for failure in practice. You can test under adversity. You can ask the hard questions before the audience asks them for you.
Or you can count your white swans and wait for Vienna.
One black swan. One failure. One evening where everything fell apart in front of two hundred people. That evening taught me more about performance preparation, risk management, and the limits of confidence than any workshop, any book, or any thousand successful shows.
The turkey never gets to learn from its black swan. You do. The question is whether you will seek it out deliberately or wait for it to find you on the worst possible night, in front of the most important audience, at the moment when it matters most.
I know which one I choose now. And it started with a wrong prediction on a piece of paper, held up in a ballroom in Vienna, in front of two hundred people who deserved better preparation than I had given them.