Reconstructing Narrative Complexity in Fallout 4

This project investigates narrative complexity in video games through a computational study of Fallout 4. It develops a pipeline that reconstructs phase-linked dialogue and interaction data from 51 main-story source quests, identifies speakers and character mentions with Python and spaCy, and represents the results as quest-, scene-, and phase-level interaction graphs. These outputs reveal recurring characters, faction-specific structures, and differences between broad cast expansion and dense repeated interaction. The project shows that computational narratology can recover meaningful narrative patterns from game data, while also highlighting that interpretation depends heavily on extraction, cleaning, and validation.