Expert Scorecards
Scorport enables users to design expert scorecards with its proprietary methodology, allowing you to create robust and scalable models for decision-making. Whether your scorecard is based on accumulated data, external data sources, or expert intuition, Scorport provides the tools to build precise and dynamic scorecards. This guide will help you understand how to use the Scorecard Designer to create an expert scorecard based on weights for Categories, Criteria, and Attributes (Bins).
Scorport Methodology Overview
-
Data and Model Integration:
- Scorport supports full-scale model development services, including AI/ML and econometric modeling. Refer to the Data Services section in Additional Services for details.
- If you lack accumulated data or need to build a scorecard quickly, Scorport's Expert Scorecards offer a structured methodology for defining weights and impacts.
-
Scorecard Structure:
- Scorecards in Scorport are based on three main levels:
- Categories: High-level data groups (e.g., Demographic Data).
- Criteria: Variables within each category (e.g., Age).
- Attributes (Bins): Ranges or values of criteria with defined impacts.
- Scorecards in Scorport are based on three main levels:
Building an Expert Scorecard
Step 1: Creating Categories
- A Category represents a group of related criteria, such as Demographics or Financial Data.
- Assign weights to each category, ensuring the total weight across all categories equals 100%.
- If the total weight is less than 100%, Scorport automatically allocates the remaining percentage to a Default Deficit Category.
Example:
- Demographic Data: Weight = 60%
- Financial Data: Weight = 40%
- Default Deficit Category: Weight = 0% (if total is 100%)
Step 2: Adding Criteria
- Within each category, define Criteria to measure specific variables.
- Assign weights to each criterion, ensuring the total weight within the category equals 100%. If there is a deficit, the system assigns it to a Default Deficit Criterion.
Types of Criteria:
- Range Type: Use when the variable has a range (e.g., Age with a range of 18-30).
- Variable Type: Use when the variable represents distinct values without a range.
Example for Demographic Data:
- Criterion: Age (Range Type)
- Weight = 50%
- Criterion: Employment Status (Variable Type)
- Weight = 50%
Step 3: Dividing Criteria into Attributes (Bins)
- Each criterion is further divided into Attributes (Bins).
- Assign an impact percentage to each attribute, reflecting its influence on the final score.
- Ensure at least one attribute has a 100% impact, representing the highest relevance to the dependent variable.
Example for Age Criterion:
- Range Type:
- 0-18: Impact = 20%
- 18-30: Impact = 60%
- 30-45: Impact = 100%
- 45-65: Impact = 80%
- 65-75: Impact = 40%
Step 4: Setting Scorecard Points
- When creating a scorecard, specify the minimum and maximum points for the entire scorecard.
- Example: 0-1000 points.
- The system automatically calculates points for Categories, Criteria, and Attributes based on their weights.
Step 5: Fine-Tuning and Dynamic Adjustments
- Scorport allows you to fine-tune and dynamically adjust:
- Category Weights: Changes here impact the distribution of points across all associated criteria and attributes.
- Criteria Weights: Modify the relative importance of individual criteria within a category.
- Attribute Impacts: Adjust the contribution of individual bins to the dependent variable.
Dynamic Scoring
- Scorport’s methodology ensures that any adjustment to weights or impacts is recalculated across the entire scorecard in real-time.
- This dynamic approach allows you to measure, optimize, and finalize scorecards quickly and efficiently.
Final Output
- The completed scorecard becomes an actionable tool that you can integrate into Strategies to measure and evaluate any incoming data.
- Use the scorecard as part of a broader decision-making framework within Scorport’s Strategy Designer.
Why Scorport Expert Scorecard Methodology?
- Customizable and Dynamic: Adapt scorecards to changing needs and data availability.
- Data-Driven and Expert-Driven: Incorporate existing data or rely on expert judgment to build effective models.
- Seamless Integration: Leverage the scorecard in strategies, making real-time decisions with the Executable API.
Scorport's Expert Scorecard Methodology is your go-to solution for building flexible, scalable, and effective scorecards. Whether you have data ready or rely on expert intuition, Scorport ensures that your decision engine operates at peak performance.