Chapter 1 — The ground
What improv is, and what it is not
Improvised theatre is composed in front of the people watching it and does not survive the night. Everything else about the form follows from that one fact.
Here is the shortest honest definition: improv is theatre that is invented while it is being performed. Nobody wrote the dialogue. Nobody knows how it ends, including the people saying it. When the lights go down, the thing that existed is gone, and it cannot be performed again because there is nothing to perform from.
That sounds like a party trick. It is closer to a discipline, and the discipline exists because of a paradox at the centre of the form: total freedom produces nothing worth watching. A stage where anyone can say anything at any time yields chaos, and chaos is boring within about forty seconds. What makes improv work is a dense set of shared constraints — agreements about how to listen, what to accept, when to leave — held so tightly that the performers can afford to have no plan.
What is actually unscripted
Beginners usually assume the whole evening is conjured from nothing. It rarely is, and pretending otherwise does the form no favours.
What is genuinely invented in the moment: every line of dialogue, every character, every relationship, every location, and the entire plot. That is not a small list. It is essentially all of the content.
What is usually decided beforehand: the shape. Most groups agree in advance on a format — how many scenes there will be, whether they connect, how the piece opens, roughly how long it runs. A Harold has a known architecture. A Montage does not, but "we are doing a Montage" is itself a decision made before anyone walks on. Rehearsal is real, and it goes into the container rather than the contents.
The useful analogy is jazz, and it is useful for exactly one reason. A jazz quartet has not written tonight's solo, but everybody in the room knows the tune, the changes and the form. They are improvising inside an agreed structure. Improvisers do the same thing, and the reason a scene collapses is almost never a shortage of imagination — it is two people playing in different structures.
Short form and long form
Short form is a sequence of separate games, each governed by a rule that generates comedy: play the scene backwards, replace every line with a question, perform it in the style of a horror film. Each game runs two to five minutes, an audience suggestion sets it off, and then it ends and the next one starts. This is the improv most people have seen, because it is the improv that televises.
Long form throws out the games. One suggestion, taken once at the top, produces a single connected piece running anywhere from twenty-five minutes to a full evening. Scenes recur. Characters come back. A thing established in minute four pays off in minute thirty-one because someone remembered it. The comedy comes from pattern and recognition rather than from a rule imposed on the scene.
Neither is the advanced version of the other, whatever anyone tells you. Short form is genuinely harder to do well than it looks — the rule is doing work, and if the performers lean on the rule instead of on each other, the whole thing curdles into a game show. But long form is where most of the last forty years of craft development has happened, and it is what most of this course is about.
What it is not
It is not sketch comedy. Sketch is written, rewritten, cut, and rehearsed until it is tight. The confusion is understandable — a great many sketch writers trained as improvisers, and several famous sketch institutions run improv theatres — but a sketch exists on paper before anyone performs it, and a scene does not.
It is not stand-up. Stand-up is one person, an authored point of view, and material refined over months of repetition. Improv is collaborative by definition; the entire skill is being changed by what the other person just did. A stand-up who bombs can rework the bit. An improviser cannot, because the bit is already gone.
It is not "just being funny." This is the misconception that damages beginners most. The people who arrive determined to be funny generally aren't, because being funny on purpose requires you to stop listening long enough to think of something funny — and the thing you stopped listening to was the scene. The reliable route to a laugh runs through paying close attention to another person, which is an unglamorous thing to be told in week one.
What the audience is actually watching
Not the jokes. Or not only the jokes.
An audience at a good improv show is watching people make decisions together without a net, in real time, and get it right. There is a specific pleasure in seeing someone offered something impossible and take it anyway — a tightrope pleasure, closer to watching a high-wire act than to watching a comedy. Part of what you are enjoying is that it might not work.
This is why recordings of improv are so often disappointing. The tension that made the room electric does not survive the transfer, because on video the outcome is already fixed. You were not laughing at the line. You were laughing at the line arriving.
The best improv looks, from the outside, like the performers must have discussed it beforehand. That impression is the whole craft. They didn't, and getting a room to believe they did takes years.
Where to go next
If you take one idea from this page, take the paradox: constraint is what makes the freedom usable. The next chapter is about the first and most misquoted constraint in the form — the agreement principle usually flattened into two words on a motivational poster.