RN groups? The 10 RNe have some distribution of decline rates ( t ₃ in days) and orbital periods ( P _orbital in days), but there is no apparent clustering. Yes, we can always draw lines in such diagrams to create divisions (with the best try being to separate out those systems with red giant companions), but to be meaningful the groups must be either well-separated or caused by separate physical processes, while other properties should correlate with the putative division. None of the prior divisions amongst the wide variety of properties satisfy the need for a useful division.