Toby Best

Queen Mary University of London

Toby has always held video games as an integral part of his livelihood, ever since catching his first Pokémon on the Game Boy Color. The ever-developing evolution of technology, from the humble NES and R.O.B. preventing the video game market crash in 1983, to the Wii’s motion controls, to augmented and virtual reality today, has been a key inspiration, and one of the reasons why he studied Mathematical Computation at University College London. He also has a keen interest in tabletop roleplaying games, such as Dungeons & Dragons and Pathfinder. His research interests involve the potential of combining roleplaying games' collective storytelling and interactive narrative with the power of artificial intelligence and deep learning.

A description of Toby's research:

Artificial Intelligence is the field of creating digital agents capable of decision-making and rational thought to fulfil a core goal or aspect. For tabletop and video games, an implemented AI would attempt to ‘solve’ the game by finding optimal winning strategies. However, tabletop role-playing games (TTRPGs) are driven by the power of collective storytelling and interactive narrative, as opposed to set rules, and therefore have a more open-ended goal - maximising player enjoyment for all participants. This involves a Game Master (GM) player as both narrator and referee, controlling the non-playable characters (NPCs) and the campaign behind the screen, whereas players usually control one player character (PC) each to interact with the world. There is no ‘failure’ state compared to traditional games, as campaigns can continue until players lose interest or the narrative is ‘complete’; even all PCs dying (known as a total party kill) can drive the narrative in a new direction. This project aims to study and piece together the different elements that would go into a Game Master AI, building on current state-of-the-art game-playing AI, such as Director AIs in games such as Valve’s ‘Left 4 Dead’, and studying the implications of such developments for players and game designers alike. For example, whether it could replicate the playing experiences of a human GM as a replacement, or enhance the experience by working with a human GM.