For the first couple of years of my performing life, I thought comedy in a magic show worked like this: you write funny material, you memorize it, and you deliver it on stage. If the material is good and your delivery is sharp, people laugh. If they do not laugh, you write better material.
This is not wrong, exactly. But it is incomplete in a way that I did not understand until I read John Graham’s framework for organizing comedy in a magic show. Graham divides all comedy material into three distinct categories: jokes, lines, and bits. And when I mapped my own material onto these categories, I discovered that I was operating almost entirely in one of them while ignoring the other two.
That discovery changed the texture of my performances more than any new effect or technique I had learned.
Jokes
Jokes are the most recognizable form of comedy. They are prepared, structured pieces with a clear setup and punchline. The audience knows, on some level, that a joke is happening. They can feel the architecture. “This guy walks into a bar…” is the classic form, but jokes in a magic show do not need to be that obvious. They just need to have the quality of being constructed — there is a deliberate path from premise to payoff, and the audience recognizes that someone wrote this.
Graham points out something important about jokes that I had never considered: because the audience knows a punchline is coming, they are primed for it. Their expectations are already oriented toward humor. This means the bar for the punchline is higher than you might think. If you signal “joke incoming” and the payoff is merely okay, the audience’s response will be less than okay. They expected something better. The setup created anticipation, and the punchline did not meet it.
This is why joke-jokes can be risky in a magic show. You are competing with every comedian the audience has ever seen. Every sitcom, every late-night monologue, every comedy special on their streaming service. If your prepared jokes are not at that level, the comparison works against you.
There is a way to mitigate this, and Graham suggests it elegantly: personalize the joke. Instead of “this guy walks into a bar,” make it “last week, I was at a conference in Innsbruck, and one of the organizers…” You are still telling a joke, but now it feels like a story. It feels like something that happened to you. The audience’s joke-detection system lowers its guard, because stories feel authentic in a way that setups and punchlines do not.
I use this technique constantly now. Every joke in my show is embedded in a story. Some of the stories are true. Some are constructed specifically to house the joke. But they all feel like real experiences, and that framing gives the punchline room to breathe because the audience is not bracing for it.
Lines
Lines are different. They come across as organic, spontaneous observations that arise naturally from what is happening in the moment. The audience does not detect a setup. There is no sense of a punchline approaching. A line feels like a real person reacting to a real situation, and when it is funny, the laugh has a different quality than a joke-laugh. It is more surprised, more genuine, more connected to the specific moment.
Here is the critical insight from Graham: many of the best lines are born from actual ad-libs. Something genuinely spontaneous happens during a performance. A spectator says something unexpected. The performer responds in the moment, without planning, and the response gets a laugh. That laugh is real, earned, and organic. And then the performer writes it down afterward and starts engineering opportunities for it to happen again.
This is where lines live — in the space between pure spontaneity and pure planning. The line itself may have originated as a genuine in-the-moment reaction. But once you know it works, you look for situations where you can deploy it again. And if those situations do not arise naturally, you subtly create conditions where they are likely to arise.
Graham describes this as finding lines by learning to relax on stage and letting the show breathe. When you are rigidly following a script, there is no room for the unexpected. But when you allow pauses, when you engage with the audience as individuals rather than as a collective backdrop, moments arise that produce organic humor. Some of those moments will be one-time events that you enjoy and move on from. But some will contain a kernel of repeatable comedy that you can cultivate.
I realized, reading this, that I had been so focused on preparing material in advance — writing jokes, scripting patter, memorizing transitions — that I had left no space for the show to generate its own comedy. My scripts were airtight. Every second was accounted for. There was literally no room for a line to emerge, because I never paused long enough for the audience to contribute the raw material that lines are made from.
Bits
Bits are the third category, and they were the most revelatory for me. A bit is comedic byplay within or between routines — a piece of business that generates humor not from a single punchline but from a sustained interaction or recurring gimmick.
Graham gives an example from Mac King’s act that illustrates the category perfectly. During King’s show, a volunteer signs a card, and she “accidentally” signs it on the back. This looks completely genuine. Even other magicians watching the show believe it is a real mistake the first time they see it. But it is a bit — a carefully engineered piece of business that King constructed after it happened to him once in a real performance. He saw the comedy potential, figured out how to make it happen reliably, and now it occurs in almost every show as a moment that feels entirely spontaneous.
This is the essence of a bit. It starts from reality. Something happens during a performance that is funny, engaging, or interesting. The performer recognizes the potential, then develops a way to make the moment recur. The audience experiences it as a happy accident, a unique moment that belongs to this performance. The performer knows it is repeatable, but the seams never show.
Bits can also be simpler than that. They can be running gags that develop over the course of a show. A water bottle that the performer drinks from every time a joke does not land, which becomes funnier each time it happens. A word or gesture that recurs in different contexts. A piece of byplay with a spectator that evolves over multiple interactions. The bit builds over time, each iteration adding to the cumulative comedy, until the audience is primed to laugh at even the slightest reference.
The Audit That Humbled Me
After reading Graham’s framework, I sat down and categorized every piece of comedy in my show. I went through my scripts, my notes, my performance recordings. I tagged each funny moment as a joke, a line, or a bit.
The results were stark. My show was approximately ninety percent jokes. Prepared, scripted, polished jokes embedded in stories and patter. I had exactly two lines that had emerged from real performance situations and been incorporated into my scripts. And I had zero bits. Nothing that developed over the course of the show. Nothing that built. Nothing that felt like it was born from the unique chemistry of performer and audience in a specific room on a specific night.
I was performing comedy at my audience rather than creating comedy with them. My show was a monologue dressed up as a conversation. The audience was watching a person deliver prepared material, and while the material was decent, it lacked the quality that makes live performance different from recorded performance: the feeling that what you are witnessing could only happen right now, in this room, with these people.
Building Lines Into My Show
The first thing I did was create space. I loosened my scripts. Not dramatically — I still knew every word I was going to say at every point. But I added deliberate pauses. Moments where I was supposed to stop talking and engage with the audience. Moments where I asked questions and actually waited for answers. Moments where I made eye contact with specific people and let whatever happened, happen.
Within a few performances, lines started emerging. At a corporate event in Vienna, a spectator I was working with during a mentalism piece said something unexpected, and I responded without thinking. The response got a laugh. After the show, I wrote it down. At the next performance, I looked for a moment where a spectator might say something similar, and I created the conditions by asking a question phrased in a way that made that kind of response more likely. The spectator said something in the same neighborhood, and I deployed the line. It landed again.
Over months of this process, I accumulated a catalog of lines that were organically sourced and situation-tested. Each one felt spontaneous when I used it. Each one had that quality of a real person reacting to a real moment. And the cumulative effect on the audience was significant. My shows started feeling less like presentations and more like interactions. The laughter became less polite and more genuine.
Developing My First Bit
The bit took longer, because bits require infrastructure. You cannot just drop a bit into a show. You need a setup that establishes the element, a series of moments where the element recurs, and a payoff where the running gag reaches its peak.
My first bit developed accidentally. During a show in Graz, I took a sip of water between routines, and something about the timing — the way I paused, looked at the audience, and drank — read as a reaction to a line that had not landed well. Someone in the front row laughed, and then a few more people caught on. I had not intended it as comedy, but the audience read it that way.
At the next show, I deliberately took a sip of water after a line that got a lukewarm response. Same reaction. A few knowing laughs. At the third show, I did it again, and this time someone in the audience said, “Get the water!” before I even reached for the bottle. They were anticipating it. They were in on it. The bit was alive.
Now the water bottle is a subtle running element in my shows. I do not overuse it. Graham’s warning about overdoing running gags is well taken. But two or three times per show, at carefully chosen moments, the water bottle appears, and the audience that has been paying attention rewards themselves with a laugh of recognition. It is a small thing. It takes perhaps fifteen seconds of total show time across the entire performance. But it creates a sense of shared experience that my jokes alone never could.
The Three Categories Working Together
The real power of Graham’s framework is not in any single category but in the interplay between all three. A show that is all jokes feels rehearsed. A show that is all lines feels improvised but unstructured. A show that is all bits feels like gimmick comedy with no substance.
The goal is a blend where each category does what it does best. Jokes provide the reliable backbone — prepared material that you know works and that you can count on to deliver laughs at specific moments. Lines provide spontaneity and connection — moments where the audience feels that the performer is genuinely present and reacting to what is happening in the room. Bits provide continuity and development — the sense that the show is building, evolving, creating a shared experience that grows richer as the evening progresses.
I am still adjusting the balance. My shows are probably still too joke-heavy. But the addition of lines and bits has changed the audience’s experience in ways that I can feel in the room, even if I cannot precisely measure them. The laughter is more varied. The energy is more dynamic. The audience is more engaged, because they sense that their participation matters — that what they say and do might become part of the show.
And that, I think, is the real lesson. Comedy in a magic show is not just about making people laugh. It is about making the show feel alive. Jokes are written. Lines are discovered. Bits are cultivated. And a show that contains all three feels like a living, breathing event rather than a rehearsed presentation.
You need all three categories. I am proof of what happens when you rely on only one.