Alessandro Canossa

The digital realm of play offers a novel and increasingly rich source of data for understanding the human condition. Unlike traditional research methods, games provide dynamic, high-resolution data capturing the fluid essence of human behaviour through every click, choice, and interaction. I envision a future where games are intentionally leveraged as controlled experimental environments – digital “petri dishes” – to study human psychology, social interactions, and cognitive processes at an unprecedented scale. By implementing a strategic roadmap, I am trying to unlock the transformative potential of game-based data, leading to profound societal benefits while proactively addressing inherent challenges.

My vision for realising this future rests on a pragmatic four-step roadmap:

• Step 1: Automatic Data Instrumentation (Read): The foundational layer involves establishing a universal data collection framework capable of capturing comprehensive player behaviour across diverse game engines and platforms. This includes cross-platform, cross-engine instrumentation and memory-level data capture, moving beyond manual developer integration. An automated and open-source framework will minimise overhead, while integration with qualitative methods like surveys and interviews will provide richer context for behavioural analysis.

   

• Step 2: Universal Modding System (Modify): With robust data capture in place, the next step is creating a unified modding framework that allows researchers to easily adapt games for experimental purposes. This includes modifying game rules, adding new mechanics, and adjusting difficulty to test specific behaviours. Even for games lacking native modding tools, dynamic injection via memory manipulation should be possible. This not only benefits research but also offers game studios advantages in rapid prototyping and balancing.

   

• Step 3: Procedural Content Generation & Generating Whole Games (Create): Moving beyond modifying existing games, the future includes the AI-driven generation of entirely new game content and even whole games tailored to specific research needs. Procedural content generation (PCG) can automatically create game worlds, levels, and mechanics. Furthermore, AI-driven “synthetic humans” (NPCs mimicking real players) can populate these environments, allowing for initial testing of different behaviours before human interaction. These AI agents can also adapt game parameters dynamically to match research requirements, enabling scalable and personalised research environments.

   

• Step 4: Evaluation & Understanding: The final step involves deriving meaningful insights from the data collected in modded or AI-generated game environments. This necessitates AI-assisted dashboards, explainable AI models, and sophisticated visualisation tools to identify patterns and trends. Benchmarking and reproducibility are crucial, allowing future studies to build upon past findings. Ultimately, the goal is to develop a “meta-theory” of human behaviour in digital spaces, informed by the insights gained.

The realisation of this roadmap promises a wide array of societal benefits:

• Improved Policy Making: Games can provide valid and reliable data to simulate real-world scenarios, enabling policymakers to test the effects of different interventions and make more informed, data-driven decisions. This can lead to policies that better address citizens' needs and improve their quality of life.
• Enhanced Social Understanding: Game-based research can provide a deeper understanding of human behaviour and social dynamics. This knowledge can equip educators and social workers with tools to promote positive social interactions and address social issues. Furthermore, it can foster virtual companionship, ease intercultural communication, and contribute to personalised assistive technologies.
• Advancements in Psychology: Detailed behavioural models derived from game data can lead to a deeper understanding of human cognition, emotions, and decision-making. The development of digital twins or synthetic humans could provide highly detailed, dynamic models for studying psychological processes in controlled environments. Predictive analytics based on cyber personalities could also aid in understanding and predicting individual and group reactions.
• Advancements in Healthcare: Changes in player behaviour could serve as indicators for the early detection of mental health issues. Monitoring player activity, sleep patterns, and nutrition data could identify potential physical health risks. Furthermore, player data could be used to tailor personalised treatment plans and develop adaptive medicine and virtual therapists. Rigorous validation and integration with healthcare systems are crucial for these advancements.
• Educational Advancements: Games can create more engaging, adaptive, and effective learning experiences for students. They can provide educators with innovative teaching tools and realistic simulations, ultimately improving student outcomes and engagement.
• Improved Urban Planning: Games can accurately model real-world urban environments, allowing urban planners to test out designs before real-world implementation. This can lead to more liveable, efficient, and sustainable cities, while also providing tools to simulate the environmental impact of urban development.

The potential of games to serve as powerful tools for understanding human behaviour is immense. I envision the research community, game developers, policymakers, and the public to actively engage in discussions surrounding the ethical and practical implications of this transformative vision. We must collectively address the challenges and work towards realising the significant societal benefits that game-based research can offer.

Given the possibilities, the potential pitfalls, and the evolving technological landscape, the central open question remains: How can we collaboratively and responsibly harness the power of games to unlock a deeper understanding of the human condition and translate these insights into tangible benefits for individuals and society as a whole?

By embracing a strategic roadmap focused on data instrumentation, modding, AI-generated content, and robust evaluation, we can transform video games into unparalleled research platforms. This shift has the potential to revolutionise policymaking, enhance social understanding, advance psychology and healthcare, and improve education and urban planning. While acknowledging and proactively mitigating the inherent challenges related to ethics, data privacy, and ecological validity is crucial, the future of game-based social science holds the promise of unlocking profound insights into the “angels and beasts” within the human psyche, driving innovation across diverse fields and contributing to a more nuanced understanding of ourselves.