I recently attended a preview of a new online work by a well-regarded Boston theatre company. The performance, which was marketed as an “Experimental Game,” took place over Zoom, with over 100 audience members in virtual attendance. I was looking forward to the show because the premise seemed compelling, although a bit of an “in joke” for theatre people: It framed the work of a single playwright as an “operating system” in which characters are stuck in a seemingly endless loop of disappointment and failure. The interaction pretext was that the audience could end the cycle and put the characters on a different path. The mechanism through which we were asked to do this was voting, using a smartphone-based web interface. The play took place entirely through Zoom, with the audience visible during the show. Notably, they chose to leave the Zoom chat on, which is very rare in online theatre. And this is where things went horribly and beautifully off the rails. I say beautifully because as a connoisseur of emergent behavior, this was one of the best instances of undesirable emergent behavior I’ve seen in a live theatre event to-date.
The action took place on one main screen. A live actor in a virtual set served as the MC. There was also a disembodied AI voice, most likely a live actor. At the beginning of the show, we were asked to vote on which play characters we wanted to “help.” The votes were submitted, and the play was chosen. I have to grant the company that all the theatry parts were excellent. The staging and virtual sets (created by a game designer and shot in a green screen studio) were well-designed and executed. Having experimented with this format, I was impressed with both the design and production value. The acting, most of which was pre-recorded, was superb, particularly given how challenging it is for stage actors to adapt to mediated performance.
I want to reemphasize that this entire experience was framed as a game, with the MC, who was very active, regularly calling out individual audience members by name, and referring to us collectively as “gamers.” As the story played out, we were given several more opportunities to vote. But as the show progressed, the audience (which, by design, included a combination of both game and theatre folk) became skeptical as to whether or not their votes were actually being counted. A lively text convo ensued during the show, as audience members began to compare their votes in chat to determine if they were actually having any impact at all. As the show drew to a close, we concluded collectively, that they were not, and an audience rebellion ensued. At the end of the day, although what we had seen was a well-produced online play, the general consensus was that this was not by any definition of the term, even a very broad one like mine, a game. Some came to the conclusion that the faux interactivity was making a statement about the playwright’s world view, but others just felt they’d been conned. The emergent behavior of this chat was the most interactive and interesting thing about the show.
After the show there was a “talkback” where I learned the important lesson that audience members are not supposed to “talk back” in a “talkback.” When the moderator explicitly asked us what we thought about the audience agency, I spoke up (after having praised other aspects of the production) and pointed out that whether or not the agency was real, if your audience thinks it’s not, then you’ve got a problem. The artistic director of the company became defensive and patronizing, asking me what kind of agency I had during the pandemic. I made the additional mistake of trying to answer what turned out to be a rhetorical question (I actually got a lot done during the pandemic including starting a nonprofit, finishing one book and starting another, co-producing a conference and a virtual larp festival) because he didn’t pause long enough for me to answer. Afterward, I invited anyone who wanted to talk further to gather in the Playable Theatre Discord, which functioned as a kind of online aftershow bar, to debrief and blow off steam.
This experience led me to think a little more deeply about using voting in live performance, whether online or in person. Should voting be used at all, or is it too superficial a method for meaningful interaction? This was particularly vital to me since I’ve been working on a show in which voting is the core mechanic. So I began to ponder some of the other shows I’ve attended that used voting. In so doing, I realized that voting, like any interaction modality, is really a question of craft and quality. The gesture of voting in itself is nominal and low-overhead for audience members. It doesn’t require them to work very hard, and for that reason, it’s ideal for conventional theatre audiences that may prefer a lightweight form of interaction. However—and this is key—voting can be lightweight without being meaningless. And as with any artistic technique, it is simply a matter of storytelling and craft. To wit, I’m going to discuss a few experiences I’ve had in which voting was more or less effective and why. Based on these examples, I’m going to conclude with some recommended rules of thumb for using voting in a meaningful and effective way in live theatre.
The first example I want to use is the other online piece by the same company who did the show described above, which had a similar problem. The show was explicitly promoted as putting the audience in the role of jurors. I’ve been on a jury before, and have always thought the jury format would be a good premise for interaction so I was interested to see how they handled it. I should add that this show was exceptionally well-reviewed and I actually heard about it on NPR. The show consisted of a monologue (again, well executed), followed by a vote. There were no lawyers, no testimony from witnesses or experts, nothing that suggested a courtroom. The vote took place at the very end of the play using Zoom’s onboard voting system, and we were asked to decide if the accused was guilty or not. There was no discussion amongst audience members. Just a single vote, after which the result was displayed, and the accused was deemed guilty or not guilty. And that was the end. The fact that I have no recollection of the final decision is telling. I was disappointed on a variety of levels, but the thing that irked me the most leads me to my first principle. Set audience expectations appropriately. If you’re putting them into a role that implies agency, put them into that role and give them the promised agency. If you can’t do that, then just put on a good play and leave it at that.
An example of how to do this right is Dacha Theatre’s :robot_face:, which I reviewed in a previous blogpost. Performed live over Twitch via Zoom (with a few intermittent prerecorded mock advertisements), the experience was framed as a “game show” with two audience modes. One was the contestant mode, a premium ticket which had players on the screen for the duration of the show, programming actor-played AI’s in teams with the help of an actor/programmer; the second was the spectator mode, which was facilitated through Twitch chat. After each “programming round,” the AIs were interviewed by members of a panel of judges. The judges discussed their impressions of the AI’s, discussing which they thought was most convincing, but the final determination was made through voting by the Twitch audience. Importantly, voting was visible to the audience so everyone could see in real time how others were voting, making it a highly communal activity. The game show format set expectations for audience interaction, then administered that interaction appropriately. My metric for the effectiveness of the voting mechanic is based in part on the fact I played in both modes. As a contestant, the win state is determined by the audience, so I had a high level of investment: Since the entire outcome of the game show was determined by the audience vote, it had very high stakes. Furthermore, when I played as a Twitch spectator, my partner got extremely worked up about which AI he wanted to win. To me that was also an agency win. He actually cared about the vote. And here’s the rub: We weren’t even playing the interactive version, so we couldn’t even vote ourselves. In both cases, I remember the outcome of the vote, also an agency win.
So that leads me to my next example. A vote can be meaningful if the audience cares about the outcome. This means either a vote has stakes, as in the example above, or the vote has a significant impact on how the narrative plays out. One of my favorite examples of this is a British piece called Choose Your Own Documentary. Similar to the first example, it’s an MC’d show with pre-recorded film, although performed live in a theatre, with remote-controlled audience voting. The story revolves around the narrator’s quest to find the author of some diary pages that fell out of an Old Choose your own Adventure novel. The live host, Nathan Penlington, is also the narrator of the film itself, and the voting choices revolved around the decisions he and the directors make while shooting the documentary. At various points the audience was asked to decide: Should we pursue this lead or follow a different path? The key thing here has to do with the content: The choices we were being asked to make had a consequential impact not only on what we were going to see next, but on the ethics of the filmmaking process itself. The audience was very vocal about promoting their choices, and there was one point in the show where I got particularly worked up because the choice we were being asked to make had to do with whether we were going to press one of the subjects to unwillingly do an interview. As an ethnographer, I felt this was unethical and adamantly tried to get the rest of the audience to vote my way, which they did not. So for me personally this vote was highly consequential, but more broadly, it visibly changed the path of the narrative. (I should also add that due to a technical glitch in the middle of the show, we discovered that the entire project was implemented in PowerPoint, which is an astonishing technical feat in and of itself.) This leads to another rule of thumb, also one which is practiced in :robot_face: — make voting outcomes visible to the audience. This is really just a basic tenet of both game and interface design—user feedback is key so that people understand the results of their actions.
Green Door Labs’ Club Drosselmeyer uses a variety of interaction modes including multiple voting mechanics, all of which include the techniques above. These also employ additional techniques which I find to be highly effective. One is to create factions. One year there was a drive to purchase “war bonds” on behalf of different groups that audience members were assigned to. The group that raised the most money in war bonds was lauded at the end of the fund drive. Because it’s framed in the fiction of the game as helping the war effort, people are motivated to purchase bonds and have a stake in the outcome. This also uses the technique of making the activity of voting itself interesting. Another example is a voting mechanism that dictated the outcome of a swing dance “bot battle” at the end of the show (see banner image). Throughout the night, players collected “features” they wanted to add to each robot in the form of punch cards, and dropped them into each bot’s respective box. Besides being fun, this voting mechanism also gives players a vested interest in the outcome. At the end of the show, the collected features were combined to give the robots the capabilities to win the battle. In addition to making the act of voting interesting, the voting thus results in a theatrical spectacle.
It’s interesting to note that in all of the examples given, including the ones that fail, voting is a part of the diegetics of the fictional world, meaning it’s integrated into the experience in a way that suits the narrative. So I would say that, though helpful, a diegetic conceit is not a guarantee for making voting meaningful, and should be entirely avoided if it’s only a superficial gimmick.
So, to return to our leading question—To Vote or Not to Vote?—I would say, voting can be extremely effective in live theatre with at minimum the following four requirements:
- Set expectations and deliver appropriately
- Create high stakes outcome and
- Give the audience a reason to care about it
- Provide clear feedback as to the results of voting
Also highly recommended, , any and all of the following, but only after the criteria above are met:
- Make voting part of the diegetics of the world
- Create factions
- Make the activity of voting itself fun or interesting
- Have voting result in a theatrical spectacle
I’m sure there are many other rules of thumb pertaining to voting, but I feel like this is a good start. I do think voting is a promising mechanism for audience interaction, and it can be extremely satisfying if designed and executed well. Hopefully the tips above will prove useful for people creating live participatory performance, but when in doubt, if you haven’t already done so, I highly recommend adding a game designer to the team.