[EM] Condorcet-Approval hybrid method

[Russ Paielli]

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

>> I would start by simply choosing the CW if one exists, or paring the 
>> field down to the Smith set otherwise. Then I would eliminate the 
>> candidate with the lowest approval and repeat.
>>
>> I thought of this yesterday while I was working out, and I thought I 
>> had come up with something big. Then I searched the EM archives and 
>> discovered that Kevin Venzke had mentioned it way back in 2003.
>>
>> Oh, by the way, I would *not* allow equal rankings. Why not? I just 
>> don't like them.
> 
> 
> Not a very convincing reason to me.

Well, at least it's honest!

>> They strike me as an unnecessary complication
> 
> 
> How are they a complication?  If anything, equal rankings make it 
> *easier* to construct a pairwise matrix.

Equal rankings complicate matters for several reasons:

-- They create a potential debate about how they should be counted. 
Actually, this may not be an issue if you use the Condorcet-Approval 
hybrid scheme I explained above, so this objection may not be valid in 
this context. (The old "margins vs. wv" debate will be irrelevant!)

-- They complicate the vote-counting procedure and corresponding 
software. This may seem trivial in an academic environment, but in the 
real world it will be another potential source of bugs.

-- They complicate the voter interface and voting procedures. Yes, equal 
rankings are simple to implement with a vector of radio buttons next to 
each candidate's name, but I can virtually guarantee that such an 
interface will be considered unacceptable for general public use.

-- Last but not least, they complicate the explanation of the method to 
the general public. Just getting the public to understand and accept 
ranked voting will be a major endeavor in itself. Any unnecessary 
complications will just irritate the public and provide ammunition for 
those who dismiss the method as too complicated.

>> and little more than a way to game the system.
> 
> 
> 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

I don't understand your point here. Are you ranking any of those 
candidates equal? And why would you even rank/approve the last 
candidate? Or are you proposing to rank more than one candidate at each 
of the four levels?

> If there are a large number of candidates, this could be considerably 
> easier for the voter than casting a fully-ranked ballot.

You wouldn't need to rank any more candidates than you would approve in 
an Approval election.

--Russ