[EM] Condorcet-Approval hybrid method

[Daniel Bishop]

Russ Paielli wrote:

> Daniel Bishop dbishop-at-neo.tamu.edu |EMlist| wrote:

>> ...
>> There's a potentially important practical advantage, in that it 
>> allows voters to cast a Cardinal Rankings-style ballot.  For example, 
>> you could let:
>>
>> Rank 1 = ideal candidate
>> Rank 2 = candidate I have minor disagreements with
>> Rank 3 = candidate I have major disagreements with
>> Rank 4 = candidate I wouldn't vote for even if he were running 
>> against Hitler and Stalin
>
> [Questions rearranged for ease of explanation]
>
> I don't understand your point here. Are you ranking any of those 
> candidates equal?  Or are you proposing to rank more than one 
> candidate at each of the four levels?

Yes.

For example, consider an election with 12 candidates.  Your ballot might 
look like

_1_ Favorite
_2_ Good #1
_2_ Good #2
_2_ Good #3
_3_ Tolerable #1
_3_ Tolerable #2
_3_ Tolerable #3
_3_ Tolerable #4
_4_ Bad #1
_4_ Bad #2
_4_ Bad #3
_5_ Evil

> And why would you even rank/approve the last candidate?

First of all, I disagree that "ranking" and "approving" should be 
equivalent.  A better approach is to have the voter mark their least 
favorite approved candidate, and give an approval vote to everyone at 
that rank or better.  That forces the approval votes to be consistent 
with the rank list, while still letting a voter say "I don't approve of 
either Bad or Evil, but I'd rather have Bad than Evil".

The point of ranking a last candidate is exactly the same as ranking a 
first candidate: to express a preference.  But your proposal makes 
voting much harder for people with a strong last preference.