[EM] acceptable Chicken Proof methods (was chain climbing)

[Kristofer Munsterhjelm]

On 04/29/2014 12:53 AM, Forest Simmons wrote:
>
> Kristopher,
>
> you suggested ratings or grading, including MJ as possibilities.
>
> But these cannot (at least in their present forms) be chicken proof.

I suspect this is because voters can take them closer to Approval 
without repercussions. As I mentioned in another post here, all methods 
where candidate X's rating only takes the scores for X into account and 
the where function that maps the rating vector into a rating is 
monotone, one may move one's vote closer to Approval-style without being 
harmed. And Approval is, as we know, not CD-proof.

I'm thinking something that uses ratings/grades in a way that goes 
somewhat in a different direction than the rankings could work. Cardinal 
pairwise is an example (say, with one of the resistant Condorcet methods 
that have been mentioned). The very broad idea would be that the ratings 
let the method distinguish between honest ballots and strategic ones 
where this would otherwise not be possible. Of course, it's possible for 
the strategizers to manipulate the ratings, too, so this distinction 
must primarily come from the ballots that are not involved in the strategy.