> > The term "defeat-dropper" is a self-explanatory > newly coined (perhaps slang) reference to the pairwise > methods that "drop defeats", such as Ranked Pairs, > Beat Path, River etc.that are all equivalent when > there are three candidates. SCRIRVE and Raynaud are > examples of Condorcet methods that are not > "defeat-droppers".
That it is not self-explanatory is evidenced by that fact that I have asked before for folks to be more clear in their descriptions of proposed methods whether they mean "drop ballot contributions that resulted in a defeat" or "drop the candidate based upon number of defeats in the pairwise matrix." It is not at all clear what "defeat dropper" means, since my eye dropper doesn't dispense defeats. > > > Plurality passes Non-Drastic Defense. Suppose that > "to rank" means " to rank in a rank-balloting system". > In that case, no one can write a Plurality example > that complies with the premise of Non-Drastic Defense. > That means that no one can write a Plurality NDD > failure example, and that Plurality passes NDD. > > Yes, and by that logic, so does a pork chop. Exactly. So by LOGIC the definition of non-drastic defense is either a universal truth (like "pork chops exist") or is untestable. If it is untestable it is useless from an anylitical standpoint, and if it is a universal truth it should be called an axiom, not a criterion. ---- Election-methods mailing list - see http://electorama.com/em for list info