[EM] I hit upon why rating is easier than ranking.

Kristofer Munsterhjelm km_elmet at lavabit.com
Mon May 9 00:46:22 PDT 2011


⸘Ŭalabio‽ wrote:
> ¡Hello!
> 
> ¿How fare you?
> 
> In list Election-Methods run out of Electorama.Com, I hit upon why
> rating is easier and faster than ranking:
> 
> With rating, one determines the best candidate and gives that
> candidate the rating +99.  One determines the worst candidate and
> gives that candidate a score of -99.  One rates the other candidates
> scores relative to the best and worst candidates.
> Lacklustercandidates get the default score of 00.
> 
> With ranking, one must determine and remember the full rank order.
> It is very easy to forget:
> 
> “¡Darn!  ¡I should have ranked Candidate T between Candidates I  and
> L!”
> 
> This problem does not exist with rating.

It exists in the manner it would also exist in 0-to-n rated voting. If 
you forget to rate candidate X, he would get the default bottom rating 
of 0, same as if you forget to rank candidate X, he would get the 
default bottom "unranked" rank.

> Another problem with ranking is when multiple candidates are equally
> good or bad.  With rating one can rate them the same.

All the Condorcet methods and most other good ranked ballot methods lets 
you equal-rank. In a hypothetical election, you could easily do

FDR > Reagan > Hitler = Stalin

if you think both of the dictators are awful, or for that matter, just 
truncate your vote:

FDR > Reagan

which has pretty much the same effect.

> Another problem with ranking is how to rank the lacklusters relative
> to each other when one knows nothing about those nobodies.
> 
> As well as being easier on the voter, rating is easier on the
> pollworkers:
> 
> ¡Just sum the votes!
> 
> ¡Peace!
> 





More information about the Election-Methods mailing list