Give Me Your Gun is an interactive theater style game which on the controversial topic of gun violence in America. It was showcased at the 2016 Games for Change Festival to a crowd of almost 200 people.
Imagine a Twitch Plays interactive theater style game, and you'd have a rough idea of what Give Me Your Gun (GMYG) feels like. Crazy, with everybody typing in different things; chaotic, as some people push the boundaries of the game and try to break it; yet surprisingly engaging, as the topic of gun violence is close to every American's heart. With that in mind, this game aims to help people understand the motivations and communicate respectfully with people of opposing viewpoints.
My role in the team was that of, for lack of abetter word, a generalist. I researched the topic of gun violence in America, brainstormed game ideas, programmed most of the front end of our web interface, analyzed the design of the game, and handled some of the production work.
This game was developed over a period of 16 weeks as a semester long project at the Entertainment Technology Center. It was a great learning experience for not just our team, but our actors and faculty advisors as well, given the completely novel nature of this experience. For more information on the game development and our main discoveries in this new medium, check out our weekly development blog.
In this live game, an audience controls a conversation that is happening on stage between two actors with opposing gun control view points. By keying in questions on a custom made web application, the audience controls what one of the actors says to the other, guiding the dialogue to a (hopefully meaningful) conclusion.
To guide the audience along, whenever a key story fact is achieved, it is shown on screen. The web interface is temporarily locked and a live cut scene acted out, with hints to unlocking the next key fact. The name of the player who unlocked the key fact is also shown in order to give them a sense of agency.
On the back end, the team manually removes duplicate questions – not just those of similar wording, but those of similar meaning as well. The first relevant question which unlocks a key fact is tagged with its corresponding number and pushed to the actors (e.g. if I have a fact "I ate eggs for breakfast today", a question like "What was the last thing you ate" or "What did you have for breakfast" could both be tagged). The team also locks the audience's web interface during a live cut scene, as well as pushes top voted questions to and unlocks key facts on the projector screen.
If you're wondering why so many of the backend tasks were done manually instead of being automated, it's because these tasks require a solid understanding of the underlying meaning of a question. Writing an NLP agorithm for this within the span of 16 weeks was just not feasible, as much as we wanted to.