Methodology
Data sources
- World Bank Open Data (SIPRI-based series) — military expenditure, armed forces personnel, population, GDP. Latest available year per country, 2023–2025.
- Open equipment listings (IISS-derived, via Wikipedia) — tanks, warships, combat aircraft, helicopters, satellites.
- Arms Control Association — nuclear warhead estimates.
- Strike-drone estimates — in-house scale built from open reporting on combat UAV and loitering-munition programs (tiers 0–4), with an estimate of annual output.
Power Index
Each component is log-normalized against the world leader (score = log(1+x) / log(1+max)), then weighted:
| Component | Weight |
|---|---|
| Defense budget | 23% |
| Active personnel | 14% |
| Combat aircraft | 14% |
| Fleet strength score | 14% |
| Nuclear warheads | 10% |
| Main battle tanks | 9% |
| Strike drones (est.) | 8% |
| Attack helicopters | 4% |
| Military satellites | 4% |
Fleet strength weights ships by class: carriers ×10, nuclear submarines ×6, cruisers ×4, destroyers ×3, frigates and amphibious ships ×2, corvettes and conventional submarines ×1–2.
Conflict model
The simulator computes an effective power for each side from air, land, sea and drone components (30/36/22/12% in a full-scale war), multiplied by a technology factor derived from spending per active soldier. Win probability follows a power-law contest function; projected duration grows as forces approach parity; loss ranges reflect force exposure and are shown with ±30% uncertainty bands.
Drones
Every country carries a drone-warfare tier from 0 to 4 (4 = mass combat employment and large-scale production: the US, China, Russia, Ukraine, Turkey, Iran, Israel) plus an estimate of annual strike-drone output. In the simulation this is a standalone power component — 12% in a full-scale war and 20% in an air campaign — and drone losses are modeled as fleet expenditure over the projected duration of the conflict. As of v0.3, drones are also part of the tabular Power Index (8% weight).
What the model ignores
Doctrine and training quality, terrain and geography, logistics depth, alliance commitments, morale, industrial mobilization and cyber capabilities. Results are statistical illustrations for education and entertainment — not predictions.