Wikipedia:WikiProject Social choice and voting systems

Structure

There is an overview of voting systems on the "Electoral system" page on English Wikipedia.

Ensure that every article has a category:

• Category:Electoral systems e.g. Single Transferable Vote
• Category:Voting theory e.g. Tactical voting
• Category:Voting theorists e.g. Marquis de Condorcet
• Category:Voting system criteria e.g. Monotonicity criterion

Avoid putting an article in multiple categories.

Issues to consider

Voting systems (more commonly known as "electoral systems") are a domain that has been analyzed from sharply different and contrasting disciplines: game theory, sociology, political science, decision science, cognitive science and economics. Our articles should give a clear summary of the analysis from as many of these domains as we can get, without overly relying on jargon from any. Where there is jargon, find the article that explains the jargon, and link to it. If there isn't yet an article, perhaps it's time time to create an article (see Help:Your first article to learn how to do this). If you do create an article, make sure that the terms of art are clearly explained.

Numbers

1. Count weakly-ordered ballots: Ballots should not be "dropped", and vote totals should not be "normalized", halfway through a voting process. This means you should avoid excluding exhausted ballots or indifferent ballots (ones that support neither candidate over the other). There's good reasons for this:
   • For many voting rules, these two measures of defeat strength can produce different results.
   • A 60:0:40 win is very different from a 30:50:20 win (the first is a landslide, the second means implies voters don't care).
   • Readers can always drop indifferent ballots to recalculate the results themselves. However, it is impossible to calculate the total number of exhausted ballots from a normalized vote total.
2. Droop quota: per this discussion, the Droop quota is defined as exactly ${\textstyle {\frac {\text{votes}}{{\text{seats}}+1}}}$. Try to avoid alternate quotas created by rounding or adding one extra vote. However, if you do have to round, be sure to round up.
3. Final scores: the final score for an election is the value that traditionally determines the order in which . Each candidate's score is scaled from 0% to 100%, with each candidate's "total" divided by the number of voters or valid ballots cast. For example:
   • Approval—results should be reported as approval ratings, out of 100%, not by dividing by the total number of approvals for all candidates.
   • RCV—report each candidate's performance in the last round before they were eliminated.
     • Avoid batch elimination, if possible—do one-by-one elimination until only a single candidate is remaining.
   • Minimax, ranked pairs, and Schulze—report the defeat strength of each candidate's weakest undropped defeat.
   • If ballot reports allow it, report both first-preference votes and the final score.

Advice

Remember that Wikipedia is an encyclopedia, so avoid minor details about the mechanics of counting or obscure criteria. Try to describe how a voter interacts with the system, and what ballots look like—what the voter expresses (e.g. rankings or ratings). Also describe how the winners are determined.

Give one or more examples if they help, and feel free to copy sample data from articles such as instant runoff. For single-winner systems, the example using the state of Tennessee and this map can be used quite effectively. Examples should strive to demonstrate possibilities for tactical voting, to be explained in the section below. They should also strive to demonstrate differences in outcomes between this system and other, similar systems, as well as unusual rules for the system.

Try to make mentions of prominent places where the system is used.

A section on controversy, issues, and advocacy is usually appropriate. This may be related to other sections, in particular information about tactical voting, the effect on factions, candidates, and the jurisdiction, as well as information on the voting system criteria the method passes and fails. Tactical voting can usually be explained fairly well in a short section.

Priority articles

Priority articles for improvement, in order:

1. Score voting & approval voting
   • Approval voting needs a reworked lead that discusses how it behaves strategically:
     • Debunking of old "bullet voting" myth
     • Myerson-Weber strategy
     • Leader heuristic
     • Dichotomous/two-party preferences
   • Score voting
     • Similar to above, but shorter
     • Discussion by experts in lead
2. Condorcet methods
3. Instant-runoff voting
   • Discussion of practical/implementation issues (e.g. spoiled/exhausted ballots)
   • Merge redundant sections
4. Round-robin voting (C2/majority matrix methods)
5. Plurality voting & first past the post (Merge needed)

Requested articles

These articles would be useful to have.

1. Condorcet dominance theorems
   • How Condorcifying a ranked voting rule creates a new rule that dominates the original
   • Discuss spoiler effects (Condorcet minimizes number of spoilers vs ranked voting rules)
   • Discuss strategy (Condorcet minimizes instances under which strategy can occur)
2. Defeat strength
3. Beatpath (See Draft:Beatpath)