[Election-Methods] peer-reviewed work that is critical of IRV

[Abd ul-Rahman Lomax]

At 10:41 AM 9/28/2007, Howard Swerdfeger wrote:
>Not peer review, but Ka ping yee of...
>http://zesty.ca/voting/sim/
>has run sum simulations of election systems in 2d space
>and it shows quite a few problems with IRV.
>these guys (warren smith) also have some "Yee" diagrams
>http://rangevoting.org/IEVS/Pictures.html
>these show behaviour under different elections.
>
>
>In General IRV is bad at ties and near tie elections...in general

Right. But the zealous guardians of the IRV Official Line have 
specifically objected to Yee as biased and self-published, and to 
Warren Smith .... don't even mention him, the banshees make such 
piercing noises that we would not be able to sleep for a week.

Right now, the easiest path to get such information into the article, 
I think, is through Pro and Con arguments. My plan is to rearrange 
the article into an "Arguments" section that would examine each claim 
(made by either side, pro or con) in turn. Pro and Con argument 
sections, I've seen many time, do not allow direct comparison of arguments.

The question of certainty regarding fact is one which has long been 
an interest to me. People think that certainty is elusive or 
impossible, but actually, it's not all that difficult. The 
"encyclopedic" writer writes with certainty, if he or she is being 
thorough and very careful. If you source a statement, and you don't 
claim that the statement is true, but your source actually makes that 
statement, you are writing the truth, period. That's not the end of 
the story, for there is also the question of balance, and Wikipedia 
articles also require, if they can get it!, balance. It is not enough 
to collect a bunch of pro arguments, even if all sourced and properly 
qualified, while neglecting con arguments, or vice versa. It is easy 
to lie with the truth, and, in fact, it is the most successful form 
of lying, for lying through false statements is much more easily 
exposed. (It can still be difficult, but that's another story.)

It will take time, but I consider the IRV article to be politically 
quite important, and worthy of serious attention at this time. Google 
"Instant Runoff Voting," and, indeed a whole serious of topics, and 
what is at the top? Wikipedia. Many people are going to take their 
information from that article. We serve the public, and the public 
interest, and the interest of better elections in the future, by 
acting to ensure that the article is fair and balanced. If people are 
going to adopt IRV, I think they should do so with eyes wide open, 
not dreaming that it will balance their checkbook and put a cherry on 
every sundae. And, of course. "reduce negative campaigning." It's 
been really beautiful to watch these night-dwellers squirm when their 
beautiful propaganda piece is being transformed into something more 
accurately informative. They really can't stand the truth! And it's showing.

And, if they knew it, they wouldn't be doing what they are doing, for 
they are essentially tarring their own cause. It looks pretty 
suspicious when you actively try to suppress dissent, and the 
beautiful thing about Wikipedia is that every edit is recorded and 
the author identified. While these people are using the ability to 
edit "anonymously" and through sock puppets or meat puppets to large 
effect, I don't think they realize just how visible they are. And it 
will come out. A cause that must resort to deception is a lost cause.

And if anyone thinks my edits to the article have been wrong, or 
unfair, or biased, well, it wouldn't be surprising at least some of 
the time. That's the Wikipedia process, but I am actively 
participating in the attempt to find consensus, and that's what 
counts. I'm being pretty careful, though. I have *not* been attacking 
IRV, just attacking propaganda, spin, and just plain false claims. 
It's amazing what happens when one actually follows some of the 
sources used to back up claims. Sometimes what is in the sources is 
the opposite of what the claim implied.

FairVote is usually more sophisticated than simple false claims, 
though it may be turning out that some of the historical information 
they have provided about Bucklin, which we already knew was so 
heavily spun that you could weave a coat with it, is also false. 
Usually what they do is to spin and exaggerate with sometimes subtle 
shifts of language. The whole Robert's Rules issue is a very good example.

http://en.wikipedia.org/wiki/Talk:Instant-runoff_voting#Robert.27s_Rules_of_Order_in_Introductory_Section

http://en.wikipedia.org/wiki/Talk:Instant-runoff_voting#Continuing_problem_with_Introduction.2C_Robert.27s_Rules

Does Robert's Rules of Order, Newly Revised, "recommend" "Instant 
Runoff Voting"? FairVote would certainly want us to think that. What 
is the source for the claim. Is it Robert's Rules of Order? No. 
Actually not. The source is .... FairVote. Is the quotation taken out 
of context? Well, it could be, but in this case it appears that it is 
not. Rather, the claim is what one can come up with, easily, if one 
does not very carefully read what is there. Most people who are not 
election experts, in fact, would miss the major problem, and, irony 
of ironies, it is over an issue that impacts a major Pro-IRV claim.

Robert's Rules dislikes any election being won by a plurality, and 
the default is that it can't happen. If the bylaws are silent on it, 
no officer can be elected without the consent of a majority of those voting.

And, guess what? The rules are explicit that a "majority" means a 
majority of all those who have cast a ballot containing a valid vote 
for an eligible candidate. And such a vote is not "invalidated" by 
being cast for the wrong candidate, not one of the top two in the 
last round of an IRV election.

There is actually a series of problems with the claim that "Roberts 
Rules recommends IRV."

The source for the claim is another Wikipedia article:
http://en.wikipedia.org/wiki/Voting_methods_in_Robert%27s_Rules_of_Order

Which refers to RONR as a source, and is accurate as far as I've 
noticed, but then sources as well:
<http://www.fairvote.org/?page=1797>Robert's Rules of Order on 
Instant Runoff Voting, Fairvote.

Notice the title. FairVote wants to frame the discussion in the rules 
as being about "Instant Runoff Voting." That's not what RONR calls 
what it describes. And it is not exactly "recommending" what it does 
describe. It is giving an example of a preferential voting method, 
and it is clear that there are others, and it is not recommending one 
over another. Nor is it really recommending preferential voting, for 
it explicitly describes the serious problems with it, the possible 
failure to find a compromise candidate. It prefers election by 
majority vote, obtained by repeated balloting, which, in fact, 
properly done, can discover a Condorcet winner (at least "election by 
motion," which is a default method under Robert's Rules, will do this.)

The Wikipedia article on Voting Methods in RRO is pretty good, and 
accurate. I have edited the link above to replace, in the displayed 
text, "on Preferential Voting." I certainly won't be surprised if 
someone squawks.

I *want* them to squawk. When the bird squawks, the hawks know where they are.