There is a performance I do not talk about much.
This is not because I have forgotten it — the opposite problem, if anything. It is because the gap between what I had hoped it would be and what it actually was remains, even now, uncomfortable to sit with. I went in with a plan. The plan collapsed. The recovery was partial and halting. The audience was polite in the specific way that polite audiences are when they are being kind to a performer who is visibly struggling.
For a period after that night, my response was to not think about it. This is the identity-protection mechanism that Carol Dweck’s work describes so precisely: the fixed mindset treats failure as information about who you are rather than information about what happened, and the natural response to information about who you are is to look away from it.
The event identity — “the night I failed badly in front of real people” — was not useful. But the data inside that event was extraordinarily useful, once I was ready to look at it.
What Dweck Says About Failure
The central distinction Dweck draws in “Mindset” between fixed and growth orientations is clearest at the moment of failure. This is the stress test where the orientation reveals itself.
In the fixed mindset, failure is a statement about the self. If you fail at something, it is because you are the kind of person who fails at this kind of thing. The failure is not an event that occurred — it is a revelation of your fundamental nature. And the natural response to revelations of your fundamental nature that you find threatening is to protect yourself from them: avoid, deny, rationalize, stop exposing yourself to situations where the revelation might recur.
In the growth mindset, failure is a data set. Something happened, and the question is: what does this specific failure tell me about what specifically went wrong, and how does that information change what I do next?
The difference sounds simple. The behavioral consequences are substantial.
What the Failure Was
I will not reconstruct the full event — partly because of the genuine discomfort, and partly because the specifics matter less than the anatomy of the failure.
What I can say is that the failure had identifiable causes, and those causes were not random bad luck. In retrospect — and this is where the data becomes valuable — the causes were visible beforehand. I had made decisions in the preparation and design of the performance that left me vulnerable to exactly the things that went wrong.
The material was calibrated for an audience I had imagined rather than the audience I would actually encounter. The pacing assumed a level of engagement that the room did not have. There were moments where I needed the audience to do something and they did not do it, and my contingency planning for those moments was inadequate.
None of these were unforeseeable. They were foreseeable and I had not fully prepared for them. The failure was, in that sense, honest feedback about the preparation.
The Anatomy of the Data
When I was ready to look at the failure as data rather than verdict, I structured the analysis the way I would structure any post-mortem in a consulting context. What happened? What was supposed to happen? Where was the gap? What caused the gap?
The answers were uncomfortable but clear.
The audience engagement problem traced back to a decision I had made about how to open the performance. I had chosen an opening that worked very well in certain room configurations and very poorly in the configuration I encountered that night. I had not built flexibility into the opening to adapt to different configurations. When the room was set up in a way that undermined the opening, I had no prepared alternative.
The pacing problem traced back to overconfidence in my estimation of how fast a cold audience would warm up. I had performed the same material for warmer audiences and calibrated my timing based on those experiences. The cold audience needed more time, and my pacing did not give it to them.
The contingency problem traced back to insufficient preparation for things going wrong. I had prepared for the performance to go well. I had not prepared adequately for the performance to go differently than expected.
These are all addressable. None of them are “you are not the kind of person who can do this.” They are “these specific things went wrong in these specific ways for these specific reasons.”
What Changed After
The changes I made after that failure were more specific and more durable than any changes I made after successful performances.
I redesigned the opening section of my standard performing set to include explicit flexibility — not a single opening that I hoped would work, but a family of openings that I could choose between based on real-time assessment of the room. This took significant additional preparation time. It was completely worthwhile.
I recalibrated my pacing sense by deliberately performing in a wider range of room types and paying specific attention to the warm-up arc — how long different audiences needed before they were genuinely engaged. I developed explicit markers for when a room was ready and practiced adjusting my timing based on those markers.
I built a much more systematic contingency practice. Not just “what do I do if this specific thing goes wrong” but “what is my general principle for responding when something goes unexpectedly, and how do I maintain forward momentum while adapting?”
None of these changes would have happened without the specific failure that made them necessary. The successful performances had not surfaced these weaknesses. They had concealed them. The failure exposed them precisely.
The Paradox of Successful Performances
There is a counterintuitive truth here that took me a while to articulate.
Successful performances can be the enemy of genuine improvement. When a performance goes well, the feedback is positive but often imprecise. You know that something worked. You do not always know exactly what worked, or whether it worked because of what you did or despite what you did, or whether it would work under different conditions.
The failed performance produces much more precise feedback. The things that went wrong are visible. Their causes are often traceable. The gap between what you planned and what happened is concrete and analyzable.
The successful performance tells you “something is working.” The failed performance tells you “this specific thing failed in this specific way for this specific reason.” The second type of information is more actionable.
This does not mean seeking failure — deliberately performing in conditions where you are likely to fail is not the right strategy. But it means that when failure occurs, treating it as information rather than verdict is not just the psychologically healthier response. It is the strategically correct one. The failure contains the most specific and usable feedback available.
The Identity Trap
The reason the fixed mindset defaults to treating failure as verdict rather than information is that the identity protection instinct is genuinely powerful. The failure feels personal because performance is personal. You are not presenting a spreadsheet — you are presenting yourself, your preparation, your craft, your ability to do this thing you claim to be able to do.
When that goes wrong, the threat is not just to your plan. It is to your self-concept as someone who can do this. And the mind’s response to threats to self-concept is defensive.
Dweck’s contribution is not to tell you that the defense is wrong. It is to offer a framework in which the defense is unnecessary — because failure is not a threat to self-concept when failure is understood as information about events rather than revelation of identity.
The performance failed. I did not fail as a person. The conditions were such that my preparation was insufficient for what was required. That is the accurate description. It is also the useful one, because it points directly at what can be fixed.
The Most Valuable Data
If I could choose to have had the bad night in exchange for never having to go through it — if the offer were on the table — I would not take it.
The changes that came from that failure improved every performance that came after it. The flexibility I built into my material, the recalibration of my timing sense, the contingency planning that became a genuine system rather than a vague intention — these are present in everything I do now.
The comfortable performances did not produce those changes. The failure did.
The worst performance of the journey turned out to be among the most instructive events of the journey. Not because failure is good, not because I am glad it happened — but because the information it contained was precise, honest, and completely unavoidable.
The growth mindset is not about being cheerful about failure. It is about being clear-eyed about what failure is and what it contains. Failure as data is not comfortable. But it is the most direct feedback the craft gives you.
Pay attention to it.