[EM] DMC/RAV vs. Approval

[Russ Paielli]

Hi Anthony,

Thanks for replying to my little poll.

Anthony Duff anthony_duff-at-yahoo.com.au |EMlist| wrote:
> Hi Russ,
> 
> I think DMC/RAV is less acceptable for public elections because the
> instructions for the voter will be too complicated for some.

Well, that is certainly a concern of mine too.

> I think a ranked ballot counted by a condorcet method, completed by
> anything reasonable, and approval, both would be excellent methods
> for elections.
> 
> I don't think that RAV adds enough benefit to either to warrent
> giving away the simplicity of the instruction given to the voter.  

The extreme simplicity of Approval is certainly a major plus for public 
acceptability and implementation. On the other hand, many IRV advocates 
dislike Approval because it forces them to "rank" their approved 
candidates all equal. They prefer IRV because it lets them rank the 
candidates in order of preference.

Most of us here realize that IRV ranking is misleading (because only 
some preferences are actually counted), but the ranking capability is a 
major selling point nevertheless. That's why I am beginning to think 
that DMC/RAV may actually be more "marketable" than Approval. You can 
tell the IRVists that DMC/RAV also lets you rank the candidates, yet 
unlike ordinal-only Condorcet, you don't need an eight-week course to 
calculate the winner.

As for the additional complexity specifying an Approval cutoff, we can 
tell voters that they can rank only their approved candidates if they 
wish, in which case they won't need a separate approval cutoff.

--Russ

> However, technically, RAV is superior, as it collects and uses more
> information.

Absolutely.

> Anthony