[EM] Poll on voting-systems, to inform voters in upcoming enactment-elections

[Richard, the VoteFair guy]

Kristofer's description (further below) of RCIPE -- Ranked Choice 
Including Pairwise Elimination -- is a bit flawed.  "Plurality" is not 
involved.  And correctly counting so-called "overvotes" is an important 
component of RCIPE.

Here's a correct description:

RCIPE:

This takes ranked ballots.
Modify IRV to eliminate "pairwise losing candidates" when they occur.  A 
pairwise losing candidate is a candidate who would lose every one-on-one 
contest against every remaining candidate.  (The mistakes in Burlington 
and Alaska would not have occurred.)
Elect the candidate who gets majority support after the eliminations.
Also, correctly count two or more candidates who are ranked at the same 
choice level.  Specifically, for example, pair up two ballots that rank 
A and B at the same currently-highest choice level.  One of the paired 
ballots is counted as support for A, the other is counted as support for B.


Bravo to Kristofer for writing these descriptions and links!!

Richard Fobes
The VoteFair guy


On 4/12/2024 9:19 AM, Kristofer Munsterhjelm wrote:
> On 2024-04-11 23:52, Richard, the VoteFair guy wrote:
>> I suggest creating a list of links to the associated Wikipedia or 
>> Electowiki articles, one for each method.  One person supplying all 
>> the links will save time for the rest of us.  Personally I'm going to 
>> need to read about some of these methods.
> 
> Good idea. Here are some links and brief descriptions. I've tried to be 
> neutral; I may post later with my thoughts about the methods, and what 
> notable criteria they pass and fail that I know of.
> 
> I've only done the methods that have a wiki page. At the bottom is a 
> list of those methods that don't. Their proposers have already described 
> them, but they might need a more clear description; in any case, I'll 
> leave that to them :-)
> 
> Smith//Score:
>      This is a composite method. 
> https://electowiki.org/wiki/Composite_methods
>      It takes rated ballots.
>      It first eliminates every candidate not in the Smith set, and then 
> does Range voting (Score voting) on the remaining candidates. No 
> normalization is done after the elimination, so a vote A: 10, B: 7, C: 0 
> still gives B 7 points and C 0 even if A isn't in the Smith set.
>      In Daniel Carrera's words: "The ballot would look like a Score 
> ballot. To process the ballots, the scores are converted into rankings 
> (equal rankings allowed), and the highest scoring candidate inside the 
> Smith set is elected." See also Toby's last post.
> 
> Approval with manual runoff:
>      First do a round of Approval voting. Determine the top two 
> candidates, then hold a later second round to determine who of the two 
> wins. This is like STAR, but spacing the rounds apart makes it possible 
> to hold debates, discuss the differences between the finalists, and so 
> on. On the other hand, it's more expensive.
> 
> Smith//Approval (explicit - specified approval cutoff):
>      This is a composite method. It takes a ranked ballot with an 
> approval cutoff marked.
>      Every voter approves of (gives a point to) every candidate ranked 
> above his approval cutoff. The Smith set member with the most points wins.
> 
> Smith//Approval (implicit):
>      Like Smith//Approval, except the voter can't specify an approval 
> cutoff. Instead, every candidate ranked is counted as approved, and 
> every candidate not ranked as not approved. The intent is to deter burial.
> 
> Woodall:
>      https://electowiki.org/wiki/Woodall%27s_method
>      This method takes ranked ballots. First make a note of the initial 
> Smith set. Then keep doing IRV until only one candidate of the initial 
> Smith set remains. Elect that candidate. Note: This is not Benham - the 
> initial Smith set never changes.
> 
> Schwartz-Woodall:
>      The same as Woodall, but with the Schwartz set. The Schwartz set is 
> slightly less prone to ties.
> 
> Copeland//Borda:
>      https://electowiki.org/wiki/Ranked_Robin
>      This method takes ranked ballots.
>      First eliminate every candidate not in the Copeland set. Then 
> calculate the Borda scores of the remaining candidates and elect its 
> winner. This can be done by using the Condorcet matrix alone.
> 
> Minmax(wv):
>      https://electowiki.org/wiki/Minimax_Condorcet_method
>      This method takes ranked ballots.
>      Based on the ballots, consider each one-on-one matchups between 
> candidates. Elect the candidate who has the most comfortable lead 
> against the candidate who does best against him, one on one.
> 
> Plurality:
>      You know this one. Choose-one ballots, equal rank not allowed, most 
> first preferences wins.
> 
> Majority Judgement (category):
>      https://en.wikipedia.org/wiki/Majority_judgment
>      Various tiebreakers exist, e.g.
> https://en.wikipedia.org/wiki/Highest_median_voting_rules#Tiebreaking_procedures
> 
>      This takes graded ballots. The candidates give each candidate a 
> grade, usually one of Excellent, Very Good, Good, Acceptable, Poor, or 
> Reject. For each candidate, calculate the highest grade so that a 
> majority gives that candidate at least that grade. The candidate with 
> the highest such majority grade wins.
>      The method is often indecisive, which means a tiebreaker has to be 
> used. There are many, but they all try to preserve the dynamics of MJ.
> 
> IRV:
>      https://en.wikipedia.org/wiki/Instant-runoff_voting
>      This should also be familiar to everybody here. It takes ranked 
> ballots. Repeatedly eliminate the Plurality loser until a candidate 
> obtains a majority of the non-exhausted first preferences.
> 
> STAR:
>      https://en.wikipedia.org/wiki/STAR_voting
>      This takes rated ballots. The ballot typically has six ratings 
> (zero to five inclusive), but the method could work with any number. The 
> method counts the total number of points for each candidate. The 
> pairwise winner of the two candidates with the most points wins.
> 
> Schulze:
>      https://electowiki.org/wiki/Schulze_method
>      This takes ranked ballots. The method is complex but basically 
> amounts to a cloneproof Smith generalization of Minmax.
>      Let a beatpath from a candidate A to another candidate Z be a 
> series of pairwise defeats A beats B beats C beats... beats Y beats Z. 
> The strength of the path is measured by the strength of its weakest 
> individual pairwise contest. Let "the" beatpath from A to Z be the 
> beatpath of strongest strength.
>      The Schulze winner is the candidate whose beatpath to anybody else 
> is at least as strong as their beatpath back to him.
> 
> Baldwin:
>      https://electowiki.org/wiki/Baldwin%27s_method
>      This takes ranked ballots.
>      Baldwin is the Borda analog of IRV. In each round, the candidate 
> with the lowest Borda count is eliminated. The last man standing wins. 
> Also called Total Vote Runoff by Foley and Maskin.
> 
> Black:
>      https://electowiki.org/wiki/Black%27s_method
>      This takes ranked ballots.
>      If there is a Condorcet winner, elect that CW. Otherwise elect the 
> Borda winner.
> 
> Approval:
>      https://electowiki.org/wiki/Approval_voting
>      This takes an Approval ballot: vote for one or more. The candidate 
> with the most marks wins.
> 
> Benham:
>      https://electowiki.org/wiki/Benham%27s_method
>      This method takes ranked ballots. Repeatedly eliminate the 
> Plurality loser until there's a Condorcet winner among the remaining 
> candidates. Elect the Condorcet winner when that happens.
> 
> Smith//DAC:
>      This is a composite method. 
> https://electowiki.org/wiki/Composite_methods
>      https://electowiki.org/wiki/Descending_Acquiescing_Coalitions
>      It takes ranked ballots.
>      First eliminate every candidate not in the Smith set. Then perform 
> the DAC algorithm to determine the winner.
>      The DAC algorithm is complex, with the result being a cloneproof 
> method otherwise close to Plurality.
>       Essentially, the DAC algorithm works like this: For each ballot, 
> for each set of candidates a voter acquiesces to, increase that set's 
> count by one. Then sort the sets in order of count, descending. Starting 
> at the top with the current viable set being every candidate, eliminate 
> from the viable set every candidate who is not in the set at the list's 
> current position, unless that would eliminate everybody. The candidate/s 
> remaining once you've traversed the whole list is/are the winner/s.
>      A voter acquiesces to a set if he doesn't rank any candidate 
> outside the set strictly above a candidate within the set.
> 
> RCIPE:
>      https://electowiki.org/wiki/Ranked_Choice_Including_Pairwise_Elimination
>      This takes ranked ballots.
>      Modify IRV to, instead of eliminating the Plurality loser, 
> eliminate the current Condorcet loser among the remaining (uneliminated) 
> candidates. Only eliminate the Plurality loser if there is no Plurality 
> loser. Once a candidate has a majority of the non-exhausted first 
> preference votes, elect that candidate.
> 
> RP(wv):
>      https://electowiki.org/wiki/Ranked_Pairs
>      This takes ranked ballots.
>      First count the pairwise matrix. Create a sorted list of every 
> pairwise victory, sorted by its strength. Break ties by random ballot.
>      Proceed from the top down the list, locking in a pairwise relation 
> unless it would create a cycle.
>      At the end, the locked-in contests give an order of candidates. 
> This is the election order: the candidate on top is the winner.
>      Strictly speaking, the tiebreaking is by random voter hierarchy, 
> but I've skipped the details about how that is constructed.
> 
> I think that should be correct, but if I forgot something or described 
> them incorrectly, let me know.
> 
> The methods not described are:
>      Margins-Sorted Approval
>      Double Defeat, Hare
>      Margins-Sorted Minimum Losing Votes (equal-rated whole)
>      Gross Loser Elimination
>      Max Strength Transitive Beatpath
> 
> -km