GrasshopperSim

An Evolutionary Sandbox & Game Theory Visualizer

A WebGL-powered ecosystem where artificial lifeforms inherit traits, compete for resources, and showcase emergent behavioral strategies over generations.

---

🧬 Theoretical Foundation

GrasshopperSim is not just a visualizer; it is an interactive laboratory exploring the fundamental mechanics of Evolutionary Algorithms and Game Theory.

The Evolutionary Engine

Agents in the simulation possess a discrete genome composed of continuous traits:

• speed: Energy expenditure rate vs. resource acquisition potential.
• jumpDistance & jumpHeight: Physical locomotion strategies.
• aggressiveness: The propensity to engage in combative interactions over resources.

Through continuous time simulation, agents that secure sufficient resources (health) live long enough to pass their optimized trait matrices down to the next generation.

Game Theory in Action: The Hawk-Dove Dynamic

The simulation organically manifests the classic Hawk-Dove game. When two agents compete for the same food node, their aggressiveness trait dictates the encounter:

• High-aggressiveness ('Hawks') risk fatal damage but gain a monopoly on resources.
• Low-aggressiveness ('Doves') share or flee, ensuring survival but missing optimal energy gains. Over time, the population approaches an Evolutionary Stable Strategy (ESS), visible in real-time through the metrics dashboard.

---

🏗️ System Architecture

Our engine decouples the reactive UI from the dense physics simulation, ensuring stable 60FPS updates even at high population counts.

graph TD
    classDef core fill:#1e1e1e,stroke:#3b82f6,stroke-width:2px,color:#fff;
    classDef agent fill:#2d3748,stroke:#10b981,stroke-width:2px,color:#fff;
    classDef env fill:#2d3748,stroke:#f59e0b,stroke-width:2px,color:#fff;

    A[Global Simulation Store]:::core -->|State Hydration| B(3D Render Engine)
    A -->|Metrics Polling| C(Telemetry Dashboard)
    
    B --> D[Grasshopper Agents]:::agent
    B --> E[Food Nodes / Terrain]:::env
    
    D -->|Sensory Data| F{Behavioral Heuristics}
    F -->|Action Choice| A
    E -->|Resource Decay/Spawn| A

Loading

The Life Cycle Loop

sequenceDiagram
    participant E as Environment
    participant G as Grasshopper (Gen N)
    participant S as Simulation State
    
    E->>G: Spawns Food
    Note right of G: Calculates distance & risk
    G->>S: Move / Attack / Eat
    S-->>G: Updates Health & Energy
    
    alt Energy > Threshold
        G->>E: Lays Egg
        Note right of E: Mutates Traits (Speed, Aggro)
        E->>S: Registers Baby (Gen N+1)
    else Energy < 0
        S->>G: Despawn (Starvation)
    end

Loading

---

🚀 Getting Started

Local Deployment

Ensure you have Node.js and npm installed.

# Clone the repository
git clone https://github.com/yourusername/grasshoppersim.git

# Install dependencies
npm install

# Instigate the simulation
npm run dev

Build for Production

To bundle the WebGL canvas and React bundle into a unified, high-performance static build:

npm run build
npm run preview

---

📊 Telemetry & Observation

The environment heavily features real-time metric tracking via recharts and echarts-gl:

• Population Velocity: Track rapid birth/death spikes in response to environmental carrying capacity constraints.
• Trait Drift: Graph average speeds and jump heights as the environment applies selective pressure.
• Generational Lineage: Track the longest surviving family names natively generated via the engine.

Note: The simulation state is volatile. Custom logic limits rapid changes to enforce evolutionary limits mathematically.