The Change Machine Manifesto

A while back, I posted about a form that I was working on with a newly formed house team at the Arcade Comedy Theater.

Well it has now been 1.5 years, and Change Machine is still going strong. Recently, I boiled down all of my garbled ideas into (what I hope is) a clear description of the form that we now put up. I called this the Change Machine Manifesto. I did this because a) the team was still struggling to make sense of what I wanted at times and b) I thought it would be useful for me to try to sum it up simply.

Reproduced here for simplicity, because it is a form that nicely embodies how I like to do improv, and in case anyone is interested:

The Change Machine form is about possibilities to expand our world. 

We prefer teeing up these possibilities and discovering their outcomes together over following a developing plot/premise. 
Together, we build one coherent universe:

  • We follow an essential thread from the previous scene rather than sweep it away to start over 
  • We prefer the edit with the most possibilities to the one that points to a specific premise 
  • We prefer more players on stage to allow for more possible edits  

 We build from from an initial moment: 

  • We begin using only what information is in our scene partner's eyes 
  • We build scenic momentum by adding information with every line 
  • Above all else, we improvise playfully at all stages of the form

You CAN Edit a Scene Too Early

I was taught that you can never edit a scene too early. This may also be expressed as "by the time you think 'it's time to edit', it's already too late" or "follow your feet" / "listen to your body."

I don't necessarily disagree with any of these, but I would like to propose a different way of thinking about the problem.

Rather than think of the edit, let's think of the scene.

This way of thinking pre-supposes that after a certain point (that is usually a couple of minutes in) our scenes are going to fall flat. And before we get there, we should edit. I think that's a pretty negative way of thinking. And what does that thinking have to say about masters like TJ & Dave or Cackowski & Talarico. What would their shows look like if they were to edit at the first sign of a possible "traditional" edit opportunity?

I believe that following the "never too soon" thinking leads to short-sighted improv. The improv itself actually starts to build to a big laugh line with no expectation of getting past it, with living in those consequences. I prefer that we think of ourselves as capable of improvising each scene for the length of the show if necessary. We focus on repeatably creating meaty scenes and editing not to "save" the scene from potential future ruin but just because it is time for the story to move on (or whatever however your form prefers its edits).

This model trust scenes and players to survive past that first big laugh. And in doing so trains them to deal with the consequences of their moves.

When I first present this idea to groups, we often play 5 or 10 minute scenes, then go back and talk about the first moment that their Harold edit instincts kick in. It is often within 30 seconds to 2 minutes. Yet more often than not, the scene work that plays 3, 4, 5 minutes out is beautiful and unexpected territory, and we are thankful to have seen it. When we go in strapping in for a long ride, a number of things happen:

  • We more often push past the part of the scene that we've "figured out" in our head and begin stumbling again–listening to our scene partner and playing on the edge of failure
  • Small specific choices early on and even mistakes get magnified into show-affecting themes
  • Solid meaty scenes tend to get solider and meatier. Fledgling scenes tend to get more fledgling*
  • We heighten past the point we used to think was the top and again find amazing subconscious places to go (thanks to Chris Trew for first introducing this idea to me).

* Another key point: this is a place to retain your Harold edit instinct. When a scene did not start well, I would go back to: you can't edit it too early. This is less a note about editing and more a note to improve our ability as a team to make sure we get into that easy, less work, scene's building momentum place quickly. This will of course never be 100%.

YETI - A (fairly) repeatable process to get to meaty scenes quickly

I posted about what improv means to me lately, but how do you do it??

There's no easy answer. I have been teaching a workshop for a while now about how to start scenes (or really shows) in a way that I feel most reliably leads to this sort of work and, in my experience, has a big rate of success leading to rich, meaty scenes. This was developed out of how Change Machine has been starting scenes for 1+ years, and how I would later get my guests to start on Gray Matter. This is how I have the most fun on stage and feel like I am not doing a ton of work. This is not true of everyone, and if your fun is elsewhere I take no offense if this is not a good system for you. A few people have asked me for notes or a description of YETI, so I thought I would put it up here.

So here is an overview of YETI or Everything You Need is In Your Partner's Eyes.

The acronym walks through the first moments of a scene:

  • Yourself
  • Eyes
  • Tension
  • [Discover]
  • Information

Let's walk through each of those in turn. Note this process begins with the lights down, before the show properly begins:


To begin, both (or any number of) players make a small, physical, human choice. Some players when they hear this contort themselves into downward facing dog. The goal here is not to "transmit" information to your scene partner (e.g. I am playing baseball) but just to stumble into some choice that you, yourself made. Human here means that your shape should be the shape we find a person when no one is looking. 


Make eye contact with your scene partner. This cannot be overstated. Until the final step (Information), this eye contact should not be broken. More often than not, when a player breaks eye contact I can see in that moment that they feel like they've figured out what the scene is. They've got the plot down. So they disconnect. The eyes go out to the crowd or up. But I promise you the flow of information to and from your scene partner is still at paramount importance. It is very difficult to be planning ahead when staring into your partner's eyes. 

When the eye contact is established, the lights come up!


Things just keep getting harder don't they? Now I push you to take all of those ideas that are storming your mind and let them flow back out. Your instinct is to think, "she is sitting and looks angry, I bet we're kids and I broke her toys." But I will push you to react physically. If she looks angry, perhaps you back slowly away. Perhaps you close your posture. Your unbroken eye contact will ensure she reacts immediately, perhaps easing closer and closer attempting to open you back up. Throughout all of this we are delaying as far as possible labeling anything. We are just reacting to what is happening right in front of us as dramatic tension builds.


I put discover in quotes because a) it does not fit in the acronym and b) it really is not very distinct from the following step. But I will separate it here for clarity. 

At some point, the tension gets so high that it boils over and you discover the first line of dialog when one of you cannot help but speak. You will feel the difference between this and your typical improv initiation. Compared to

"Can I get a suggestion?"
"Flask, thank you"
(internally) "Hmm, flask makes me think of the town drunk, so I'll say"
"Hey, you better not be drinking whiskey in the street, Hank!"

Ugh, all that thinking all that work. In this model we take all of this natural behavior, our built-in reactions to body language and facial expressions and as it happens we begin to build a picture in our head. In the previous example let's say you chose to back away slowly and your scene partner inched ever closer, her angry look turning more and more psychotic. Your arms now open wide as you back up faster and faster until you hit the side wall. You feel protective of something certainly. Just as she is inches from your face you yell, "You can't have her!"

No work, no thinking. Just makes sense. And already by the first line of dialog there is so much happening.


In a lot of improv, someone comes out with an idea, and we give them space to get that idea out. In this worldview, no one has an idea. We do not sit idly back and assume someone else will do the work. From the first line on we add information in every line. This is harder than it sounds. This is beyond yes and. This is not resting, not repeating what we know, but building the momentum of the scene at breakneck speed (also tends to lead to action):

A: "You can't have her!"
B: "You are not powerful enough to stop me!"
A: "I am willing to sacrifice more" (mimes a blade, at her own neck)
B: (backing now away) "You can't. You won't. Neither of us would survive."
A: "I never wanted it to end this way. But you forced this on us, Gwyneth! You did this to us!"

I just made that up as I wrote it. Not quite the same as being on stage but you get the idea. See how every single line, new information is added. The second line, it may be our instinct for B to say "Yes I can!" or "Give her to me now!" but that doesn't really tell our partner or the audience any new information. We know B wants her and A doesn't what B to have her.

This also happens to be an incredibly high energy scene. In TJ and Dave terms (as I understand them), it has high heat and heavy weight. It does not have to be that way. That was just the scene I made up writing this post.

Here's another example:

Change Machine at the Pittsburgh Comedy Festival on 8/23/2014

I really love this stuff. The more people ask good and interesting questions, the more I refine my own understanding of it. If you're interested in learning more, let me know!

ACTION, or If Romeo and Juliet had Been a Fairly Typical Improv Show...

This is my prediction of what Romeo and Juliet would be like if it was done by many improv teams (abridged):

  • Romeo and Juliet fall in love
  • The Prince says Romeo can't hang around. 
  • Romeo is all like, "but I love Juliet"
  • The Prince is like, "Oh man that sucks."
  • Juliet maybe is like "Guys, if you don't figure this out, I am gonna...kill myself!" 
  • Romeo steps in like, "Yo J, don't kill yourself I love you too much."
  • Juliet is like: "I don't know maybe I should. But Romeo does love me..."

And on and on. Not as compelling as the original is it? What's the problem? We never get to the point where DRAMA transpires and people are able to emote about it.

I think a lot of improvisers are afraid of action. They are afraid to make bold moves, to offset the course of a scene or show in progress. They are afraid to drink that poison and leave their scene partners to deal with the consequences. 

But we go to the theater and to movies to watch people deal with the consequences of their dead lovers! Not to watch them debate whether or not to drink the poison (notice I didn't pick Hamlet).

The more I coach teams to move toward action, the more I see truly incredible and surprising moves. Often their teammates are surprised but once we move past the initial surprise we get into such rich and deep scenes where we deal with the fallout  rather than swimming in the muck of indecision.


  • I am not promoting unmotivated "invented" action. If we're at your family reunion and you stand up and start throwing baseballs at your Aunt June, that doesn't make any sense and confuses everyone. If we're at your family reunion and the scene has been about how your childhood was taken away by your Aunt June, who made you work in her store since you were 4 and you never got to do anything normal kids do. Then your biological father shows up with a bag of baseballs in tow to make up for lost time, and you stand up and start throwing them at your Aunt June, bravo!
  • This sounds like the old improv adage: Show Don't Tell. I support that thinking, and they are related. But to me, this is a shade different. Show Don't Tell means rather than talking about going to the store, just go to the store. Going to the store may not be that big a move though. It may just be something we were talking about. I am particularly advocating that you make bold choices that we are afraid of specifically because they could potentially change (or deepen) the course of the piece. 

After 12 years, I am learning what improvisation means to me.

I used to say that every 2 years, I learn that everything I thought I knew about theater was completely wrong. 

My pace of improvisation epiphanies has been similar, but recently they are at least trending in the same direction. At the core of this is a solidifying understanding of what sort of work–to use Susan Messing's nomenclature–gets me off. I'm going to attempt to put into words what improv I strive for these days.

Improvisational theater is an art form that is capable of doing something that no scripted performance art can: it allows 2 (or more) performers to enter a blank stage with blank minds. Neither they, nor the audience, have any idea what the scene or show that is about to be performed will be about yet the lights come up and immediately something is happening. Something engaging. Those actors rather than needing to invent a story in the moment can just play the moment. They use an almost monastic level of control to allow their thoughts and ideas of what the scene is and where it is going to come into their head and flow out. And moment to moment, the scene unfolds before them. This allows for unique and beautiful creations that would not be possible in sketch or plays. 

In a YETI workshop (more on this soon) I taught recently at The New Movement in New Orleans, a player began with a strong posture choice–slumped shoulders, legs wide, very masculine. Her scene partner crossed the stage cautiously. I asked her if she felt her partner came too close or stopped too far away and she said, "I felt my character was on a bus, sort of a creepy unsavory dude." This is the default for a lot of improvisers: what is the narrative? The plot of the story. I asked them to replay the scene and focus on just what information was in front of her. Make that great choice but try to hold off on figuring out where she was and who she was and just listen to her scene partner. How was she holding herself? How was she regarding her? To use TJ and Dave's terminology1, what was the "heat" between them?

With little training, this is something we can all instinctually learn to pick up and react to. And with some training, we can become quite good at it.

And in improvisation, the process is the product. The audience is along for the ride. They watch this with us. They are on the edge of their seats. They discover as we discover and get surprised as we, often genuinely, get surprised.

So, what are the implications of this way of thinking? I will list a few of them here, though I am definitely still exploring this:

  • If approaching improv this way, comparing your work to its written counterparts 2 (sketch, plays, etc.) is a misguided endeavour. We should study those counterparts in our effort to become better improvisors, though as critics of our work and others', I do not believe we should hold up written work as the standard. It is perhaps more akin to other improvised collaborative art such as improvised jazz.
  • This way of viewing improv does not discount the other ways as invalid. I do believe there are some truly funny people (one referenced in this article) who are very talented at more premise-based approaches. Perhaps there should be different names for the two ways of improvising. We currently distinguish between short form and long form improv, and I think maybe this distinction is more important. Maybe call that "Writing on the Spot" or call this "Collaborative Scenework."
  • With great power comes great responsibility. With written work, if I want to explore or wander into dangerous/taboo/offensive territory, I can run it by a friend who has more experience than I do with mysogony, abuse, addiction, racism, etc. to ensure that I am handling the issue appropriately and being funny for the right reasons. Even if I am improvising in a premise mode or planning ahead in my mind, I can run a quick broadcast delay. The approach I am describing here relies on the performer to have that responsibility in themselves.


  1. In their new and incredible book, Improvisation at the Speed of Life: The TJ and Dave Book, TJ and Dave define the "heat" and "weight" of a scene. This is an incredible way of looking at opening moments, and the whole book is great too!
  2. Matt Besser in an interview with SplitSider: "We always comp are our scenes to sketch comedy because we believe a great improvised scene is something that you can write out and it’s a great sketch."

The End of the World Show: Press + Performances

In the last post in this series, I left off at our tech rehearsal which would have been a day or two before opening night. Let's wind back the clock again...


About 2 months prior to the show, I sent out my first press release. I am about the furthest from a marketing expert, but I was passionate about this show and I had help from the marketing crew at the venue. The first release was focused on the show itself: what was unique and why would an average person be interested in seeing it. I also included some hi-res press photographs.

A month later (still ~1 month out), I sent a follow-up release that was more focused on the (now confirmed) cast, performance dates, ticket prices and info. This had updated photos as well. 

Between these and some contacts I had, particularly in some smaller online publications I got a couple interviews, a ticket giveaway, and a reviewer coming to our first show. The more people began to cover it, I got the show picked up by a few larger publications. We ended up selling out two of the four shows and getting a great house at the other two.


The performances is where the director is really able to let go of the reigns and trust that the preparation was done now. The show now belongs to your cast and crew. I do give a few notes during the run, but mostly on technical issues that help the show to run more smoothly ("Let's try to have the overlords do the house open announcement since they're in make-up at that time"). It's difficult to make major artistic adjustments during the run. If you do, probably better to call a rehearsal and practice than rely on notes.

Those of you who have done short runs of theater productions will be familiar with the bittersweet ending to such a production. It has taken over your life, and you will be getting a lot of free time. But you have also built something special and unique and made a new family and that is going away. But like all of theater and improv specifically, its ephemeral nature is part of its appeal. 

The End of the World Show: Tech

Though it has been more than a year since my last post about this process, I wanted to pick this back up as I have been encouraging more and more people lately to think about projects.

In parallel to rehearsing with my cast, I began work on the tech aspects of the show. Ideally, I would have had a Technical Director to collaborate with on this, but for this show, that was also me. I did, thankfully, have a wonderful board operator at the theater.

Tech Break-Down

For End of the World, tech consisted of:

Final Set Design - Musician would set up stage right, cast on stage and Overlords off stage left with low colored lighting. Projection gave the appearance of looking out of the ship and onto Earth.

Tessa Karol as an alien overlord in costume, Hair and Makeup at Tech Rehearsal.

  • Pre-Recorded Audio and Video: This show had a lot of audio/video components. For me this involved sourcing and compiling this content, and in some cases working with artists to create original material. 
  • Live Music: Most of the shows had a live synth player adding environmental ambience and underscoring the scenes. We had to find the right place to put her physically in the space, hook her into the board, sound-check and incorporate her into the rehearsal process (for this show, she was just at the tech run through).
  • Set Design: We did not build any sets, but that did not mean that there was no set design. We had two distinct areas of the stage we wanted to set apart, as well as the live music. We played with where to place them physically, lighting options, and incorporating some furniture.
  • Costumes, Hair and Makeup: To clearly distinguish the aliens from the humans in the show, we used costumes, hair and make-up. We had a very low budget, so the costumes were gathered from my and the casts' closets, hair services donated by a cast member, and make up done by a generous local face paint artist in exchange for hanging her art work in the lobby. 
  • Pre- and Post-Show Experience: I wanted the experience to begin from the moment the audience entered the lobby. The conceit was that they themselves were an alien counsel and by the end of the show would vote to save or destroy the human race. Before the house opened, they spent their time transforming themselves into their alien alter-egos using tin foil, pipe cleaners, egg cartons, face paint and costume pieces. The end of the show allowed them to vote to save or destroy using their smartphone and watch the vote live on screen, sharing the results on social media. All of these pieces had to be built and organized.

Creating Media

The concept of this this show seemed essential to explain for the audience to enjoy, but also complicated. The solution I came up with was to present a short video at the top that fit with the pre-show experience. I worked with two incredible animators to create it in an amazingly short amount of time: Spencer Diaz–an Imagineer at Disney Research–and Nico Zevallos–a fine art and robotics student at Carnegie Mellon. The resulting work did a wonderful job of leading into the short scripted introduction:

Introduction animation used in the End of the World Show

They also created two other animations that we played depending on the audience vote at the end of the show.


QLab project for End of the World Show

The show ran about 90 mins and had about 20 cues in it. That's probably low to normal for a sketch show but high for an improv show. I wanted to make this as easy as possible for the person in the booth (let's call her the Stage Manager), so I programmed all the cues (audio and video) into an amazing and intuitive piece of software called QLab which you can rent for $3 per day. I rented it for the week leading up to the show and the run of the show and it was remarkably affordable for what it does.

With QLab, you can program all of your cues to operate just by hitting the space bar. Cues can play video, put up static images, play audio, do all of those at the same time (which I did frequently), fade sound in and out, time cues to play during other cues and so forth. And it's all pretty easy to set up. Then your tech person will just need to hit the space bar at each cue.

Technical Runs

With all that preparation done, you are ready for your tech run. Ideally you will do (at least) 3 kinds of runs at least once. If time or venue availability is limited, you may need to combine some of these but I don't recommend it:

  • Paper Tech: You want to walk through each of the cues with the Stage Manager on "paper." This will be whatever you have for a script. If it is a fully improvised show, this may be more of an outline. But if it is a more ambitious production with technical aspects, you definitely do want to write down clear instructions before the first rehearsal.
  • Cue-to-Cue: Here is where you involve the actors and any other production folks. You want to be in the venue and use all the actual people who will be doing anything on show night. The goal is to walk through each cue, run a little bit before, run the cue, a little bit after then skip to the next cue. I usually use these to also to rehearse entrances, exits, any devices (edits for example) that we may not have rehearsed in the space, where do we stand when we're not in a scene, etc.?
  • Dress: This is a full run of the show in costume, make-up, full tech, top to bottom. This should be treated like a performance. It will also give you a sense of timing.

At this point, you should be confident about opening night!

This post is part of a series about creating the End of the World Show

Austin Days 4&5: Environment, Heightening and My New Best Friends

Each day has a focus but none of these are as simple as "let's do some space work" or "don't forget to be somewhere." Chris Trew taught in the morning on Day 4 and he made a point that really resonated with me. A question came up about why some argument scenes or transaction scenes are really successful, and Chris said basically to throw all of those "No transaction/teaching/stranger/therapy scenes" rules out the window. 

We learn a lot of pillars (in TNM lexicon these are weapons and paths but pillars is a simple metaphor). You have a game pillar, a character pillar, an environment pillar. As long as a few of those pillars are supporting your scene (probably two, or one really great one, if you have three BAM), you'll be cooking. So a transaction scene between strangers with strong Points of View and clear game, probably going to be great! Similar with environment. How can we use it as a tool to give us another place to push off from or another source of discovery/contrast?

One of my biggest discoveries of the week was on heightening: STOP DOING IT IN MY HEAD. I learned to stumble into heightening the way I discover at the top of the scene. Go out there, not knowing what the next thing is I'm going to offer, but trust that I get patterns and I get heightening and whatever comes out will be great. This was a revolution for me and led to some of the most joyous moments (mostly me watching them because I struggled with this). But I had a scene in the show on Thursday (often referred to as the blowjob scene) where I did this accidentally. I now hold this up as my ideal of how I want to feel when I'm heightening: on stage, not the back line.

TNM Training Camp 2014

And maybe I should have led with this, but I truly feel like I have 9 new best friends (plus CJ, Amy, Chris and Tami). It's like the closeness I have felt on house teams, playing with the same people for months, doing good shows and bad, great scenes and crappy scenes, I feel all of that with every one of my training camp compatriots that I met 5 days ago. If you are looking to breakthrough your work, and work hard, I can't recommend this enough.

Austin Days 2&3: Emotion and POV

Each day has a focus. Day 1 was "and like a samurai," focusing on the consistently building off your partner every line. I forget the phrases for 2 and 3 but 2 was drilling into emotion, reacting with emotion over logic and diving in hard (I struggled most here so far) and 3 was creating a strong point of view for your character in the first moment of the scene, contrasting/complementary points of view and so forth. 

Since we've had some shows under our belts, we've also been watching tape. Each day after lunch we watch the video last night's performance altogether and CJ (one of our instructors). We watch the whole thing together start to finish and he pauses throughout to make quick comments, ask questions, and issue challenges for the afternoon's work.

I also haven't quite gotten my finger on what's different about notes here. Every teacher did a set or two with us, and gave us notes scene-by-scene. But they never seemed to make people feel worse, or choice-coachy. Sometimes they were quick, sometimes more of a discussion (initiated clearly by a question from the teacher), sometimes they would ask the class to chime in to "popcorn" ideas about other moves in a given scenario.

Speaking of popcorn, these classes, like the TNM book (Improv Wins) are full of catchy phrases and useful metaphors. Teaching this way helps boil things down and communicate complex ideas more clearly, and helps us remember them before shows or the next day. Chris says these come to him usually in the moment then get refined over time, but they are useful. 

After each session, we all summarize our take-aways with a word or phrase. Here are some I have found useful so far:

  • And like a samurai
  • Discover something in every line
  • Stumble into emotion
  • Always build momentum
  • Add depth not details
  • Let the walls down. If the walls are up, explain the walls.
  • Find something you can take to 7
  • Joy Anger Love Seduction Fear
  • Contrast over Conflict
  • Find a signature move

The Megaphone Marathons are also starting to bring out some great teams. Below is the amazing Austin team, Opposites, featuring our Day 3 instructor Mark.

Austin day 1: The intensive begins

The New Movement

TNM Intensive runs from 9am-4:30pm Mon-Fri. We go out for lunch and have an optional dinner time activity before coming back for a performance each night. The performances are part of the Megaphone Marathons, with some groups from out of town.

Some day 1 observations:

  • It's a long day, mentally and physically. So I may not be as eagerly blogging as I had hoped.
  • There's a journaling component, which is great. I always bring a notebook to classes and ep0/usually to rehearsals but this longer format allows short spaces in session to journal.
  • A lot of the individual feedback comes from the teaching structure. There is an official lesson, with a rotating cast of teachers each session. But there is always another teaching sitting in the house, watching your work and jotting down lots of notes and data to provide you feedback on your goals for the camp.
  • Like tech conferences, much of the best experiences already happen outside of the formal programming. Walking back from dinner I was able to unstick some of what was confusing me about the philosophy presented. I enjoy just talking to improvisers from other cities about their experiences and scenes where they're from. This week will take 10 strangers and turn them into a team, and that's pretty great.
  • Scene sprints - using a complicated game clock, every so often we do 10 sprints of :27 second scenes with :03 seconds in between (in pairs throughout the room). This is an energizing, if exhausting way to get a lot of practice with a new idea.

See you tomorrow!