[EMAIL PROTECTED] wrote: > Mr. Ingles wrote- > > When looking at votes-as-cast, it's easy to show how Borda can elect a > unanimously despised candidate: > > Voter rating > <---preferred despised---> > ---------------------------------------- > 40 A C D E > 60 B C D E > > In other words, C, D, and E are all despised candidates. Even though B > has a 60% majority, C wins the Borda count with 200 points compared to B > with 180. > ---- > D- This is yet another example that later choice votes should NOT be counted > if they hurt an earlier choice.
Ah yes, but if the C candidates were closer to the preferred end of the spectrum then that "later choice" SHOULD be counted. Which simply makes the case for the argument that Borda is improved by truncation. In truncated Borda, if a candidate is close to the preferred end of the spectrum then rank that candidate, if not, then don't include that candidate. This gives the voter the power to decide whether it is worthwhile to allow his 2nd or 3rd choice a chance of hurting his 1st or 2nd choice. If the above voters take advantage of truncation, then B will be selected. And of course, the ultimate extension of this argument is to just use Approval. -- Richard
