[EM] three tier approval

[Richard Moore]

LAYTON Craig wrote:
> Richard wrote:
> 
> 
>>The three-tiered rating could actually conceal the CW, so 
>>that you couldn't tell who it is from the ballots. If A and 
>>B are ranked equally on most of the ballots, but A beats 
>>every candidate including B, then A will appear to be the 
>>CW. But B could still be preferred over A by a majority, and 
>>that doesn't show due to the collapsing.
>>
> 
> When I said the condorcet winner, I meant the three tiered approval
> condorcet winner, not the sincere condorcet winner.  I was trying to say
> that the collapsing and the way everything is scored makes it much less
> likely (than RP) that pairwise victories will need to be skipped in the RP
> count.

I was just trying to interpret your meaning, which I took to 
be that "more likely to pick a CW" is relative to ordinary 
Approval's likelihood to do so. For the record, I almost 
wrote "sincere CW" instead of "CW" but then I realized that 
what I wanted to say was: "It conceals who the CW would have 
been (sincere or not) if full rankings had been allowed 
(just as Approval does)," so I struck "sincere". As for 
"three tiered approval condorcet winner", I don't know how 
meaningful that concept is. I could define "approval CW" for 
ordinary Approval but it's not a very helpful concept.

>>Of course, this is what happens in ordinary Approval, too. 
>>This method and Forest's Five-Slot Approval are 
>>finer-grained versions. I think your point system is a novel 
>>idea. I don't remember how Forest recommended scoring FSA.
>>
> 
> Novel?  Is that a compliment or a criticism? :-)

It's neither; it means that I haven't seen that particular 
scoring method proposed (though maybe I've missed it). 
Whether the scoring idea is a good one or a bad one needs to 
be studied.

> If I recall correctly, Forest wanted to score it as an Approval ballot
> (grades of A,B or C counting as approved, grades of D and F counting as
> dissapproved).  If I'm right, the point of the system was to give voters
> expressivity without changing the counting method.

Yes, that was one method but I think he suggested at least 
one other method. I could be mistaken. If that's the only 
method for counting FSA then it's coarser than I thought 
(equivalent to Approval), though it provides a way to gauge 
the strength of the winner's mandate.

Richard