[EM] approval strategy

[Kevin Venzke]

Russ,

 --- Russ Paielli <6049awj02 at sneakemail.com> a écrit : 
> I'm replying to myself because I would like to expand on the point I was 
> making.
> 
> A couple of proposals were made to alleviate the voting dilemma I 
> pointed out above. I did not reply to them because I honestly don't know 
> if they will work or not. I do know, however, that they compromise the 
> ultra-simplicity of Approval, which will make them harder to get 
> adopted. That certainly does not mean they are not worth discussing, 
> just that I'd like to focus on Approval and IRV for now.

Given the strategy problems you are concerned about, I hope you will
reconsider the "vote for and against" method I suggested. Perhaps it
does compromise the "ultra-simplicity" of Approval, but not by much:
Instead of being able to vote for any number of candidates, the voter
votes for one and against one. The "one man one vote" doubt mostly
disappears, also.

> Well, call me slow, but I am starting to realize that Approval has the 
> very same problem. As soon as a third party starts truly competing, 
> supporters of that party must decide when to drop their vote for the 
> "lesser of two evils." It's basically the same decision they must 
> ultimately make in IRV as to when to finally take the plunge and put 
> their real party at the top of the list.
> 
> So is Approval really any better than IRV? 

It doesn't seem so when you (apparently) are talking about a flank party 
candidate becoming competitive. But in Approval there is at least no
risk in voting for your preferred candidate. Also, IRV is not good at
finding compromise choices with few first preferences, whereas Approval
can do this easily, no matter how unviable that candidate seems, provided
that the voters like the candidate enough.

> Well, Approval *is* 
> considerably simpler than IRV, which a major factor for public 
> acceptability. And it *is* monotonic, which obviously desirable. And it 
> is "summable," which greatly simplifies implementation and reduces 
> security risks. Approval clearly has a lot going for it over IRV. But 
> when push comes to shove, how effective will it really be at giving 
> "minor" parties a chance to actually get elected?

If you want "minor" parties to "actually get elected," you might like
the VFA method. Suppose a majority vote against the Republican. Then
for the Green to win, he just needs more "for" votes than the Democrat.
And like I suggested in my first message, the strategy dilemma has
been moved from the Green voters to the Republicans.

> If Approval is ever adopted, prepare for some irate voters who pull back 
> too early on their vote for the "lesser of two evils" and have it 
> backfire on them. After they get burned on that one, many of them will 
> go back to their old ways and the duopoly could be back as strong as 
> ever. Or so it seems to me. Am I wrong?

If the Green and the Democrat are clones, I think there will be a
disaster every time.

If Democrats do not feel they need/want to give Greens a vote, then the
Greens shouldn't withdraw their vote for the Democrat.

So I think you're not wrong. But I'm not worried, since I don't think
the point of Approval is to elect Greens. (Though, it would be nice to
know how many supporters they have.)

Kevin Venzke



	

	
		
Découvrez le nouveau Yahoo! Mail : 250 Mo d'espace de stockage pour vos mails ! 
Créez votre Yahoo! Mail sur http://fr.mail.yahoo.com/