Students' quest: find a way to study video game characters

Video gamesEveryone knows that the characters shown in video games don’t accurately reflect the diversity of the people playing them. When you take the controls of your favorite video game, your first — and often, only — option typically is to play as a white male character.

But how exactly does everyone know that? And if it changes, how do we track the progress? While there is plenty of anecdotal evidence that women and minorities are underrepresented, there is precious little hard data about the characters in games.

“A lot of people play games, but there actually hasn’t been much work studying representation in games at all,” said Brian Waismeyer, a student in the Information School’s Master of Science in Information Management program. “What intrigued us was, why not?”

Waismeyer and three fellow MSIM students are addressing that question with their Capstone project. The culmination of iSchool students’ degree program, Capstone connects students with businesses and organizations to solve real-world information problems. The students will present their projects at the annual Capstone event on May 26.

Working with the iSchool’s GAMER group, the students identified a pair of information problems that are barriers to studying video games. For one, there is no standard database or library of video games from which a researcher could garner a representative sample. And secondly, there is no established way to describe a game’s protagonist – the main character you “become” when you operate the controller.

The students took on the task of designing a controlled vocabulary – a standard set of attributes that can be ascribed to any game and any protagonist. The job was more difficult than it might sound. In a game like “Tomb Raider,” you know that you play as Lara Croft, but in many games the lines aren’t all that clear. The protagonist’s race or gender may not be specified, or the lead character may not even be human.

With so many types of games, the students needed to provide clear instructions for how to categorize everything catalogers will encounter, said Samantha Coulter, one of the students on the project.

“The controlled vocabulary has a robust set of instructions for cataloging,” she said. “We ourselves used the controlled vocabulary to catalog some games, so we feel comfortable using it. It works. It’s effective.”

Kameron So, another student on the project, said that by building a structure for describing protagonists in the games, the group is opening doors to further study.

“The controlled vocabulary, I think, is a really important document because we have facets that really are not studied in academia,” he said. “For example, the facet of sexual and romantic relationships in video games. That type of data on protagonists doesn’t exist. We really think this has the potential of creating something helpful.”

To address the other problem identified by the group – the lack of a reliable data source – the students turned to techniques they’ve learned in data science classes at the iSchool. They scraped data from several sources to build a list of thousands of games that span the information age, from blips on the screen that pre-date “Pong” to the latest edition of “Halo.”

They then narrowed that list to a representative sample of about 2,600 games that researchers can study with confidence to make generalizations about the broader game archives. The students are naturally curious about what secrets are hidden in the data, but for now they’re satisfied with laying the groundwork for further study.

“I want to make sure the foundation is solid,” said Kate Sousa, another student on the project. “The foundation we are building, for this to be able to ramp up, I want to make sure that is 100 percent good to go.”

The Capstone project will help the GAMER group reach a key milestone in its long-term effort to organize and research video games and interactive media. The controlled vocabulary will be included in the iSchool’s Video Game Metadata Schema, which will soon be published in the Open Metadata Registry. Assistant Professor Jinha Lee said the metadata schema will become the go-to source for scholars to classify and comprehensively describe games.

“Our goal is to hopefully have all these different organizations that have game collections to start using it,” Lee said. “Ultimately we want something like a union catalog where we can use one standard schema to describe these games and we don’t have to duplicate the efforts.”

The Capstone students, she said, took on an ambitious, challenging project that will pay dividends for researchers in the GAMER group and elsewhere in the field.

The students hope their work leads to discussions based on facts, rather than assumptions, about video games.

“To be able to drive people to have data-driven conversations versus anecdotal, emotional and opinion-based conversations, I think that would be a huge potential win for us,” So said.