Conectado: Connecting games, data, and research
Serious games have proven to be educational tools with numerous positive effects, capable of promoting learning, changing behavior, or improving training and skills development, among other effects. A big challenge is our ability to translate complex player activity into meaningful insights, both a data collection and processing problem. We spoke with Artem Vartanov, a fourth-year student in the e-learning group at La Universidad Complutense de Madrid (UCM), led by Antonio Calvo Morata and Baltasar Fernández Manjón about work he did as an intern to explore these issues using the serious game Conectado.
Conectado: A Serious Game to Decrease Bullying Behavior
Conectado is an example of a serious educational video game. Developed by researchers in the e-learning group at the Universidad Complutense de Madrid (UCM), it was designed to raise awareness about bullying and cyberbullying among teenagers aged 12 to 17. The game places the player in the role of a victim experiencing harassment over five in-game days, aiming to foster empathy and emotional engagement with the issue.
Figure 1. Screenshot of opening scene of Conectado game.
Conectado combines questionnaires and in-game player analytics to validate effectiveness. It has been extensively tested in schools across Spain with over 1,000 students. Pre-post testing indicates that playing Conectado significantly increases students' awareness of bullying, and over 80% of teachers report it as a useful tool for classroom discussions.
Scaling Impact
With the validation of Conectado, the UCM researchers have made the game freely available and open-source, as a web application and a desktop application for Windows and Linux operating systems, with translations in Spanish, English, and French. Conectado and serious games like it are particularly valuable because they compress time and eliminate real-world danger or costs, allowing students to safely experience and experiment with intricate concepts that cannot be accessed directly.
However, Conectado is not just a game. Serious games are designed to collect large amounts of gameplay interaction data, used to study human learning as an emergent property of the intersection of game mechanics and player activity. A highly promising direction is applying human-supervised machine learning to analyze this massive amount of data to detect critical behavioral patterns, such as player struggle.
Data Standards
A major issue in game analytics is the complete lack of a shared standard for recording player actions. Because games vary so widely in design and mechanics, building a unified data format that researchers can reuse without massive engineering overhead has proven incredibly difficult. With the increasing visibility and use of Open Game Data, the UCM e-learning group became interested in a collaboration. After discussions with Field Day Lab, it was decided to integrate Conectado into the Open Game Data repository. The main reasons were to make Conectado globally available and to facilitate widespread use of the game and its data. Furthermore, the integration aims to increase game discoverability and build trust by being accepted into the repository, which has a very rigorous game admission process.
During a summer internship at Field Day Lab, Artem worked on integrating Conectado into the Open Game Data infrastructure. “Conectado was originally developed using the data format called xAPI, an adaptation of a standard originally developed for military training and tailored specifically for serious games,” Artem explained. Open Game Data is developing a more detailed standard for in-game data logging, so a mapping from one format to another had to be developed and described so that in-game decisions and interactions could be accessed and analyzed. “One of the technical issues we had was language-based. The Open Game Data schema was built using ASCII for English-language code and games, and we needed to add unicode character support.” He also identified opportunities to improve integration workflows.
With those mappings, it then becomes possible to extract player event data needed to study struggle and other complex behaviors. “The next step,” said Artem, “is the training of LLMs to recognize complex behavioral patterns in gameplay data.”
“Working with Artem and Baltasar was a perfect example of how infrastructure can amplify great scholars,” says David Gagnon. “During his internship, Artem developed new research tools that allow qualitative researchers to openly explore game data, then train AI models to detect patterns. Applied to Conectado, this short project became his capstone project, but the tool is already being used by other researchers to study other games. Ironically, the first paper was submitted for review at the upcoming IEEE Conference on Games in Madrid, so we may be seeing each other again soon - this time on their turf!”
Artem’s capstone project [add link] covers some of these aspects: from using AI techniques to improve serious games to the use of different data formats and the application of LLMs, aiming to explain player behavior by analyzing gameplay logs.
Resources
Conectado home page with links to source code and player interface
Link to play Conectado (around 30 mins of gameplay)
Concectado gameplay data on Open Game Data repository
References
Calvo-Morata, Antonio; Alonso-Fernández, Cristina; Freire, Manuel; Martínez-Ortiz, Iván; Fernández-Manjón, Baltasar (2021) Creating awareness on bullying and cyberbullying among young people: Validating the effectiveness and design of the serious game Conectado. Telematics and informatics, 60: 101568. https://doi.org/10.1016/j.tele.2021.10156. Open Access Version: https://pubman.e-ucm.es/drafts/e-UCM_draft_363.pdf
Calvo-Morata, Antonio; Rotaru, Dan Cristian; Alonso-Fernandez, Cristina; Freire-Moran, Manuel; Martinez-Ortiz, Ivan; Fernandez-Manjon, Baltasar (2020) Validation of a Cyberbullying Serious Game Using Game Analytics. IEEE transactions on learning technologies, 13 (1): 186-197. https://doi.org/10.1109/TLT.2018.2879354
Calvo-Morata, Antonio; Freire-Moran, Manuel; Martinez-Ortiz, Ivan; Fernandez-Manjon, Baltasar (2019) Applicability of a Cyberbullying Videogame as a Teacher Tool: Comparing Teachers and Educational Sciences Students. IEEE access, 7: 55841-55850 https://doi.org10.1109/ACCESS.2019.2913573
