SimArena Documentation

SimArena is a powerful toolkit for creating, configuring, and running simulations with multiple agents in various scenarios. It provides a robust foundation for AI research and development, with support for custom objectives, brains, weapons, and maps.

---

Project maintained by arbyun Hosted on GitHub Pages — Theme by mattgraham

Usage Guide

Here are some common ways to use SimArena:

Quick Start

For detailed setup instructions and code examples, check out the specific documentation pages related to each component.

For configuration examples, check out our README file on GitHub. Alternatively, explore the configurations folder in the project for more examples.

Common Use Cases

1. Basic AI Testing

Create simple simulations to test different AI brain behaviors:

• Compare BrainA vs BrainB performance
• Test agent decision-making in different map layouts
• Evaluate weapon effectiveness across different scenarios

2. Game Balancing

Use SimArena to balance game mechanics:

• Run multiple simulations with different weapon configurations
• Test team compositions and map layouts
• Analyze combat effectiveness through the Map Analysis System

3. AI Development

Develop and test new AI behaviors:

• Create custom Brain classes that extend the base Brain
• Use Map Analysis data to make strategic decisions
• Test AI performance across different objectives (Deathmatch, Capture Point, etc.)

4. Research and Education

For academic projects and learning:

• Study emergent behaviors in multi-agent systems
• Analyze tactical decision-making patterns
• Generate data for research projects

Getting Started

1. Simple Approach: Create simulations directly in code (ok for quick tests)
2. Configuration Approach: Use JSON configuration files with templates (recommended for reproducibility)

Simple Approach Example

// Create a simulation with a randomly generated map
var simulation = new Simulation(width: 20, height: 20);

// Create the objective and tracker
var objective = new DeathmatchObjective(SimulationObjective.TeamDeathmatch, 2, 10);
var tracker = new DeathmatchTracker(objective);
simulation.SetObjectiveTracker(tracker);

// Create agents with brains
var (x1, y1) = Brain.GetRandomWalkableLocation(simulation.Map);
var brain1 = new RandomBrain(simulation.Map, team: 0);
var agent1 = new Agent(x1, y1, brain1, "Agent Red");
brain1.SetAgent(agent1);

var (x2, y2) = Brain.GetRandomWalkableLocation(simulation.Map);
var brain2 = new TacticalBrain(simulation.Map, team: 1);
var agent2 = new Agent(x2, y2, brain2, "Agent Blue");
brain2.SetAgent(agent2);

// Add agents to the simulation
simulation.AddAgent(agent1);
simulation.AddAgent(agent2);

// Run the simulation
while (!simulation.IsGameOver)
{
    simulation.Update(deltaTime: 1.0f);
}

Console.WriteLine($"Game Over! Winner: Team {simulation.WinningTeam}");

Configuration Approach Example

// Load the configuration
var config = GameConfiguration.LoadFromJson("configurations/team_deathmatch.json");

// Create the simulation
var simulation = new SimulationWithConfiguration(config);

// Run the simulation
while (!simulation.IsGameOver)
{
    simulation.Update(deltaTime: 1.0f);
}

Console.WriteLine($"Simulation '{config.Name}' completed! Winner: Team {simulation.WinningTeam}");

What’s Next?

Once you’ve got the basics working, explore further documentation:

• Custom Brains - Create your own AI behaviors
• Configuration - Learn about the template and configuration system
• Objectives - Set up different win conditions
• Weapons - Configure combat mechanics
• Maps - Create and customize maps
• Map Analysis - Analyze simulation data