About the conflict prediction system (ViEWS)

The Violence Early-Warning System (ViEWS) is a publicly available, data-driven forecasting system. It generates monthly predictions of the number of fatalities in impending state-based conflict 1-36 months ahead, as well as probabilistic assessments of the likelihood that given conflict thresholds will be met – for each country and 55×55 km location in Africa and the Middle East. 

ViEWS is directed by Professor Håvard Hegre and based at the Department of Peace and Conflict Research at Uppsala University and Peace Research Institute Oslo. 


ViEWS started as an ERC Advanced Grant project in January 2017 in response to the need for a systematic and uniform means to monitor locations at risk of conflict.

The problem

Every month, thousands of people are killed in large-scale political violence. These conflict are highly destructive – beyond the immediate mortality and morbidity, conflicts lead to profound long-term consequences that impede the realisation of sustainable development. They severely hinder growth and poverty reduction, and they have severe detrimental effects on public health with the adverse impacts felt for decades after the shooting stops. Conflict is ‘development in reverse’.  Preventing and containing armed conflict is consequently high on policy-makers’ agenda. Early action, however, requires early warning. With a good understanding of where armed conflict will occur in the near future, for how long it will persist, and how lethal it will be, concerned governments, IGOs, and NGOs may engage in diplomacy efforts, alert the international community about the situation, prepare for humanitarian assistance, and allocate resources where most needed.  This is what ViEWS offers. 
The recent history of state-based violence in Africa, as recorded by the UCDP as of 1 June 2020. Red-colored cells suffered fatalities from state-based violence in April 2020. Purple cells have not experienced such fatalities in years.
Predicted probability of at least 25 fatalities in state-based violence in October 2020 (red dots) and September 2020 (grey dots), as seen from forecasts generated in August and July 2020, respectively.

The solution

The Violence Early Warning Systen (ViEWS) was created to remedy the problem above and bridge the knowledge gaps concerning areas of risk. The system offers a novel integration of isolated causal factors – identified through decades of peace research – into a theoretically and methodologically consistent forecasting system that is further informed by conflict data spanning nearly 30 years in time. ViEWS systematically monitors all locations at risk and produces uniform predictions of the number of fatalities in impending state-based conflict, coupled with probabilistic assessments of the risks of state-based, non-state, and one-sided violence, at both the country- and sub-national levels.
Coupled with state-of-the-art methodologies and an open-source ideology, the result is an ambitious and objective early-warning system at the frontier of research – a fully transparent system that is continuously developed, tested, and iteratively improved.

The forecasting tool

Based on the solution above, ViEWS releases two sets of monthly predictions for each country and 55×55 km territory (defined by the PRIO-GRID) in Africa and the Middle East – for each of the next 36 months: 1) Continuous predictions of the number of fatalities in impending state-based conflict at both levels of analysis, and; 2) Dichotomous predictions of the likelihood that at least 25 battle-related deaths (BRDs) will occur in each country and month, and at least 1 BRD will occur in each PRIO-GRID cell and month – both from each of three different types of political violence (state-based, non-state, and one-sided violence, as defined and recorded by the UCDP).  The forecasts are released through the ViEWS API and the interactive dashboard (in beta mode, limited data only), with written summaries in the Risk Monitor series. 
Prediction map for illustration of the ViEWS system
Predicted number of fatalities from state-based violence per country and month in October 2022, based on input data up to and including July 2022.