Banter System: The Lotus Mind Engine (v5.1.3)

The Banter System enhances atmosphere without distracting from gameplay. It creates the feeling of a strategy night with friends—except your friends are AI constructs.

"Like the lotus flower that rises pure from muddy waters, our AI emerges with personality, wisdom, and purpose."

Architecture Overview

The **Lotus Mind** is a tri-layer cognitive architecture:

┌─────────────────────────────────────────────────────┐
│                  LAYER A: EventBus                  │
│              (Detection & Routing)                  │
│  Monitors game events (Turn Start, Card Buy) and    │
│  triggers detection logic                           │
└─────────────────────────────────────────────────────┘
                         │
                         ▼
┌─────────────────────────────────────────────────────┐
│                  LAYER B: BanterEngine              │
│              (Personality & Selection)              │
│  Checks character DNA (Rabbit vs Tortoise) and      │
│  selects contextual dialogue from pools             │
└─────────────────────────────────────────────────────┘
                         │
                         ▼
┌─────────────────────────────────────────────────────┐
│                  LAYER C: BubbleManager             │
│              (Expression & Priority)                │
│  Manages on-screen bubbles, preventing overlap      │
│  and enforcing priority rules                       │
└─────────────────────────────────────────────────────┘

Priority System

Not all speech is equal. The **BubbleManager** ensures critical information is never hidden by chatter.

Character/Type	Priority	Notes
Emergency	15	System messages (Errors, Winners)
MuBot	10	Strategic advisor (Wisdom)
Joko	10	Hype commentator (Celebration)
Habot	5	AI opponent (Taunts)
Jabot	5	AI opponent (Analysis)
Narration	1	Generic game events

Rule: Higher priority bubbles can kick lower priority bubbles off the screen immediately.

Visual Constraints

Max Bubbles: 2 concurrent bubbles on screen
Duration: 4–10 seconds based on word count
Queue: Max 10 items (excess dropped to prevent spam)

Future Roadmap

Current architecture supports static dialogue for YouTube Playables (offline safe). Future versions will explore LLM API integration for dynamic responses.

See: Characters for personality details.