The people that talked about the rules at Codurance was none of the people that defined the original rules. We had to "reverse engineer" the thinking, find reasons and then discard, change or let as they were (one of the objectives of the write up was to put down those reasons). Explicit recursion did take a while. The reasoning we came with is to learn to use map/reduce functions (do take into account that our group has an ecletic knowdledge of FP and hybrid languages: Clojure, Haskel, Elixir, Scala, F#, ...). Can that binary search tree be converted into a linear search if you rearrange the tree into a list? Yes. Now you can use map /reduce. Is it performant? Most probably not. Are there other ways? Time to discover. Interestingly, most of the katas/exercises on which we are practicing these rules are not heavily algorithmic in nature. Maybe we do need to add some to understand better and modify the "rules" with the new knowledge.