On nature's strategy for assigning genetic code multiplicity