This series examines what risk scores represent, how tools are built and evaluated, and how to interpret their outputs in real decision contexts.
The goal is straightforward: make risk assessment methods easier to understand—and easier to evaluate critically.
An introduction to a basic but often overlooked issue: not all risk scores mean the same thing, even when they appear similar.
What risk scores actually represent—and how to interpret them in practice.
How risk assessment tools are constructed, tested, and evaluated.
What common metrics like AUC measure—and what they can obscure.
How risk assessment tools are used in real decision contexts, including pretrial and corrections.