[EM] Fluffy the Dog and group strategy

[Anthony Simmons]

>> From: Richard Moore <rmoore4 at home.com>
>> Subject: Re: [EM] Fluffy the Dog and group strategy

>> Reminds me of something I wrote a couple months back.
>> Someone posted an example, which if I recall was very
>> similar to Fluffy. I replied that it may be impossible to
>> tell from the ballots alone whether the third candidate was
>> in fact a centrist who was unpopular because the electorate
>> was highly polarized (and most people voted top ranking for
>> *their* extremist candidate), or if he/she was the village
>> idiot whose handful of close friends and family members
>> provided the only top-ranking votes. So should any method
>> give a different result depending on whether Fluffy is a
>> centrist candidate in a population that is unwilling to
>> compromise, or is indeed just a dog? How can the method make
>> this distinction?

In this case, is it a real distinction?  The method will tell
whether Fluffy is preferred over the opposite extremist or
not.  If the populace is merely unwilling to compromise,
Fluffy will get most of the preferences.  If Fluffy is a dog,
he/she will either get few preferences, showing that the
voters won't approve a dog, or will get many preferences,
showing they don't care.

If we need a method of determining whether Fluffy is a dog,
perhaps that is more in the domain of a vet than an election
method.