About ConflictWatch

What this is

ConflictWatch aggregates academic data on organized violence — armed conflict between governments, between organized non-state groups, and violence directed at civilians. The figures come from the Uppsala Conflict Data Program (UCDP), a long-running research project that compiles recorded fatality estimates from public reports and codes each event to a place and a date.

The numbers presented here are recorded estimates, not exact counts. They are compiled after the fact from many sources, published as low, best and high ranges, and they routinely undercount — especially while events are still unfolding. ConflictWatch exists to provide historical and statistical context, with a focus on data quality and the human cost, not to track live situations.

• 📚 How organized violence is defined and categorizedUCDP’s three types — state-based, non-state and one-sided violence — and the fatality thresholds that separate an “armed conflict” from a “war”.
• 🗂️ How the data is collected and geocodedWhere UCDP figures come from, how each event is tied to an administrative area and date, and why there is a lag between an event and a confirmed record.
• 📉 Why the numbers are uncertain and usually undercountEstimate ranges, the difference between battle-related and total deaths, and why recorded figures should be read as a floor, not a final count.
• ⚖️ Sources, licensing and neutralityWhy ConflictWatch relies on an openly licensed academic source, how attribution works, and the commitment to a neutral, non-partisan presentation.

Reading conflict data responsibly

Conflict figures are recorded estimates about real human loss. Reading them well means leading with uncertainty, naming what each number includes, and being careful not to overstate what the data can show. This is context, not operational intelligence.

Data sources

• UCDP (Uppsala Conflict Data Program) — A leading academic source for data on organized violence, with documented definitions and methodology. Openly licensed under Creative Commons Attribution (CC-BY).
• UCDP definitions and methodology — The public codebook describing the categories (state-based, non-state, one-sided), the 25- and 1,000-death thresholds, and how battle-related deaths are recorded.

Frequently asked

Are these figures exact death tolls?

No. They are recorded estimates compiled after the fact from public reports, published as low, best and high ranges. They carry significant uncertainty and routinely undercount, so they are best read as a floor rather than a final count.

What is the difference between an “armed conflict” and a “war” here?

They are thresholds, not opinions. UCDP counts an armed conflict once it reaches at least 25 recorded battle-related deaths in a calendar year, and labels a conflict a “war” only when it reaches at least 1,000 such deaths in a year.

Why might these numbers differ from a news report or another dataset?

Different sources use different definitions, sources and units. ConflictWatch uses UCDP, which counts battle-related deaths on a consistent academic basis; figures are revised as information arrives, so they will not always match in-the-moment reporting or datasets like ACLED.

Do the figures include deaths from hunger, disease or displacement?

Generally no. The headline figures are battle-related deaths — those resulting directly from fighting. Indirect or “excess” deaths from disrupted healthcare, hunger and displacement are estimated separately and are often much larger.

Is this a live feed of current events?

No. This is aggregated historical and statistical data for context. There is a lag between an event and a confirmed record, and figures firm up over time. For current situations, defer to humanitarian organizations and official sources.