Shahed Conflict Simulator

An interactive simulation modeling US/Iranian conflict dynamics, focusing on whether Iranian missiles/shaheds will exhaust US interceptor stockpiles or whether the US can destroy Iranian launch capacity first.

Key Finding: In Monte Carlo simulations using reasonable parameter ranges, Iran wins approximately 10% of the time.

Created by NunoSempere of Sentinel and Samotsvety. Code available on GitHub.

Features

Interactive Web Interface

Real-time Parameter Adjustment - Sliders for all key parameters with instant visual feedback
Scenario Presets - US Favorable, Iranian Favorable, and Baseline scenarios
Animated Simulation - Day-by-day progression showing conflict dynamics
Live Charts - Three interactive charts showing capabilities over time and daily activity
Winner Determination - Automatic analysis of which side achieves victory
Parameter Documentation - Info buttons with detailed ranges and sources

Monte Carlo Analysis

1 Million Simulations - Comprehensive statistical analysis
Parameter Sampling - Random draws from documented reasonable ranges
Fast Execution - ~600,000 simulations/second using Bun
Detailed Statistics - Win rates, conflict duration, and outcome distributions

Model Overview

State Variables

Iranian Side:

Stockpile (drones)
Production capacity (drones/day)
Launch capacity (launches/day)

US Side:

Interceptor stockpile
Interceptor production (interceptors/day)
Strike capacity (strikes/day)

Key Parameters

Interception success rate (90% baseline)
Production vulnerability to strikes
Launch capacity vulnerability to strikes
US strike system vulnerability to hits
Strategic allocation (production vs launch sites)

Model Dynamics

The simulation models a day-by-day conflict where:

Iran launches drones limited by stockpile and launch capacity
US intercepts drones with available interceptors
Drones that penetrate defenses degrade US strike capacity
US strikes degrade Iranian production and launch capacity
Simulation ends when any critical capacity reaches zero

Getting Started

Prerequisites

Bun (recommended) or Node.js
Modern web browser

Installation

# Clone the repository
git clone https://github.com/NunoSempere/shaheds.git
cd shaheds

# Install dependencies
bun install
# or
npm install

Running the Web Interface

# Start development server
bun run dev
# or
npm run dev

# Open http://localhost:5173

Running Monte Carlo Simulation

# Run 1 million simulations
bun run monte-carlo
# or
npm run monte-carlo

Expected output:

MONTE CARLO SIMULATION - 1,000,000 RUNS
...
OUTCOMES:
  🇺🇸 US Victory:          902,325 (90.23%)
  🇮🇷 Iranian Victory:      97,672 (9.77%)
  Stalemate:                   0 (0.00%)
  Ongoing (90 days):           3 (0.00%)

CONFLICT DURATION:
  Average:  6.40 days
  Median:   5 days

Running Command-Line Simulation

# Run baseline scenario for 30 days
bun run sim
# or
npm run sim

Project Structure

shaheds/
├── src/
│   ├── simulation.ts      # Core simulation logic (CLI)
│   ├── frontend.ts        # Web interface logic
│   ├── parameterInfo.ts   # Documentation and ranges
│   └── montecarlo.ts      # Monte Carlo analysis
├── index.html             # Web interface
├── plan.md               # Mathematical model documentation
├── parameters.md         # Parameter ranges with sources
└── README.md

Parameter Sources

All parameters are based on observed data from the 2026 conflict and open-source intelligence. See parameters.md for detailed ranges and sources including:

UK Centre for Information Resilience estimates
Military Times reporting
CSIS Missile Defense Agency data
Observed conflict data from 2026
Sky News analysis
Defense One articles

Key sources:

Iranian production: ~10,000 drones/month (CIR estimate)
US interceptor stockpile: SM-3 (414), THAAD (534), PAC-3 (thousands)
Interception rates: 90-93% observed in 2026 conflict
Launch vulnerability: 0.12 launchers/strike (calculated from 300/400 launchers destroyed)

Technologies

TypeScript - Type-safe simulation logic
Vite - Fast build tool and dev server
Chart.js - Interactive data visualization
Bun - Fast JavaScript runtime for simulations
Inter Font - Professional typography

Deployment

Vercel (Recommended)

# Install Vercel CLI
npm install -g vercel

# Deploy
vercel

Or connect your GitHub repository to Vercel for automatic deployments.

Other Options

Netlify - Build command: vite build, Publish directory: dist
GitHub Pages - Run bun run build and deploy dist folder
Cloudflare Pages - Automatic Vite detection

Model Assumptions

Simplified Conflict Model - Focuses on drone/interceptor dynamics, excludes many other military factors
Deterministic within Parameters - Daily calculations are deterministic; randomness only in Monte Carlo parameter sampling
No Escalation Ladder - Doesn't model nuclear escalation or other conflict expansion
Production Isolation - Assumes US interceptor production is safe from Iranian strikes
Homogeneous Systems - All drones/interceptors treated equally (no distinction between Shahed-136 vs Shahed-191, etc.)

Key Insights

From Monte Carlo Analysis

US wins 90% of scenarios with randomly sampled parameters
Conflicts end quickly (median: 5 days, average: 6.4 days)
Launch capacity destruction is the primary path to victory
Cost asymmetry remains challenging even with high interception rates

Critical Factors

Launch vulnerability - Most impactful parameter; determines how quickly US can degrade Iranian offensive capacity
Initial interceptor stockpile - Must last long enough for strikes to eliminate Iranian launchers
Interception rates - Even 90%+ rates strain stockpiles during sustained campaigns
Strike allocation - Balance between degrading production vs launch sites

Contributing

This is a research project by NunoSempere of Sentinel and Samotsvety. For questions or suggestions, please reach out.

License

MIT

Acknowledgments

Created by: NunoSempere of Sentinel and Samotsvety
Data sources: Military Times, CSIS, UK Centre for Information Resilience, Sky News, and other defense publications
Inspired by: Observed dynamics in 2026 US-Iran conflict and Ukraine war data

Disclaimer

This simulation is for educational and analytical purposes only. It represents a simplified mathematical model based on publicly available information and should not be used for actual military planning or decision-making. Parameter estimates contain significant uncertainty and the model omits many relevant factors in real-world conflicts.

NunoSempere/shaheds