The Challenge
Static dialogue systems break immersion in narrative games. I needed conversations that adapt to time of day, relationship history, story progress, and player choices - creating believable characters that remember past interactions.
Traditional branching dialogue becomes exponentially complex with multiple variables. I needed a dynamic system that could handle contextual awareness without requiring manual scripting of every possible conversation state.
System overview sketch documented in Obsidian — rule-based architecture with modular flow
Technical Implementation
Inspired by Elan Ruskin's GDC talk on dynamic dialogue, implemented with custom fact dictionaries and priority selection.
Fact & Priority System
Tracks facts: Past actions, current situation, relationships
Picks the best response: More specific situations have higher priority
Automatic fallback: Plays default dialogue, so never breaks
Event-Driven Flow
Triggers: Player actions update game state
Supports: Multi-character conversations
Integrates: Portrait and audio systems
Contextual Awareness
World-aware: Time, weather, location, story progress
Memory: NPCs remember past interactions
Dynamic: No manual scripting of combinations needed
Results
Before
Static dialogue trees with manual branching for every game state. Characters felt robotic and conversations became repetitive quickly.
After
Dynamic conversations that adapt automatically. Characters remember interactions and respond contextually to any game state.