User:BetterVotingAdvocacy/Big page of ideas: Difference between revisions

If you're going to use [[:Category:Pairwise counting-based voting methods|Category:Pairwise counting-based voting methods]], I encourage considering using [[Pairwise counting#Negative vote-counting approach]] and [[Rated pairwise preference ballot]].

 

 

== Images ==

 

One way to think about how large the difference in required markings can be between pairwise counting and IRV is to consider an election scenario with a candidate who is a majority's 1st choice, where voters rank every candidate. In both Condorcet and RCV, the majority's 1st choice will win, but in RCV, only the 1st choices of each voter need be counted, resulting in one mark per ballot, while with pairwise counting, supposing there are 10 candidates, at a minimum 25 marks will need to be made per ballot (see [[Negative vote-counting approach for pairwise counting#Semi-negative counting procedure]]). However, pairwise counting will give more detailed information as to how much support each candidate has overall.

 

== Condorcet ==

* [[Approval voting]] and [[Score voting]] (and most [[cardinal method]]<nowiki/>s) have some kind of strategic equilibrium for the CW, especially if they are a [[Majority Condorcet winner]].

 

A basic reason to prefer [[:Category:Condorcet-cardinal hybrid methods]] over most other Condorcet methods (or at least, over the [[:Category:Defeat-dropping Condorcet methods]]) is that they allow the voters who prefer a CW to defend that candidate without needing to do Favorite Betrayal as much (though there may be errors with this analysis). As a general example, suppose there are two main candidates, with one being the CW, and there are some 3rd parties without about half as many pairwise votes in favor of them as the main candidates. The voters who prefer the losing main candidate can bury the CW under the minor candidates, and in the ensuing cycle, the non-CW main faction will win. There isn't anything that voters who prefer the CW as 1st choice can do to fix this, but the voters who rank a 3rd party 1st and the CW above the non-CW main candidate can do FB to prevent their favorite candidate from pairwise beating the CW. This ends the cycle and allows the CW's pairwise victory over the other main candidate to take precedence again. In rated Condorcet methods, however, FB isn't quite as necessary if the CW [[majority-beat]]<nowiki/>s the non-CW main candidate; this is because those who prefer the CW can do [[Min-max voting]] to give the CW maximal points by the majority and the non-CW no support by a majority; this will guaranteeably give CW enough points to win. More specific example of this at <ref>https://www.rangevoting.org/CondStratProb.html</ref> and some explanation of how majorities can force their preference in rated methods in the [[Approval voting]] article.

It is likely possible that the [[tied at the top rule]] can be made to work with something like Smith//Approval.

 

Note that when demonstrating the result of [[Smith//Score]] or similar methods, it isn't necessary to discover the entire Smith set in every case to find the winner. This is because if you confirm that some candidate is in the Smith set, and they have more points than all other candidates, then regardless of who else is in the Smith set, this candidate will win.