Accurately evaluating (i.e., reasoning over) large numbers of definitional rules (automated or not) can produce unexpected or non-intuitive results. Often, the end-user or business analyst will want to ascertain how the results were produced — that is, to audit the way the logic was applied. Doing that requires starting with the results and tracing back through the chain of rules that led to the results.
Automated support for such logic traces is an important platform capability. A logic trace permits flaws and opportunities for improvement in decision logic to be identified. It also provides a means for less knowledgeable workers to learn about business know-how in hands-on fashion.