PhD Research Application · April 2026
Ritwik Sharma
Game Designer · Systems Engineer · Researcher
Research Question
How do autonomous agents with private motivations, unreliable information, and social relationships produce narrative behavior that human observers recognise as coherent and meaningful?
ritwiksharma097@gmail.com +91 94623 32146 LinkedIn ritwiksharma-phd.pages.dev GitHub: Zombita Bikaner, Rajasthan, India
Documentation Index
Research Background
MA, Game Design — University of the Arts London (2021). MA dissertation examined branching, narrative illusion, and the conditions under which player choice produces genuinely emergent story — the same question this research approaches from the agent architecture side. Six years of production system-building since graduation, all converging on the same intellectual problem. Author of Under The Cherry Blossoms (2024), a published novel.
Why This Proposal Exists
The research question did not come from a literature review. It came from an idea in 2018, developed through an MA at UAL, and then six years of building systems that kept implementing the same concepts without a formal research frame. The Zombita platform — built in production before this proposal was written — implements the core research variables as working infrastructure. The proposal is the formal version of what the systems already do.
Three Research Pillars
Pillar I · RQ1
Information Reliability
How does variance in agent information reliability affect the structural properties of emergent social networks? Strategic deception, selective disclosure, and exaggeration as first-class design variables.
Pillar II · RQ2
Trust Modeling
What trust mechanisms are necessary and sufficient to produce narrative behavior perceived as socially coherent? Trust as a dynamic network property shaped by agent history and social position.
Pillar III · RQ3
Consequence Permanence
How does the irreversibility of agent actions interact with information reliability to affect perceived narrative quality? Permanence as a mechanic and a methodological variable.