The first time I saw it, I almost could not believe it was me.
I was sitting in a hotel room in Graz, laptop open on the desk, watching footage of a corporate performance I had given the previous evening. A technology company had hired me to do a keynote with some magic woven in — my usual format, the kind of engagement where I feel most at home. The audience had been responsive. The event organizer had shaken my hand afterward and said the feedback was excellent. I had walked away feeling genuinely good about the performance.
Then I watched the video.
The first thing I noticed was my left shoulder. Every time I reached a moment in the routine where I needed to do something the audience was not supposed to focus on, my left shoulder climbed about two centimeters toward my ear. Not a dramatic movement. Not something anyone in the audience would have consciously registered. But on video, playing back at normal speed, it was unmistakable. A tiny, involuntary tension response that repeated like clockwork at every critical moment.
I rewound. Watched again. There it was. And there. And there again.
Then I noticed my breathing. At those same moments, I held my breath. Not obviously — I was not gasping or going red in the face. But the rhythm of my speech changed. The flow of my words stuttered for a fraction of a second, not because I had forgotten what to say, but because my body was holding itself still while my hands were busy. My mouth kept moving, but the natural cadence of my voice shifted into something subtly mechanical.
I had been performing this routine for months. Nobody had mentioned these tells. Not Adam, not the friends I had shown it to, not any audience member. And I had certainly never felt them while performing. As far as my conscious mind was concerned, I was smooth, relaxed, and natural throughout.
The camera disagreed.
The Gap Between Feeling and Reality
This is one of the most humbling discoveries you can make as a performer: the version of yourself that exists in your mind while you are performing has almost nothing to do with the version that exists on video. What feels smooth often looks tense. What feels natural often looks rehearsed. What feels like a seamless moment often has a visible hitch that you have trained yourself not to notice.
Dariel Fitzkee wrote about this gap in “Showmanship for Magicians,” a book I had read earlier in my journey. He was relentless about the physical side of performance — hands, facial expression, posture, grooming. One of his key insights was about consciousness of the hands: when you become aware of your hands, that awareness almost always betrays itself to the audience. The goal is to make every movement so habitual that the conscious mind is entirely free for the performance itself, not monitoring what the body is doing.
What Fitzkee could not have anticipated, writing in the 1940s, was that video would make this self-betrayal visible in a way that no mirror, no friend watching from the audience, and certainly no amount of introspection could match. A mirror shows you what you look like right now, in this moment, when you know you are being watched — by yourself. A friend gives you their filtered impression, shaped by their relationship with you and their own ability to articulate what they saw. But a camera captures everything, in real time, without judgment, without filtering, and without the curse of your own self-awareness.
The camera captures what the audience actually sees. Not what you think they see. Not what you hope they see. What they actually see.
The Taxonomy of Tells
After that first brutal viewing session in Graz, I started cataloguing what I was finding. Not just in that one performance, but across every recording I could get my hands on. Over the following weeks and months, I built a list of the kinds of tells that video reveals, and I found they fall into roughly five categories.
The first category is physical tension. This is the shoulder thing I described — involuntary muscle tension that appears at moments when the performer is concentrating on something the audience is not supposed to notice. It shows up differently in different people. For me, it was the left shoulder. For someone else, it might be a slight clenching of the jaw, a stiffening of the fingers, or a widening of the stance. The body knows when something important is happening, even if the face is projecting calm, and that knowledge leaks out through the muscles.
The second category is gaze breaks. This is perhaps the most common tell in magic performance, and it was one I had in spades. At certain moments, my eyes would drop to my hands. Not for long — maybe half a second. But on video, it was obvious. My eyes were supposed to be on the audience, or on whatever I was directing their attention toward, and instead they flickered downward in a pattern that repeated at every critical juncture. If a spectator happened to be watching my face at that moment, they would see a performer whose eyes said “something is happening here that I need to monitor.”
The third category is rhythm disruption. This is what I noticed with my breathing and speech. The natural rhythm of a performance — the cadence of words, the pace of movement, the flow from one action to the next — changes at moments of heightened concentration. The performer speeds up, or slows down, or pauses in a way that breaks the pattern established by everything that came before. It is like a pianist who plays beautifully through an easy passage and then visibly bears down when the difficult bars arrive. Even if every note is correct, the shift in effort is audible.
The fourth category is what I think of as spatial tells. These are changes in how you use space — stepping backward when you should be stepping forward, angling your body away from the audience, bringing objects closer to your body as if to shield them. Spatial tells are about comfort and discomfort. When you are comfortable with what you are doing, your body opens up. When you are uncomfortable, your body closes down. On video, these spatial shifts are remarkably easy to spot.
The fifth category is timing anomalies. These are moments where the pace of the routine hitches — a beat too long before a reveal, a transition that takes two seconds instead of one, a moment where you seem to be waiting for something before proceeding. Timing anomalies often indicate that the performer is either executing something difficult and needs the extra time, or is recovering from something that did not go as planned. Either way, they create a subtle sense that the performance has shifted from flowing to effortful.
Why You Cannot Feel Your Own Tells
The reason these tells are invisible from the inside is straightforward: your attention during performance is occupied. You are managing multiple streams of information simultaneously — what you are saying, what the audience is doing, what your hands are doing, where your props are, what comes next in the routine. There is no bandwidth left to monitor the micro-expressions of your own body.
This is actually by design. If you were consciously monitoring your shoulder tension or gaze direction while performing, you would not be performing. You would be self-monitoring, which is an entirely different cognitive task that is incompatible with natural, engaging presentation. The whole point of rehearsal is to make the physical actions automatic so that your conscious mind can focus on the audience. But the automatic actions carry the tells with them, embedded in the motor patterns you drilled during practice.
Here is the uncomfortable truth: you probably practiced your tells into existence. When you learned a routine, you practiced it over and over until the actions became automatic. But those practice sessions happened in private, without an audience, without the social pressure of performance, and without the need to manage your own physical presentation. You were focused on the mechanics — getting the actions right, in the right order, at the right time. Your body developed whatever tension patterns, gaze habits, and rhythm changes it needed to accomplish those mechanical tasks, and those patterns got baked into the routine alongside the actions themselves.
By the time you started performing the routine for real people, the tells were already there. They were part of the package. And because they were automatic, you could not feel them any more than you can feel yourself blinking.
The Consulting Parallel
My background in strategy consulting gave me a useful framework for thinking about this problem. In consulting, we distinguish between “known unknowns” and “unknown unknowns.” Known unknowns are things you know you do not know — you can identify the gap and go looking for the information. Unknown unknowns are things you do not even know you do not know — you cannot look for information you do not know is missing.
Tells are the unknown unknowns of performance. You cannot fix what you cannot see, and you cannot see what you do not know to look for. This is why video is not just helpful — it is essential. Video converts unknown unknowns into known unknowns. It shows you the tells you did not know you had, which allows you to target them specifically in your practice.
Without video, you are working blind. You might feel like your performance is smooth and natural, but that feeling is generated by the same brain that is producing the tells. It is like asking a fish to evaluate the water it is swimming in. The fish does not know it is wet.
The Process of Elimination
Once I had identified my tells on video, fixing them became a specific, targeted project. I knew what to look for: left shoulder tension, gaze drops, breathing interruptions, rhythm changes at critical moments. I could go back to practice and work on each one individually.
The approach I used was borrowed from my practice methodology. I would run through the routine in front of a camera, focusing on one specific tell. Just the shoulder, for example. I would perform the entire routine while paying conscious attention to keeping my left shoulder relaxed. Then I would watch the footage and see if the shoulder had stayed down.
At first, fixing one tell made another worse. When I concentrated on my shoulder, my gaze drops increased. When I focused on my gaze, my rhythm disruptions got more pronounced. This was frustrating but predictable — I was taking something automatic and making it conscious again, which consumed the cognitive bandwidth that had been managing other aspects of the performance.
The solution was patience and repetition. I worked on one tell at a time until the correction became automatic, then moved to the next. Over weeks, the corrections integrated into the routine the same way the tells had originally integrated — through repetition until they became unconscious habits.
It was slow. It was tedious. It was also the single most productive period of improvement I have experienced as a performer.
What the Audience Actually Notices
Here is the question that might be forming in your mind: if these tells are so subtle that no audience member ever mentioned them, do they actually matter?
The answer is yes, but not in the way you might think. Individual tells are usually too small for a spectator to consciously identify. Nobody in the audience is going to say, “I noticed your left shoulder tensed at the 3:42 mark.” What they register instead is a general impression — a vague sense that something was slightly off, or that the performance felt a bit tense, or that they did not quite believe what they were seeing. They cannot articulate why, because the tells are below the threshold of conscious perception. But they affect the overall impression nonetheless.
Think of it like a photograph with a slightly off white balance. Nobody looking at the photo would say, “The white balance is shifted two hundred Kelvin too warm.” But they might say the photo feels slightly strange, or that something about it does not look quite right. The technical flaw creates a perceptual impression that the viewer cannot diagnose but can absolutely feel.
This is what tells do. They create a cumulative impression of tension, effort, and artificiality that erodes the audience’s sense that they are watching something natural and effortless. And since naturalness and effortlessness are essential to magic that feels real rather than merely clever, tells undermine the very foundation of what you are trying to create.
The Ongoing Practice
I still find tells. Every time I add a new piece to my repertoire, or modify an existing one, the video review process reveals new habits I did not know I had developed. It is not a problem you solve once and move on from. It is a continuous practice of self-observation and correction.
What has changed is my relationship with the process. That first viewing session in Graz was genuinely painful. I did not want to see myself doing those things. I did not want to know that my performance, which had felt so good from the inside, looked so imperfect from the outside. There was an ego cost to that discovery that took a while to process.
Now, finding a tell on video feels less like a failure and more like an opportunity. Every tell I identify is a specific, actionable thing I can fix. Every tell I fix makes the performance measurably better. The camera is not a critic — it is a diagnostic tool. And like any diagnostic tool, it is only useful if you are willing to look at the results honestly.
The performance I gave in Graz that night was good. The audience enjoyed it. The feedback was genuine. But it could have been better, and the camera showed me exactly how. That gap — between good and better, between what feels smooth and what actually is smooth — is where the real work of performance improvement lives.
You cannot close that gap by feel alone. You need the camera. And you need the courage to watch what it shows you.