Retired math prof's op-ed: AV over RCV in Duluth, MN

36 views
Skip to first unread message

Steve Cobb

unread,
Mar 26, 2015, 5:34:21 PM3/26/15
to electio...@googlegroups.com
Voting reform in the news:


In Response: Ranked-choice voting a mixed bag

by Kathryn E. Lenz

http://www.duluthnewstribune.com/opinion/columns/3708164-response-ranked-choice-voting-mixed-bag


Dick Burkhart

unread,
Mar 26, 2015, 8:11:06 PM3/26/15
to electio...@googlegroups.com

Interesting that she criticizes IRV (on practical, not theoretical grounds) and suggests approval or score voting instead. But a simplified Borda would be an improvement over approval and score needs some kind of constraint if it is not to degenerate into plurality.

 

Dick Burkhart, Ph.D., mathematics

4802 S Othello St,  Seattle, WA  98118

206-721-5672 (home)  206-851-0027 (cell)

dick...@gmail.com

--
You received this message because you are subscribed to the Google Groups "The Center for Election Science" group.
To unsubscribe from this group and stop receiving emails from it, send an email to electionscien...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Brian Goldman

unread,
Mar 27, 2015, 9:47:04 AM3/27/15
to electio...@googlegroups.com
score needs some kind of constraint if it is not to degenerate into plurality.
Please explain. I agree that Score can degenerate to Approval, but as long as your honest favorite and your lesser of two evils favorite are different candidates, Approval never becomes Plurality.

As to Borda, VSE suggests it reacts even worse than Approval and Score under tactical voting. The news article also raises the issue of IRV's complexity, specifically the need to educate voters on how to use the system. I suspect the same would be needed in Borda, but far less so for Approval.

Warren D Smith

unread,
Mar 27, 2015, 3:20:15 PM3/27/15
to electio...@googlegroups.com
I liked that op-ed. I'd never heard of its author before.

--
Warren D. Smith
http://RangeVoting.org <-- add your endorsement (by clicking
"endorse" as 1st step)

Dick Burkhart

unread,
Mar 27, 2015, 4:00:58 PM3/27/15
to electio...@googlegroups.com

For score (= range = cardinal) voting you could put your maximum score on your favorite candidate and the minimum score on all others. One way to prevent this is to require that your average score equal the midpoint of the possible scores (with automatic adjustments for deviations from this constraint).

 

Pure Borda is subject to tactical voting but a suitably modified Borda does quite well. For example, the simplest version of Borda for a single winner election uses only a voter’s top 3 rankings, so that you can vote “for” but not “against” candidates. In this case, undervoting can be penalized by the “modified Borda Count” of Peter Emerson. In some kinds of elections, you would need additional measures to guard against “clone” candidates. A suitable primary election could do this, if necessary using a method of proportional representation to identify distinctive voting blocks, so that only one candidate from each block is allowed in the general election.

 

Dick Burkhart

4802 S Othello St,  Seattle, WA  98118

206-721-5672 (home)  206-851-0027 (cell)

dick...@gmail.com

 

--

Brian Goldman

unread,
Mar 27, 2015, 6:12:55 PM3/27/15
to electio...@googlegroups.com
you could put your maximum score on your favorite candidate and the minimum score on all others.

This assumes that I care more about my favorite defeating my second favorite than any other pair in the set. Consider the election with the following candidates:
  • A, honest Score 9
  • B, honest Score 8
  • Z, honest Score 0
If I know nothing about how anyone else is voting, I would argue voting A=9, B=9, Z=0 is more strategic than A=9, B=0, Z=0. Similarly if I know that "A" has almost no chance of winning, with the race between "B" and "Z", the most strategic vote is A=9, B=9, Z=0, unlike strategic plurality where I would vote only for B.

the simplest version of Borda for a single winner election uses only a voter’s top 3 rankings

The VSE I linked is a 5 candidate race. So it is effectively using the voter's top 4 rankings. I do not see how ranking one less position will drastically change the strategic behavior. 
In some kinds of elections, you would need additional measures to guard against “clone” candidates. A suitable primary election could do this...

Why not instead use a suitable primary election to find the best two candidates? That way all voters can be completely honest in how they vote. This also raises the question about how to do the primary, which is by no means trivial, and is yet more complication on top of adding Borda itself. 
Reply all
Reply to author
Forward
0 new messages