Global Terrorism 1970–2017

Attack Patterns & Success Predictors  ·  181,691 incidents  ·  GTD via Nano

About the data

The Global Terrorism Database (GTD) is an open-source dataset from the National Consortium for the Study of Terrorism and Responses to Terrorism (START), University of Maryland. It covers bombings, armed assaults, assassinations, hijackings, and more — from 1970 through 2017.

This analysis narrows to incidents with known attack type, region, and outcome — 181,691 records. Success is defined as the attacker achieving their stated or inferred objective. Data is queried live from Nano (table: DS_FACTORY.GTD_NARROW).

181,691
Total Incidents
88.9%
Overall Success Rate
34.8%
Middle East Share
48.7%
Bombing/Explosion

Attacks per Year — 1970–2017

Top 5 Regions by Attack Share

Attack Type Success Rates

Prediction Model — LightGBM

.937
0.9373
AUC-ROC
Key Finding: Attack type, region, and year are the strongest predictors of success. Assassinations are hardest to execute successfully (75.7%) while barricade operations almost always achieve their goal (99.2%).
2014 ISIL Surge: Attacks jumped from ~5K/year to 17,000+ in 2014 alone — this single year accounts for nearly 10% of all incidents in the dataset.