Our teams at Georgia Tech and UC Santa Cruz have been working on an authoring tool that helps journalists quickly create bite-sized newsgames. The Cartoonist has been the working title for the tool because our intention is to create games akin to editorial cartoons, in terms of the amount of information being conveyed and the style of representation. But despite this small scope, the promise of this tool requires intense research and design.
Over the past half-year, we have been faced with a daunting question: How do you create something that can generate games for a seemingly endless list of topics?
Where to Begin?
We started by looking at classic arcade and Atari 2600 games and broke them down into their various components. We then asked questions about what these components mean individually and when interacting with each other. Does Pac-Man eating ghosts map to something metaphorically? Does splitting a dangerous object into two pieces in Asteroids have rhetorical implications? How do familiar game mechanics like shooting, chasing, jumping, racing, and getting power-ups parallel real-world actions?
Games are good at explaining systems and can work through processes to produce variable outcomes. A journalist might report on a story about a local business that gave a politician money in hopes of securing the passage of a beneficial ordinance. What we want is for a journalist to enter this kind of simple relationship into the tool and for it to generate a game that explains the process.
A Unique Concept
Trying to understand how the dynamics of news stories relate to the dynamics of games we found a middle ground of representation in the form of a concept map. This is a way of thinking about actors, relationships, and actions in a news event.
The story is distilled into verb relations between actor nodes while the game is distilled into mechanical relations between actor nodes. The authoring tool is able to group relations and nodes together to produce patterns of events. If one politician is receiving large donations when running for office against another politician, the tool interprets the effect of the donations on the race and makes up tasks for the player and goals to achieve in a game.
Consider the example above. Some citizens of Rio de Janeiro are buying drugs from the gangs, who terrorize the rest of Rio’s population. Citizens are demanding help from the Brazilian government, which is using the police to arrest the gangs, who are fighting back. It appears to be a complicated set of relationships that don’t obviously translate into a game.
But our tool can interpret these relationships as meaningful patterns: The fear of the citizens is self-perpetuating; the government is indirectly battling the gangs by enabling the police; the gangs have the resources to fight back. Rather than take each of the bubbles piece by piece, the tool looks for groups of relationships to turn into game dynamics.
What actually makes this happen is far more complex than this description implies. It involves picking appropriate and compatible game mechanics (things moving around the screen, colliding with each other, competing for resources, etc.). But it has been important to have a simple layer of representation that makes it easier to think about this process in our project and discuss it with the journalists who will use it when it is completed.
Our goal is for the journalist to never have to think about how the game is being built. Instead, they focus on what they do best — synthesizing current events — and leave the rest to us.