User:BetterVotingAdvocacy/Big page of ideas: Difference between revisions
User:BetterVotingAdvocacy/Big page of ideas (view source)
Revision as of 20:16, 9 July 2020
, 3 years ago→Work-in-progress
(→Images) |
|||
Line 127:
I'm trying to evaluate whether is a way to essentially "track" a voter if voters are allowed to weaken their votes in Condorcet. By track I mean that you could figure out some voter's preferences by looking at the election result data; here is one example<ref>http://www.votingmatters.org.uk/ISSUE30/I30P2.pdf</ref>. My guess for how you could create an example where such a thing is possible is to have an election with few voters, where only 1 of the voters weakens their vote at all. Keep in mind that this may be somewhat realistic when considering that each precinct releases its own vote totals, such that a very small precinct may be vulnerable to this type of thing, if it exists.
I made some edits (https://electowiki.org/w/index.php?title=Rated_pairwise_preference_ballot&type=revision&diff=12325&oldid=11201) where I discussed some ways of figuring out what scores a voter would give to both candidates in a matchup (Ctrl+F "actual scores in the 1st vs 3rd matchup").
* But I see a problem with the second way I described: it doesn't work properly with regular Score voting. Suppose a voter scores A:4 B:2 C:1. Their 1st vs 2nd preference is A:4 B:2, and 2nd vs 3rd is B:2 C:1. If adding these up as prescribed to obtain the A vs C preference, you get 6 points for A and 3 for C, which reduces to 5 and 2 respectively if fit within a 0 to 5 scale. But the actual A vs C preference was A:4 B:1, which has the same margin, but different absolute scores.
** One way of solving this seems to be to instead do "take the lowest absolute score given to any candidate in any of the relevant matchups, and add the cumulative margin to that candidate's score to find the more-preferred candidate's score, moving this down as necessary to be capped at the max score." This always gives the appropriate result in regular Score voting. I wonder if it's equivalent or not to" take the highest absolute score, subtract cumulative margin, and move this up as necessary to be at least the min score."
== Condorcet ==
| |||