Risk Scores

Introduction

The Risk Scores, a fundamental feature of Country Watch, offer invaluable insights into Mexico's state-specific safety and security conditions. These scores are a cornerstone of our platform, providing users with data-driven assessments of risk. Within the Information Panel displayed on the main Country Watch view, you will find two essential risk scores:

1. Life Risk: This score evaluates the prevalence of violent crimes in each state, helping users gauge potential risks to personal safety and well-being.

2. Theft Risk: Focusing on crimes against private property, this score assesses the level of property-related incidents in different regions, aiding users in safeguarding assets and making informed decisions related to property protection.

DataInt, our trusted partner, compiles and delivers the Risk Scores, drawing on comprehensive data analysis from official sources. These Risk Scores play a pivotal role in helping users understand and navigate security challenges effectively, whether they are planning travel, managing business operations, or ensuring the safety of personnel.

In this documentation, we will explore the specifics of the Risk Scores, including their methodology, interpretation, and how to access and utilize them to enhance your risk assessments.

---

How to Access the Risk Scores?

The Risk Scores in Country Watch are easily accessible within the Information Panel of the main Country Watch view. These Risk Scores provide crucial insights into safety and security conditions and are conveniently located just under the "Map Controls" section. They are presented as two distinct time series, one for each Risk Score: the Life Risk Score and the Theft Risk Score.

The Information Panel is located on the left side of the Interactive Map and is the main component our application uses to display relevant information on the selected state. The Risk Scores section is located right under the Map Controls and displays two time series (one for each Risk Score).

To access the Risk Scores and gain valuable information on security assessments, follow these steps:

1. Open Country Watch: Begin by logging in to your Country Watch account and accessing the main dashboard.

2. View the Information Panel: On the main dashboard, you will notice the Information Panel displayed on the left side of the screen. This panel serves as a central hub for accessing essential information related to risk assessments.

3. Locate the Risk Scores: Within the Information Panel, scroll down just under the "Map Controls" section. Here, you will find two distinct time series representing the Risk Scores:

• Life Risk Score: This time series provides insights into the level of violent crimes in each state, helping you assess potential risks to personal safety and well-being.
• Theft Risk Score: The second time series focuses on crimes against private property, offering assessments of property-related incidents in different regions, which is essential for safeguarding assets and making informed decisions regarding property protection.

4. Interpret the Risk Scores: Each time series displays how the Risk Score has evolved in time. Right under each time series title, you will find key data highlights that will help you understand the latest value of the Risk Score:

• Most Recent Value: The first number represents the most recent value of the risk index for the selected state.
• Ranking: The second number reflects the ranking of the selected state in terms of the risk index in question compared to other states. This number is between parenthesis.
• Nominal Change: The third number indicates the nominal change in the index versus the last month, providing insights into the direction and magnitude of change.
• Relative Change (%): The fourth number represents the relative (percent) change in the index versus the last month, helping you understand the percentage increase or decrease in risk.

The numbers under each time series title help you gain a quick insight on the selected state current score and how it has changed compared with the last month.

Methodology

The Risk Scores featured in Country Watch are derived using a sophisticated latent variable model for count data with Poisson-distributed observations. This model serves as a powerful tool for summarizing crime incidence within a given state, whether it pertains to violent crimes or crimes against private property.

Objective

The primary goal of the Risk Scores is to provide users with a concise summary of crime incidence within specific categories, tailored to the context of either violent crimes or crimes against private property. These scores enable users to assess security conditions effectively and make informed decisions.

⚠️ Please Note The Risk Scores provided by Country Watch, along with their underlying model, are valuable tools for assessing security conditions within Mexico's states. However, it's crucial to note that these scores are not intended to reflect the probability of an individual being a victim of a crime. The occurrence of crime is a multifaceted and complex phenomenon influenced by a myriad of factors, including demographic, social, political, and even geographical variables. Crime incidence is shaped by a dynamic interplay of circumstances, making it challenging to predict at the individual level.

The Risk Scores are designed to offer an aggregate risk assessment for groups of target crimes within a given state. They provide valuable insights into general security conditions but should not be interpreted as indicators of personal risk.

We recommend using the Risk Scores as part of your overall situational awareness strategy. Stay informed, exercise vigilance, and make decisions that align with your specific circumstances and objectives.

Time-frame

To capture the most recent security conditions, the Risk Scores are computed based on data from the past three months. This time frame ensures that users have access to up-to-date information, while reducing data noisiness.

The Model

Behind the scenes, DataInt employs a latent variable model for count data. In this model, it is assumed that the observed counts of target crimes are generated from latent factors, referred to as "scores." These scores capture the underlying patterns in crime incidence within the state.

The Risk Scores model is Bayesian in nature. This means that it leverages Bayesian statistics, a powerful framework for statistical inference. Before any updates are applied to Country Watch, our team rigorously validates the model's convergence. This step ensures that the model has effectively learned the underlying patterns in the data and is providing reliable and stable Risk Scores.

Data Sources

The Risk Scores in Country Watch are built upon a foundation of robust data from official sources. These sources provide critical information that forms the basis of our Risk Score modeling. Here are the primary data sources that DataInt incorporates into the Risk Score calculations:

• Executive Secretariat of the National Public Security System (Secretariado Ejecutivo del Sistema Nacional de Seguridad Pública)

• National Board of Population (Consejo Nacional de Población)

By combining data from these two official sources, DataInt constructs a comprehensive and reliable Risk Score model that enables users to assess security conditions effectively and make informed decisions. These data sources, updated regularly and grounded in official statistics, contribute to the accuracy and relevance of the Risk Scores in Country Watch.