University of Twente Student Theses
Retrofitted: Enhancing Immersion in SoloRPGs through Large Language Models
Jordan, Matthew (2024) Retrofitted: Enhancing Immersion in SoloRPGs through Large Language Models.
PDF
1MB |
Abstract: | This thesis explores enhancing player immersion in Solo Role-Playing Games (SoloRPGs) through integrating Large Language Models (LLMs). Traditional SoloRPGs, often suffer from complex rules and setups that can hinder player immersion. Attempts to transform SoloRPGs into digital formats have had some success but at the cost of reducing player choice due to pre-coded limitations. This project investigates whether LLMs, known for their human-like conversational capabilities can provide a solution that enhances immersion while maintaining player freedom. The research is guided by the primary question: "To what extent can player immersion in a SoloRPG be enhanced by a natural language interface using large language models?" Supporting this inquiry are sub-questions defining immersion, measurement techniques, and applying LLMs in creating natural language interfaces. The methodology involved creating a modified version of the SoloRPG "Quill," incorporating OpenAI's GPT-3.5 Turbo model. This version was tested against the traditional game through user evaluations that measured flow, presence, and cognitive absorption. Findings from 36 participants indicated no significant difference in flow and presence between the traditional and modified versions. However, a significant increase in cognitive absorption was observed in the LLM-enhanced version, suggesting that natural language interfaces can indeed enhance certain aspects of immersion in SoloRPGs. |
Item Type: | Essay (Bachelor) |
Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
Subject: | 01 general works, 18 languages and literature, 54 computer science |
Programme: | Creative Technology BSc (50447) |
Link to this item: | https://purl.utwente.nl/essays/102964 |
Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Repository Staff Only: item control page