G-Player is a web-based tool that allows complex visualizations and analyses of player behavior in a virtual environment. G-Player aims to provide the Northeastern Game Department with the ability to run data visualization on big-data queries to identify player behaviour trends between personality types.
British professor and game researcher Richard Bartle theorized that one could construct a classification system for behaviour how players approach and enjoy open world video games.
Bartle believes that a player’s behaviour type could be broken down along two axis: acting vs. interacting, and an interest in players vs. interest in the world. These axis provided four quadrants of behaviour types in video games: killers, achievers, socializers, explorers.
Can you analyze a player’s experience in a game world, and based on their actions determine their behaviour type? If possible, this could be done not through observation, but programmatically by the game itself - opening up a new field in video game technology where the game itself could adapt to the behaviour patterns of the player, and providing a more challenging and stimulating experience based on their identified player type.
G-Player is a data visualization web-app for uploading and viewing spatial-temporal player activity logs from video games. The project was developed in the context of visualizing a player's data as they played a custom mod within the game Fallout: New Vegas. This software attempts to provide the visual tools to enable game researchers to analyze player activity within video games. Our goal was to provide a user interface for exploring player data in order to understand actions, correlations, and trends.
G-Player was created as a team project in CS4500 Software Development - in association with Truong-Huy D. Nguyen, Game Development Researcher and Alessandro Canossa, Associate Professor of the College of Arts, Media & Design at Northeastern University.
As the lead developer for my CS4500 Software Development team, I was responsible leading the project, determining the architecture, and the majority of the front-end data visualization.
The research team wanted a tool that could be applied to any of their active studies in which player logs were stored. They needed a way to visualize hundreds of hours of player activity, including location data, game interaction, and player actions. As a result, we had to build out an API and Admin Interface for defining new data sets, and allowing the upload and input of arbitrary player data.
The front-end was a leaflet-powered mapping tool. Plotting player data based on preset or custom-defined database queries. It required the ability to view broad data sets, such as hundred of playthroughs, along with the ability to drill down into specific playthroughs to better understand individual player choices.