That's kind of strange, I only was peripherally reading the discussion and saw they were doing random/entropic strategies, which IMO you do if the problem space

That's kind of strange, I only was peripherally reading the discussion and saw they were doing random/entropic strategies, which IMO you do if the problem space is so large that you can't wrap around it using normal exploration techniques.