[EM] Ranking unacceptable candidates.

[Juho Laatu]

On 18.10.2013, at 22.21, Kevin Venzke wrote:

>> I think ranked methods should encourage full ranking. And I don't like 
>> active truncation in general. The key reason is that if people start truncating 
>> the worst candiates, and easily also their strongest competitors, then we start 
>> losing valuable information. People may also start to truncate those candidates 
>> whose supporters they expect to truncate their favourite candidates. In the 
>> worst case we end up in having a plurality style election instead of a ranked 
>> election (when people close to bullet vote, or rank only the candidates of their 
>> own opinion group).
>> 
>> All information should thus be collected, and we should avoid opposite sides 
>> truncating each others. I guess it is quite common the elected candidate is 
>> disliked by many voters of the other opinion groups. But we still want to be 
>> able to elect one of the canddate in a rational way (= using preference 
>> information collected from the voters).
>> 
> 
> I feel the major problem is that method designers have nothing to offer 
> voters to persuade them to admit that the worse/worst frontrunners are
> better than anybody. It would be extremely rare to find that one actually
> needs the "worst frontrunner" as a compromise choice. It's totally evident
> that giving a preference to this candidate can only (and should only!) have
> the effect of helping him (even if it mostly does nothing), which is the
> opposite of what the voter showed up at the polls to do.

What I would like to get from each voter is ranking of all frontrunners except that worst of them (= ranked last by default). This gives us complete ranking of all the relevant candidates. If the left wing will win, I want all right wing voters to indicate that the more moderate one of the left wing candidates is better than the extreme leftist candidate. Otherwise the method may fail to elect the moderate left wing candidate that is obviously the more preferred one.

Ranking of the non-frontrunners is less relevant. The method will probably elect the correct winner in any case. For informative purposes it would be good if voters would rank also the nono-frontrunners.

I don't like the idea that unranked candidates would be considered "not approved" since that would encourage voters not to rank the candidates of the "opposite side".

>> Additional cutoffs/categories/approvals may help us in collecting full rankings 
>> from the voters. Harmless ranking is one approach to eliminating the risks of 
>> complete rankings. 
>> 
> 
> I can point out a concrete example with "Single Contest," under which there
> is an explicit approval cutoff, and adding or rearranging preferences among
> disapproved candidates can never cause the defeat of a candidate who was
> approved. It also doesn't open up any opportunities for other voters to take
> advantage of the possibility that I might specify those preferences. 
> (Contrast with MMPO, which still calls for defensive truncation despite
> LNHarm.) I do think that this level of guarantee would make me want to submit
> a full ranking.

It would be good to have appropriate means to get rid of strategic voting and truncation if those will be a problem in some environment. I lean in the direction that in most environments and (typical large public political) elections already basic ranked methods (with theoretical strategic vulnerabilites) are likely to work fine. I hope time will show us how different societies can provide full sincere rankings with different methods (when we get these methods implemented in real political elections). This depends both on the properties of the voters and on the properties of parties and media (whether they encourage or discourage strategic voting).

Juho