The issue of a unique seed for each subdivision vs. using the same seed for all is so interesting that I added a switch between the two states (cyan group). When the same seed is re-used, edges are aligned across subdivisions, which takes some thinking to explain… It happens because the numbers are generated in random sequence and sorted, leaving gaps in shorter lists but using the same values.
Uneven_Subdivision_2018Mar31a.gh (37.7 KB)
The other error I made, effectively sorting the subdivisions by size, is also interesting and a valid objective. Subtle changes, critical differences!