Talk:Explicit approval voting

From Electowiki
Revision as of 22:30, 11 November 2019 by Psephomancy (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Analysis of elections vs the method[edit source]

Actually I guess the Analysis should be its own article about the specific elections, not the method. Then it can be in Category:Elections Psephomancy (talk) 17:45, 10 November 2018 (UTC)

Equivalence to score[edit source]

The current article states this:

[Wikimedia's S/(S+O)] is mathematically equivalent to 2-level Score voting with averaging, though the abstain votes are explicit rather than implicit, and the levels are essentially (−1, +1) rather than (0, 1), so they are affected by the psychological consequences of disapproval voting.

Is that true? I don't think it is. Here's the FAQ answer for the 2017 version of the system:

Voters submit votes using a Support/Neutral/Oppose system. The votes will be tallied and the candidates will be ranked by percentage of support, defined as the number of votes cast in support of the candidate divided by the total number of votes cast for the candidate ("neutral" preferences are not counted, so this is the sum of support and oppose votes) - Support/(Support+Oppose). The candidates with support from at least 10% of voters and with the highest percentage of support will be recommended to the Board of Trustees for appointment, which occurs once additional verification of requirements is completed.

The way that I read that:

  • Step 1: Candidates must score higher than 10% approval. This is standard approval voting
  • Step 2: All "neutral" votes are no longer counted. The resulting election is now equivalent to a standard approval election, without the ballots of the "neutral" voters.

I'm not sure this can be converted to -1/0/+1 score system. -- RobLa (talk) 06:33, 25 November 2018 (UTC)

@User:RobLa: It can't be converted to a -1/0/+1 Score system, no. (-1, 0, +1) is mathematically identical to (0, 1, 2), and that produces different results in a few of the elections, as I said in the Analysis section.
It's not psychologically identical, though, as shown by Baujard et al. Explicitly voting "Oppose" is a form of negative voting, not just an abstention like it is under Approval. In standard Approval, you approve of some candidates, and leave the others blank, which can mean either disapproval or neutrality. It could be worded better. Psephomancy (talk) 04:58, 30 November 2018 (UTC)

@User:RobLa: I think it is mathematically equivalent to 2-level Score with Abstentions, though, so I don't think that part should be changed.

For example, if you had 5 ballots filled out:

   A: Support
   B: Support
   C: Support
   D: Neutral
   E: Neutral
   F: Oppose

and 4 filled out

   A: Support
   B: Neutral
   C: Oppose
   D: Neutral
   E: Oppose
   F: Oppose

Then the S/(S+O) would be:

   A: 9/(9+0) = 100%
   B: 5/(5+0) = 100%
   C: 5/(5+4) = 55.5%
   D: 0/(0+0) = 0%
   E: 0/(0+4) = 0%
   F: 0/(0+9) = 0%

If Support = 1, Oppose = 0, and Neutral = abstain, averaged score voting would produce:

   A: (1+1+1+1+1 + 1+1+1+1)/9 = 100%
   B: (1+1+1+1+1          )/5 = 100%
   C: (1+1+1+1+1 + 0+0+0+0)/9 = 55.5%
   D: (                   )/0 = 0%
   E: (            0+0+0+0)/4 = 0%
   F: (0+0+0+0+0 + 0+0+0+0)/9 = 0%

Psephomancy (talk) 02:40, 1 December 2018 (UTC)

What would you think if we changed it to this:
Approval voting is generally equivalent to 2-level Score voting (where the levels are "0" and "1"). Wikimedia's variant is equivalent to 2-level Score voting where voters may explicitly abstain, and the default choice is to abstain. The levels in Wikimedia's system imply (−1, +1) rather than (0, 1), so they may be affected by the psychological consequences of disapproval voting.
My hope is that this wording more clearly makes the point you're trying to make, which is that explicitly saying "disapprove" makes the election less "approval"-like in nature than some other wording might. -- RobLa (talk) 22:32, 1 December 2018 (UTC)
That sounds good, yes. — Psephomancy (talk) 06:29, 5 December 2018 (UTC)