AI Tree implementation

Overview

This document outlines the design of a behavior tree system for AI pawns in Gilded Honor. The system ties progression to unlocking new nodes on the behavior tree, enabling AI pawns to become more intelligent and offer new strategic options to the player.

Objectives

• Implement a behavior tree system for AI pawns.

• Tie progression to unlocking new nodes on the behavior tree.

• Enhance AI pawns' intelligence and strategic options as the game progresses.

• Provide a challenging and dynamic gameplay experience.

Behavior Tree System

Introduction

The behavior tree system is a hierarchical structure that defines the AI pawns' decision-making process. Each node represents a specific behavior or decision, and the tree branches out into sub-nodes for more granular actions.

Nodes

1. Root Node: The starting point of the behavior tree.

2. Selector Nodes: Nodes that evaluate children from left to right, executing the first successful one.

3. Sequence Nodes: Nodes that execute children from left to right, stopping at the first failure.

4. Decorator Nodes: Nodes that modify the behavior of a single child node.

5. Leaf Nodes: Terminal nodes that represent specific actions or behaviors.

Node Unlocking Mechanism

• Progression Tiers: Divide the behavior tree into tiers, with each tier representing a level of AI intelligence.

• Unlock Criteria: Define criteria for unlocking each tier, such as player achievements, game levels, or resource collection.

• Node Unlocking: Upon meeting the criteria, new nodes within the tier are unlocked, adding more complex behaviors to the AI pawns.

Progression System

AI intelligence Progression

• Champion progress through the game by completing missions, social activities, and by receiving gifts.

• Each progression milestone unlocks new behavior tree nodes for AI pawns, increasing their intelligence.

AI Pawn Behavior Progression

• AI pawns start with a basic set of behaviors in the behavior tree.

• As players progress, AI pawns unlock new nodes, enabling more advanced strategies and adaptive behaviors.

AI Pawn Behavior Examples

Starting AI

• Basic Movement: AI pawns move towards the player or predefined targets.

• Simple Attacks: AI pawns execute basic attack patterns with minimal strategy.

Mid AI

• Advanced Movement: AI pawns use cover, flanking maneuvers, and strategic positioning.

• Coordinated Attacks: AI pawns work together to execute more complex attack patterns.

Late AI Game

• Adaptive Strategies: AI pawns analyze player behavior and adapt their strategies accordingly.

• Special Abilities: AI pawns unlock special abilities or tactics unique to their type and inherent attributes.

Technical Implementation

Behavior Tree Framework

• Utilize a behavior tree framework compatible with the game's engine (e.g., Unity).

• Implement node classes for each type of node (Selector, Sequence, Decorator, Leaf).

Progression System Integration

• Integrate the progression system with the behavior tree framework.

• Define unlocking criteria and implement logic for unlocking new nodes based on player progression.

AI Testing and Balancing

• Conduct thorough testing of AI behavior at each progression tier.

• Balance AI difficulty to ensure a challenging yet fair gameplay experience.