This approach seems very similar to working backwards when trying figure out the behavior of a function that calls other functions or libraries and when trying to a debug an issue and following the stack trace to determine the root cause of the issue. Not a perfect analogy but I find it helpful to think about it that way.