Methodology — How War Monitor Works
War Monitor aggregates open intelligence into a single live picture. This page explains where the data comes from, how the risk scores are built, and — importantly — what they cannot tell you.
Last updated June 19, 2026 · Maintained by Rumen Slavov
Data sources
- ACLED — Armed Conflict Location & Event Data: geo-coded battles, explosions and violence against civilians, with reported fatalities.
- UCDP — Uppsala Conflict Data Program: verified organized-violence events with actor attribution.
- GDELT — global news intelligence aggregating 10,000+ sources for news velocity and tone.
- OSINT & official reporting — vetted open-source channels and institutional reporting (e.g. IAEA on nuclear matters).
- Markets & tracking — oil/energy prices, AIS vessel traffic, and military flight data.
The Country Instability Index (CII)
The CII is a composite 0–100 score computed per country. It weights recent ACLED conflict events by type and reported fatalities, and combines them with security and information signals into unrest, security and composite components. A higher score indicates more measured instability over a trailing window. The CII drives the escalation-risk ranking.
Update frequency
Conflict event feeds refresh continuously throughout the day; risk scores recompute on a short interval; the daily intelligence briefing is regenerated each day. Nothing is cached for more than a few hours.
Limitations
These measures are descriptive, not predictive. They are constrained by reporting coverage — events in media-dark regions may be undercounted — and by the lag between an event and verified data. AI-generated summaries can contain errors and should be cross-checked against primary sources. Treat War Monitor as a fast situational-awareness layer, not a forecast or an authoritative casualty record.
Frequently asked questions
It builds a Country Instability Index (CII) from recent ACLED conflict events weighted by type and fatalities, combined with security and information signals into a 0–100 composite. This is descriptive of current conditions, not a forecast.
It draws on established datasets (ACLED, UCDP, GDELT) and official reporting, but it is limited by reporting coverage and data lag, and AI summaries can contain errors. It is best used as a fast situational-awareness layer cross-checked against primary sources.