SMU Guildhall Graduate Catalog Spring 2020 — Cohort 28 | Page 50

Austin Merritt Level Design « Using Perk Choice to Personalize RPG Content My thesis artifact aims to offer a proof of For this project, I researched video game large open world Role Playing Games understand that personalization involves concept for personalized content within (RPG). I was initially inspired by tackling the challenge of coming up with a way to predict which content players will enjoy in Fallout to enable the creation of levels that feel handcrafted for any player’s specific loadouts and abilities. Over the course of a year with over 200 hours of work, I developed a Fallout 4 dungeon set that changes its enemy encounters, item placement, and room layouts to support the player’s predicted style of play. Playstyle predictions are calculated through the observation of the player’s perk choices within a perk selection interface accessible at the start of the experience. Once the perk data is recorded, it is fed into various systems within the dungeon that affect the player’s experience within the space. 50 LEVEL DESIGN personalization systems, which helped me a pattern of observation and prediction of player characteristics. By understanding these elements of personalization, I was able to better frame what the experience should and should aim to achieve. Beyond the theories I learned about and applied, this project also taught me a great deal about scripting in creation kit as I constructed new systems within the game’s framework.