Automation Confusion: A Grounded Theory of Non-Gamers' Confusion in Partially Automated Action Games

Title of the Paper

Automation Confusion: A Grounded Theory of Non-Gamers’ Confusion in Partially Automated Action Games

Paper Information

• Research Area: Digital Games and Human-Computer Interaction
• Keywords: Digital Games, Automation, One-Button Control, Shared Control, Non-Gamers, Confusion

Research Background and Problem

• Identified Problem:
  • The partial automation design in digital games aims to simplify gameplay by reducing players' operational burden. However, this may lead to confusion, especially among non-gamers, regarding control—players may struggle to discern which actions are controlled by them and which are managed by the automated system.
  • This phenomenon of "automation confusion" has not been systematically studied, and its prevalence and potential issues remain unclear.
• Significance:
  • Partial automation can reduce game complexity, making it particularly appealing to non-gamers. Addressing the confusion problem can improve the gaming experience for inexperienced users and enhance collaborative gameplay among players of varying skill levels.
• Research Motivation and Related Work:
  • The study is grounded in four research areas: non-gamers' experiences with video games, partial automation in games, users' mental models of automated systems, and users' perception of automation. While these areas have been explored to some extent, there is still a lack of comprehensive understanding of the mechanisms and types of confusion caused by partial automation.

Solution

• Research Methodology:
  • This study analyzes the gaming experiences of 10 non-gamers playing two partially automated action games.
  • Data collection methods include gameplay recordings, think-aloud protocols, and post-game interviews. A grounded theory approach is employed to analyze behaviors, psychological states, and related data, developing a theoretical framework for "automation confusion."
• Key Techniques and Implementation Steps:
  • Development of two partially automated games:
    • Ninja Showdown: A simple turn-based combat game where players control some actions while others are automated.
    • Spelunky: A platform action game where players trigger multiple potential actions with a single button, while movement and some decisions are automated.
  • Introduction of Awareness Cues: Indicators such as showing the character's current or next planned actions to help users understand automation.
  • Game design includes tutorial sections to familiarize players with controls and guide them through gradual onboarding.

Research Findings

• Main Discoveries:
  • Four Aspects of Automation Confusion:
    1. Types of Mental Model Errors:
       • Over-attribution (players believe they controlled actions that were actually automated).
       • Under-attribution (players fail to realize they could control certain actions).
       • Additional Rules (players invent rules that do not align with the actual game mechanics).
       • Simplified Rules (rules fail to account for the complexity of game outputs).
    2. Attitudes:
       • Analytical Attitudes (critical vs. non-critical analysis).
       • Emotional Attitudes (feelings of frustration and lack of engagement during gameplay).
    3. Behavioral Manifestations:
       • Learning behaviors, including exploration, validation, and observation.
       • Superstitious behaviors, such as reflexive button pressing and altering button-pressing patterns.
    4. Sources of Confusion:
       • Misinterpretation of feedback (e.g., misunderstanding system-provided information).
       • Confusion due to missing or unnoticed feedback.
       • Incorrect expectations (players form inaccurate assumptions based on prior experiences or subjective reasoning).
  • Positive Effects of Automation:
    • Despite the confusion, partial automation reduces the amount of control players need to learn, enabling non-gamers to play complex games.
• Comparison with Existing Solutions:
  • This study is the first to systematically reveal the complex phenomenon of confusion caused by partial automation and provides a theoretical framework for understanding the types of confusion.
• Limitations and Future Directions:
  • Limitations:
    • Small sample size, with the study focused on non-gamers.
    • Lack of comprehensive comparison between partial automation, full manual control, and full automation.
  • Future Directions:
    • Extend research to include a wider variety of games and a broader player base.
    • Explore specific strategies for improving user-friendly cues and interaction design.

Conclusion

This study proposes a theoretical framework for "automation confusion," offering an in-depth analysis of non-gamers' mental model construction, behavioral responses, and emotional attitudes in partially automated action games. By leveraging this framework, the research provides practical design recommendations to mitigate the potential risks of automation and enhance the playability of digital games for players with diverse skill levels.