Method support poll

From Electowiki
Revision as of 16:12, 9 September 2005 by James Green-Armytage (talk | contribs) (BTW, I was the one who changed the formatting (forgot to log in))

Jump to: navigation, search

Method support poll. Anyone with an interest in voting methods is welcome to participate. Do not alter the entries of any other participant. In this context, "supporting" a method means that you would an effort to adopt this method in the place of a method that you don't "support". If you oppose a method, that means that you might attempt to prevent the adoption of that method, even as a replacement for a method not on your "support" list. There is no need to list every method that you oppose; just those that you think are worth mentioning, i.e. at least somewhat controversial. Feel free to rank your supported methods in order of preference, or supply some other kind of comparison between them. Feel free also to modify your answers at any time; there is no closing date for the poll.


Single winner election methods

Which single-winner election methods do you support for the election of government officials? Unless otherwise indicated, it will be assumed that you support the method for local, state, and federal elections.


  • James Green-Armytage
Support: cardinal pairwise with beatpath, ranked pairs, 
river, or sequential dropping, AWP, CWO-IRV,
ER-IRV(whole), Approval voting, ER-IRV(fractional), IRV
Close to supporting: DMC/RAV, CDTT, IRV, Smith//minimax
Oppose: Borda count
Close to opposing: minmax, MMPO, Bucklin, Descending Acquiescing Coalitions
  • Kevin Venzke
In general I prefer methods without favorite betrayal incentive. I 
insist on minimal defense or something similar.
Support: ICA, Approval voting
Close to supporting: Schulze, Tideman, or River (using winning votes 
or possibly AWP), ER-Bucklin(whole), MAMPO
Support for legislatures only: CDTT-Later-no-harm combination methods
Oppose: plurality, IRV, ER-IRV, margins, Borda,
DSC, rating ballots
  • Juho Laatu
Support:
Condorcet: best for contentious elections if burying threat considered not too bad
Minmax(margins): my default reference point, good utility function, strategy resistance 
maybe not that bad
Path based (Schulze etc.) and winning votes: ok but lots of work for small 
improvements/changes (that are not all positive)
Pairwise comparison methods enhanced with additional approval or rating information: ok as well if 
not too complex for the voters
Approval: clean method and at least some improvement to the commonly user 
methods
Best IRV methods (if pairwise comparison based methods are out of question)
Close to supporting: Two round system (not that bad for multiparty countries)
Not supporting: Rating based methods (maybe not feasible for contentious elections like the 
government official elections of this query, but good for non-contentious elections)
Oppose:
Plurality (except that it is ok for two party countries IF they want to stay 
that way)
Regional winner gets all votes of that region
Borda and many other ("more heuristic") methods

Legislative election methods

Which legislative (multi-winner) election methods do you support for the election of government officials? Unless otherwise indicated, it will be assumed that you support the method for local, state, and federal elections.


  • James Green-Armytage
Support: CPO-STV, single transferable vote
Close to opposing: Cumulative voting, limited voting, SNTV
  • Kevin Venzke
Support: open party list (approval component), closed party list, 
possibly a proportional approval scheme
  • Juho Laatu
Support:
Open party list enhanced with hierarchical structure
Open party list
STV (non-party-based votes are both good and bad)
Regional representation (various styles to complement other methods)
Close to supporting: Closed party list, CPO-STV (complexity problems)
Not supporting: Two party methods (ok if kept intentionally, not just because of fear of 
changes or to stay in power)

See also