[EM] bullet voting and strategy on Approval ballots.

[Juho]

On Aug 31, 2010, at 7:39 PM, robert bristow-johnson wrote:

>
> On Aug 31, 2010, at 6:28 AM, Juho wrote:
>
>> On Aug 29, 2010, at 10:13 PM, robert bristow-johnson wrote:
>>
>>> On Aug 29, 2010, at 12:55 PM, Warren Smith wrote:
>>
>> ... long discussion on Approval and IRV ...
>>
>>>> These foundational problems strongly suggest that the entire area  
>>>> of rank-
>>>> order ballots, was a mistake, a road that should never have been  
>>>> taken.
>>>
>>>
>>> i know (and respect) that such is the product you're selling.   
>>> similarly to how FairVote sells IRV.
>>
>> Yes, FairVote and RangeVoting have some similarities :-).
>
>
> and i'm still waiting for someone to tell me how i should mark  
> (Approval or Score) a candidate who is neither my favorite nor my  
> most despised.  i know that my favorite gets approval=1 or score=99  
> and that Satan gets approval=0 and score=00, but what should Jimmy  
> Carter and Gandhi and Hitler get?

Here's one approach (a partial answer) to the problem.

If there are two potential winners, then approve the better one of  
those and all candidates that are even better. Don't approve the worst  
one of the potential winners and candidates that are even worse.  
Candidates that are between these two potential winners you may  
approve sincerely whatever way you feel like, and depending on if you  
want to increase their approval count or not.

If there are more than three potential winners then voting will be  
more complex. In some cases you can find a rational way to vote but in  
some cases you can not (e.g. in the Burr dilemma case).

In Range the recommendation is the same. Approval means max points. No  
approval means min points. Candidates between the two potential  
winners should get either max or min points, or you could use also  
some intermediate number if you don't have any strong interest to  
either increase or decrease their point count.

All this was based on the assumption that the election is competitive,  
i.e. you want your vote to have the maximum impact on the outcome.

>
> and then, even if someone gives me an answer for that, then how is  
> that not an act of strategic voting?  is there no strategy involved  
> when deciding to approve or score a candidate who you might  
> marginally approve of and is not your favorite?

The rules that I gave above are strategic in the sense that you need  
to analyze playing field and your strategic opportunities and then  
cast an optimal strategic vote. The rules are maybe not strategic in  
the sense that you would try to cheat the system. But if voters were  
advised to either mark all the candidates that they "approve" or give  
the candidates as many points as they are worth (instead of being  
advised to cast their best strategic vote), then these rules are  
strategic also in the latter sense.

Juho



>
> --
>
> r b-j                  rbj at audioimagination.com
>
> "Imagination is more important than knowledge."
>
>
>
>