[EM] ruminations on ordinal and cardinal information

[Russ Paielli]

James Green-Armytage jarmyta-at-antioch-college.edu |EMlist| wrote:
> James G-A replying to Russ 
> 
>>My first comment is that this proposal is significantly more complicated 
>>than my (or Kevin's) "Ranked Approval Voting" (RAV) proposal, which 
>>simply drops the least approved candidate until a CW is found.
> 
> 
> 	Yes, I suppose the tally is harder to explain, although the interface is
> identical. However, you've not responded to the points I made in the last
> e-mail about strategic vulnerability in your RAV method.
> http://lists.electorama.com/pipermail/election-methods-electorama.com/2005-March/015226.html
> 
>>Secondly, I find it interesting that you define the magnitude of a 
>>defeat in step 3 as, "the number of voters who place A above their 
>>approval cutoff and B below their approval cutoff." This is equivalent 
>>to using the difference of Approval scores for the two candidates (with 
>>a constant offset). 
> 
> 
> 	No, it isn't. You can have a very strong defeat even when a candidate
> with a higher approval score beats a candidate with a lower approval
> score. You seem to be imagining that I wrote "the number of voters who
> place A above their 
> approval cutoff and B below it, ***minus the number of voters who place B
> above their approval cutoff and A below it***." But the second part of
> that sentence isn't in my proposal. This creates a world of difference,
> and that difference has important anti-strategic properties. Again, I urge
> you to read the cardinal pairwise paper and the strategy example posting. 
> http://fc.antioch.edu/~james_green-armytage/cwp13.htm
> http://lists.electorama.com/pipermail/election-methods-electorama.com/2004-September/013936.html

James,

I had misread your defeat magnitude as "the number of voters who place A 
above their approval cutoff plus the number who place B below their 
approval cutoff." Now I see that is not what you wrote, though the 
grammatical distinction is very subtle.

As far as I am concerned, your AWP proposal is far too compicated for 
public acceptance. I suggest you try the test I suggested earlier today: 
spend five or ten minutes explaining AWP to several people, then see how 
many can paraphrase it back accurately. That means they must be able to 
explain to a programmer all the rules that define the method in detail.

Your last step is to revert to Ranked Pairs or some such defeat-dropping 
scheme, which is already too complicated all by itself.

The defeat-dropping stuff is necessary in ordinal-only methods because 
they have no cardinal information. I don't understand why you want to 
retain those methods when simpler and more effective methods are available.

> 
>>If you simply used the winning approval score rather 
>>than the difference, I think your proposal would be equivalent to RAV.
> 
> 
> 	I don't use the difference.
> 
>>Elsewhere you have advocated using "winning votes" rather than "margins" 
>>for the pairwise measure off Condorcet defeat magnitude, but here you 
>>seem to be advocating "margins" for the defeat magnitude when it is 
>>based on approval scores. Why the difference? What is the advantage of 
>>using "margins" in your proposal? 
> 
> 
> 	This is exactly where you are misunderstanding my proposal. Cardinal
> pairwise and approval-weighted pairwise both use winning rating
> differentials rather than marginal rating differentials. That's what makes
> these methods interesting. Please read the references I give you.
> 
> 
>>More importantly, does the advantage, 
>>if there is one, justify the additional complexity of your proposal? I 
>>must tell you that I'll be surprised if it does. Then again, it won't be 
>>the first time I've been surprised.
> 
> 
> 	Yes, the benefits do justify the additional subtlety of the method.
> Because approval-weighted pairwise is more strategy resistant than RAV.

No, I don't think the benefits, if any, justify the additional 
complexity because they will prevent your method from ever getting used. 
Would you rather half a loaf of bread or zero full loaves?

--Russ