[EM] Scatter plot of clone independence versus IIA

[Kristofer Munsterhjelm]

On 02.07.2021 02:32, VoteFair wrote:
> On 6/27/2021 11:15 AM, Markus Schulze wrote:
>> Hallo,
>>
>> I recommend the following formulation for the fourth category:
>>
>> Suppose A won before cloning.
>> Suppose A is cloned to A(1),...,A(n).
>> Suppose A(1) wins when the used election method is applied only to
>> {A(1),...,A(n)}.
>> Then A(1) wins after.
>>
>> Markus Schulze
> 
> I am not understanding what you are suggesting.
> 
> It sounds like you want to sub-divide the fourth category into these two
> sub-categories:
> 
> * A wins before cloning, A(1) wins after cloning, and A(1) wins if only
> the clones participate.
> 
> * A wins before cloning, A(1) wins after cloning, but A(1) does not win
> if only the clones participate.
> 
> Do we need to distinguish between these two cases?
> 
> It seems to me that any of the clones winning instead of the original
> similar candidate is a case of the clone displacing the original candidate.

It looks like he's suggesting a related criterion (call it "independence
of non-clone alternatives") that is failed if A1 wins the ordinary
election (after clones have been added), but A2 wins when everybody but
the clones are eliminated.

That may be a desirable criterion, but if A won before the cloning
happened, I don't think it's a failure of the independence of clones
criterion.

-km