On the writing and telling of jokes

In my previous comedy-related posts, I’ve talked about getting started, continuing to practice, and failing to market a show. This time I’d like to talk about jokes themselves, rather than the stuff around the jokes. I feel qualified to do this based on my slightly-more-than-one-year of experience telling jokes and the zero books I’ve read on the subject.

Joke structure

The simplest jokes have two parts: a setup (which provides enough context to establish an expectation) and a punchline (which subverts the expectation in some way). My favorite example of this is the famous Henny Youngman (not, as I earlier wrote, Rodney Dangerfield) joke

Take my wife… please.

That’s a great joke. The setup, “take my wife,” establishes the expectation that he’s about to use his wife’s behavior as an example of something. The punchline subverts that expectation: instead of taking his wife for example, we’re supposed to take his wife away from him. In this the punchline doesn’t violate the stated information in the setup, it just forces us to reinterpret that information in a novel way.

Furthermore, the “take my wife” joke is very short. The setup provides exactly the amount of information we need in order to laugh at the punchline, and no more. Don’t tell us about your college background and the interview process if your joke is about the snacks they feed you at work.

There are exceptions to this. Norm Macdonald tells very long jokes. You are more than welcome to try to be Norm Macdonald. I think you will have more fun if you tell short jokes at this stage in your career.


An absurd anecdote can also be a “joke,” but I prefer to think of the anecdote as a container for several jokes, rather than as a single joke. When you’re telling a story try to find as many places as possible to put punchlines, or at least places you can reasonably expect the audience to laugh. If you go more than about ten or twenty seconds without a laugh, you’re not so much pacing yourself as you are reducing the energy in the room, which will cause your later jokes to miss. If you’re doing a five-minute set and expecting a single laugh at the end, you may wish to honestly re-evaluate a video of your performance and see whether the audience is actually enjoying it.

Where to find jokes

One of the biggest problems we have is coming up with sources for material. The most universally applicable way I’ve seen to come up with new material (which I got from a class at the San Francisco Comedy College) is to just get up on stage and complain about whatever is going wrong in your life. Usually I phrase this as “what are you mad about,” but some people insist that they’re not mad about anything. Fine. What is less than perfect in your life? Your roommate never takes out the garbage? There are too many people walking slowly on the sidewalk when you’re trying to get to work? You want a clean apartment but you don’t want to have to do the work of cleaning? All of those can be jokes!

Failing that, try to recount the last few days of your life. Write down a narrative of the events, with a pen. Did anything unexpected happen? The guy at work put the wrong soap in the dishwasher and filled the kitchen with bubbles? You got lost? The store was out of the thing you were trying to buy? Those can be jokes!

Evaluating jokes

When you’re trying to tell the best jokes, you can’t just sit in a room by yourself and write down the things you think are funny. You have to try them out on an audience, then use the audience’s reaction to help you determine what’s actually funny. I find it useful to record my sets, ideally on video, so that I can review them with an honest and unbiased record instead of relying on my own faulty memory.

The simplest version of this process is to take a few hours to review the videos of your last few months of performances, keep notes of which jokes have provoked the biggest laughs, and use those jokes for the big shows that matter.

If you continue to practice a joke without reviewing the audience’s reaction, you’re not getting better at telling a joke. You’re memorizing some words so that you can repeat them verbatim.

Furthermore, as much fun as it is to tell jokes for an engaged audience that’s laughing, it’s also important that you try jokes where audiences are a little less forgiving sometimes. The very supportive open mics and shows can give you a mistaken impression of how good a joke is. This is especially dangerous if you’re popular among the other comics at the mic: people might be laughing to support you rather than the joke. Then using that joke at a show and seeing it fall flat can be very disappointing.

On the nature of humor

People laugh when they are surprised and delighted to be surprised. My favorite example of this was a friend’s baby whose mind was blown by a light switch. I picked him up and let him turn the light on and off. He was surprised to see that the world worked that way, and delighted with his newfound knowledge. He laughed.

People laugh when they feel like they’re part of something, part of a shared experience. Because a joke subverts an expectation, the audience member who understands the joke now shares with the comic and the other audience members a new and unique perspective. They are part of an elite club who know the real truth about the topics discussed during the show. They laugh.

People laugh when they feel safe, when they feel relieved, when they have their own beliefs reaffirmed by the charismatic person with the microphone who is on the stage. This is why political humor or shock humor can work so well: if you say something that not everyone agrees with, but that the people in the audience agree with, they’ll feel again like their opinions matter and that they’re valid as human beings. They’ll laugh.


Tell jokes. Please. Stop wasting your time with words that don’t give the audience an opportunity to laugh. Have punchlines every ten seconds or so. Honestly evaluate your material and determine where you expect the audience to laugh. Review the recordings of your performances and see if you were right. If you were wrong, adjust the content of your material accordingly.