[EM] How choice of voting systems depend on amount of participants

Juho Laatu juho4880 at yahoo.co.uk
Tue Oct 7 01:29:29 PDT 2014


Condorcet methods are characteristically single-winner methods (but some of their characteristic properties can be used in multi-winner methods too).

There is no need to rank (or score) all candidates. Typically political elections have some major candidates (those that could win the election) and some minor canidates (those that are not expectd to win in any case). In such elections Condorcet methods work fine if most voters tend to rank at least n-1 of the n major candidates. Also shorter rankings are ok if the voter doesn't care which one of the remaining major candidates will win. And even shorter votes are still much better than not voting at all.

I assume that "simple majority" means plurality / FPTP. That method has many shortcomings. For example, if left wing (with about 50% support) has one candidate and right wing (with about 50% support) has two (major) candidates, then left almost certainly wins, even if right wing would get clearly more than 50% of the votes.

Also Condorcet methods are majority oriented in the sense that they compare pairs of candidates to see which one of them would win the other by majority (of those votes that preferred one of them to the other). Condorcet criterion (that determines which methods are Condorcet methods) says that if one of the candidates beats all others, then that candidate will win. Sometimes there can be a cycle of opinions (none of the candidates beats all the others), and then different Condorcet methods can make different choices (rarely).

Condorcet methods put some burden on the voters, since they (or most of them) are expected to list more than one candidate. Also the vote counting process is more complex than in plurality (typically done by computers since full hand counting (of all the pairwise comparisons) would be too tedious).

Juho


P.S. I also note that in countries that elect mutiple winners from single member districts one could try to use Condorcet separately in each single member distrct. That could however lead to some strange results. Condorcet methods tend to elect good compromise winners that are liked by all, but if those results are summed up, most representatives could be centrists, and the result would not be proportional at all in the sense that each party would get seats in proportion to its support.


On 07 Oct 2014, at 02:10, dikov dikov <dikov1 at yahoo.com> wrote:

> Hi,
> 
> Thank you for the reply and sorry for the pause.
> 
> I though that Condorcet methods are better when multiple places are concerned. Then scoring every entry brings relevant ranking. On another hand if only one winner is considered than simple majority system is accurate enough.  
> Is it correct, or  I am wrong?
> 
> Thank you,
> 
> DMytro
> 
> 
> 
> 
> 
> On Tuesday, September 30, 2014 12:37 PM, Juho Laatu <juho4880 at yahoo.co.uk> wrote:
> 
> 
> On 30 Sep 2014, at 07:10, dikov dikov <dikov1 at yahoo.com> wrote:
> 
> 
>> So if a contest assumes only one winner  than any method is valid, right?
>> 
> 
> There age good and bad methods. In this list (that is quite good) many people could recommend e.g. some Condorcet methods.
> 
> I brief, Condorcet methods are very good at least in a situation where
> - you elect one winner
> - the elecion is competitive (i.e. people try to vote so that one of their favourites will win)
> - you can afford the complexity of a method where the voters will rank the candidates (or at least some of them) in their preference order
> 
> (If you elect multiple winners, if the elction is not competitive (e.g. _neutral_ judges elect their favourite athlete) or if you need a simpler method, then it could be better to use some other methods.)
> 
> I write this just to make a long story short. Many (maybe most) people on this list (not all) think that Condorcet methods are good general purpose sigle-winner methods for competitive elections.
> 
> BR, Juho
> 
> 
> 
> ----
> Election-Methods mailing list - see http://electorama.com/em for list info
> 
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20141007/129df50f/attachment-0001.htm>


More information about the Election-Methods mailing list