Threat, risk and harm: scoring of OCGs (Organised Crime Groups)

Research Institution / Organisation

The Open University

In Collaboration With

National Crime Agency

Principal Researcher

Dr Paul Mulholland

Level of Research

Professional / Work-based

Project Start Date

December 2017

Research Context

The National Crime Agency (NCA) makes use of an Organised Crime Group Mapping (OCGM) process to capture information about active Organised Crime Groups (OCGs) in the UK. A scoring mechanism is then used to translate the set of captured OCG features into a risk of harm assessment score. The risk of harm of each OCG is then use to assist with the prioritization of police activity and strategic reporting. The OCGM was originally developed in 2007. Currently, over 50 police forces and agencies submit information on OCGs on a quarterly basis. The reports describe each OCG according to a number of features. Each feature is broken down into a series of sub-questions.

The aggregated data from the contributing forces and agencies is used in a risk of harm calculation process that produces a score for each OCG. The risk score weights a number of factors related to intent, capability and criminality of the OCG.

The aim of the current project is to analyse both parts of the process by looking at:
1) The method currently used by OCG contributors to describe the features of the OCGs
2) The method currently used by the NCA to determine risk of harm from the aggregated OCG data.

Findings will be used to recommend improvements to how OCGs are described (i.e. how data is captured) and how those descriptions can be used most effectively in order to calculate risk.

Research Methodology

​The project will adopt a mixed methods approach including:

  1. Data modelling, analysis and interpretation
  2. Desk-based research to identify OCG/crime scoring mechanisms and alternative ways of calculating and modelling risk of harm
  3. User interviews to develop a deeper understanding of how OCGM is interpreted and used and elicit suggestions of how it can be improved

OCGM data will be prepared and stored in a form appropriate for analysis. This will involve negotiating access to an anonymised OCG dataset, agreeing a data management policy and translating and storing the data in a form suitable for analysis.  A formal model will be built encompassing the OCGM method plus other candidate methods such as MoRILE. Statistical techniques will be used to perform an initial analysis of the OCGM data. This will support analysis of features relevant to the descriptive and predictive properties of different scoring mechanisms.
A range of statistical and/or machine learning techniques will be used to inform recommendations for improving the calculation of risk of harm.

Interim reports and publications

​Report available for members of the Centre for Policing Research and Learning.

Date due for completion

March 2020
Return to Research Map