Unrelated to the algorithm, I’ve found that doing things “by hand” is a surprisingly good way to develop intuition for otherwise opaque concepts. I wonder if it’s a function of doing work more slowly or paying closer attention. Examples include writing code written by others and working through math on paper.
I’ll commonly explore new tech by generating my own summarized documentation and through light tooling of initial configuration.
I am in the midst of this now where an existing operational “algorithm” needs better software support.
A lot of the discovery for that is starting with very general issues describing “problem” areas and then drilling them down to specific issues that might require action.
It can be tedious to go this “slow,” but it’s how I get to fast.