On Jul 23, 2011, at 3:59 PM, preferentiality wrote:
> Ranked Ballot (voters ranking candidates in order of preference) will
> give us (PRACTICABLE!) Instant TRUE Democracy for ALL the World, &
> even put an end to all war forever.
>
> Because it gives an equal chance of winning to not just all parties,
> but all combinations of programs, “RB” is the only thing that’s truly
> both just & free. Because it always elects the candidate most exactly
> in the middle of all voting, RB is top-dead-center counter extremist,
> & thus more anti-terrorist than all the many recent retrenchments
> combined & thus will even disallow the tendency of (virtually two-
> party) parliamentary systems to give the top to the biggest gang on
> the block, sometimes with violently extremist results.
Worth reading more - I am not buying the author's claims of achieving
perfection.
Dave Ketchum
> --
I see it as worth considering the thinking, although I AM NOT signing
on as backing any of it.
On Jul 23, 2011, at 11:32 PM, Dave Ketchum wrote:
> Knowing of IRV and Condorcet methods of counting ballots, the first
> paragraph below makes me wonder how valid the the author's claims
> may be. The very last few lines help.
> . STV - not used here - THANKS
> . Condorcet - also not used here - think more whether this is
> better.
>
> On Jul 23, 2011, at 3:59 PM, preferentiality wrote:
>
>> Ranked Ballot (voters ranking candidates in order of preference) will
>> give us (PRACTICABLE!) Instant TRUE Democracy for ALL the World, &
>> even put an end to all war forever.
>>
>> Because it gives an equal chance of winning to not just all parties,
>> but all combinations of programs, “RB” is the only thing that’s truly
>> both just & free. Because it always elects the candidate most
>> exactly
>> in the middle of all voting, RB is top-dead-center counter extremist,
>> & thus more anti-terrorist than all the many recent retrenchments
>> combined & thus will even disallow the tendency of (virtually two-
>> party) parliamentary systems to give the top to the biggest gang on
>> the block, sometimes with violently extremist results.
>
> Worth reading more - I am not buying the author's claims of
> achieving perfection. There are many topics for which more than two
> possible choices seem worth debating possible value.
Ranking candidates is only part of the process. How would "RB" tally its
rankings? Instant Round Robin Voting (IRRV) uses ranked ballots, but so does
IRV, and they're poles apart in quality.
I got tired of wading through the morass of wild, irrational claims, so if
the answer was in there, I never saw it. Since the post was overwhelming, I
suspect that I'm not going to like the answer even if the author deigns to
reply (there's a high probability such poorly constructed propaganda was
hit-and-run spam).
BTW, for the subject line to be true ("an equal chance of winning"), we
shouldn't have any voting at all. Instead, we would simply hold a lottery in
which each candidate has one ticket.
Cheers,
-- Jeff Fisher ><> Vancouver WA
http://jeffryfisher.net/Statesman
"Individual liberty can only be built on a foundation of individual
responsibility."
IRV is fine but not close to the answer. Change the entire system to parlmentary with PR. It's time to get tough and stop diddling around with Mikey Mouse attemps.
Frederick Ellis MBAReno, NVFrom: Jeffry R. Fisher <jrf_subs...@comcast.net>
To: covote...@googlegroups.com
Sent: Sunday, July 31, 2011 8:39 PM
Subject: [COVoterChoice] Re: RB gives an equal chance of winning
preferentiality wrote:
>
> Ranked Ballot (voters ranking candidates in order of preference) will
> give us (PRACTICABLE!) Instant TRUE Democracy for ALL the World, &
> even put an end to all war forever...
Ranking candidates is only part of the process. How would "RB" tally its rankings? Instant Round Robin Voting (IRRV) uses ranked ballots, but so does IRV, and they're poles apart in quality.
I got tired of wading through the morass of wild, irrational claims, so if the answer was in there, I never saw it. Since the post was overwhelming, I suspect that I'm not going to like the answer even if the author deigns to reply (there's a high probability such poorly constructed propaganda was hit-and-run spam).
BTW, for the subject line to be true ("an equal chance of winning"), we shouldn't have any voting at all. Instead, we would simply hold a lottery in which each candidate has one ticket.
I disagree. Because it gives more weight to 1st place votes than 2nd and
lower place votes (and especially ignores last-place votes), IRV promotes
love-me or hate-me polarization.
Such polarization isn't healthy for a democracy, and for that reason alone,
IRV is awful. Add in its opacity and IRV may be even worse than the FPP that
we have now.
What we need instead is a system that rewards intelligent win-win
compromise, and the best I've seen so far is IRRV (aka Condorcet). AV
(Approval Voting) is a close second.
> frederick ellis wrote:
>> IRV is fine but not close to the answer.
>
> I disagree. Because it gives more weight to 1st place votes than 2nd
> and lower place votes (and especially ignores last-place votes), IRV
> promotes love-me or hate-me polarization.
Many would be pleased with more weight for first place - reading these
would give the same winner as FPP, assuming most voters agreed on
winner. Needing more for winner from IRV style ballots:
. AV offers equal ranking, but voters often prefer unequal.
. For IRV discarding those FPP would see as being little liked
gets to what other those voters ranked below - sometimes nice, but can
be discarding those truly best liked, which makes many argue against
IRV.
. Condorcet looks at entire ballot and sees races between each
pair of candidates. Usually one wins all such races and, if not, the
resulting cycle can get analyzed for winner among the best liked.
. This thread has its own way of looking - competitive with
Condorcet.
--
You received this message because you are subscribed to the Google Groups "Colorado Voter Choice" group.
To post to this group, send email to covote...@googlegroups.com.
To unsubscribe from this group, send email to covoterchoice+unsubscribe@googlegroups.com.
By IRV I suppose you are referring to the "eliminative" form of Ranked Ballot, where if noone has 50% the one with the lowest first choice votes is eliminated & his voters second choices distibuted to whoever won them. While this form is the natural evolution of regular runoff voting, it is true that it gives some small greater weight to the extremes, & no doubt some individuals will be inclined to be so calculationg as to not just go ahead & feel free to vote their heart.
A better form of RB is the "addtive" form, where if noone has 50& of the first choices, noone gets eliminated, but rather the second choices get ADDED in & so on until someone finally does get 50%. This will be the one most nearly exactly in the middle of all participating, ie it results in centrism, which is just wonderfully anti-extremist, but at the same time gives an equal chance of winning to not just all parties, but all combos of programs.
However, I do not get "Because it gives more weight to 1st place votes than 2nd and lower place votes (and especially ignores last-place votes), IRV promotes love-me or hate-me polarization.". I've heard such things before, but noone has ever broken it down for me, & it would be fine should Frederick Ellis be so inclined. Anything beyond going to Additive RB, such as "con-d'or-sit" is pure obstructonism & just showing off, to my mind.
Yes, that's the only one I have seen promoted as IRV.
> A better form of RB is the "addtive" form, where if noone has 50& of the
> first choices, noone gets eliminated, but rather the second choices get
> ADDED
Interesting. I'd like to see what the math and logic experts have to say
about strategic voting and paradoxical outcomes. My first impression is that
it starts to look like AV: Two or more candidates could go over the 50%
threshold after the same addition (because more than one vote is being
counted per voter). You'd then need a secondary victory criterion, and the
most obvious is to select the candidate with the highest %.
> I mean condorcet couldn't come up with an encapsulating little descriptor
> no more burdensome than Additive Ranked Ballot?
"Instant Round Robin" is descriptive, at least for those familiar with
sports tournaments.
> IRRV? FPP? Please spell them out.
In a general discussion I would have, but in a voting methods group I lazily
assumed that everyone knew the jargon or had a reference at hand.
IRRV = Instant Round Robin Voting
FPP = First Past Post or First Place Plurality
For these and many many more, with full descriptions, analyses, test cases
etc, see Electowiki http://wiki.electorama.com/wiki/Main_Page
To unsubscribe from this group, send email to covoterchoic...@googlegroups.com.
preferentiality wrote:
>
Ranked Ballot (voters ranking candidates in order of preference) will
give us (PRACTICABLE!) Instant TRUE Democracy for ALL the World, &
even put an end to all war forever...
Ranking candidates is only part of the process. How would "RB" tally its rankings? Instant Round Robin Voting (IRRV) uses ranked ballots, but so does IRV, and they're poles apart in quality.
I got tired of wading through the morass of wild, irrational claims, so if the answer was in there, I never saw it. Since the post was overwhelming, I suspect that I'm not going to like the answer even if the author deigns to reply (there's a high probability such poorly constructed propaganda was hit-and-run spam).
Cheers,
-- Jeff Fisher ><> Vancouver WA
http://jeffryfisher.net/Statesman
"Individual liberty can only be built on a foundation of individual responsibility."
--
You received this message because you are subscribed to the Google Groups "Colorado Voter Choice" group.
To post to this group, send email to covote...@googlegroups.com.
To unsubscribe from this group, send email to covoterchoice+unsubscribe@googlegroups.com.
Two major problems with this. First, unless every voter marks a preference against every candidate it will not comply with "one
person, one vote". Second, ADDING second choices to first choices means that my second preference counts against the election of my
first preference. That failure to comply with "Later no harm" will likely lead voters to "bullet vote", i.e. mark only a first
choice - which defeats the objective of the system completely.
James Gilmour
--
You received this message because you are subscribed to the Google Groups "Colorado Voter Choice" group.
To post to this group, send email to covote...@googlegroups.com.
To unsubscribe from this group, send email to covoterchoic...@googlegroups.com.
-----Original Message-----
From: covote...@googlegroups.com [mailto:covote...@googlegroups.com] On Behalf Of Zoe Zidbeck
Sent: Wednesday, August 17, 2011 6:17 PM
To: covote...@googlegroups.com
Subject: Re: [COVoterChoice] Re: RB gives an equal chance of winning
That's been a common argument against the Bucklin method. However,
first of all, it's just not true and it neglects the important issue
of preference strength. It would be a good idea, especially for those
in Colorado, where Bucklin was invented, to look at actual election
results. In real, final elections, people did vote for additional
candidates, plenty of them.
In later usage of Bucklin, for party primaries, there was a lot of
bullet voting, sure, but that doesn't necessarily mean anything.
Ultimately, lots of people, if you allow them to, will just vote for
one candidate. That's what led Dodgson (Lewis Carroll) to propose
what recently came to be called Asset Voting, to deal with this. (It
makes STV into a truly spectacular election method! But Dodgson was
mostly ignored for more than a century).
Later No Harm is a voting system criterion guaranteed to poison
results for any method that satisfies it.
The argument of OPOV is quite foolish. In the end, either one vote
counts from each voter, or none have an effect. It *looks like* the
voter is casting two votes, but in a single-winner election, the
voter is either participating in the "final pair" or not. With
approval-style voting, which Bucklin is, just staged so that there is
"limited LNH protection," if the voter votes for both candidates, the
voter has indicated approval of both and this, then, has no effect on
which one wins. Except for one thing!
And it's an important thing.
It can make the election be by a majority of voters. In Bucklin
systems, it's voters that count, not votes.
I've argued that Bucklin would make a spectacular first-ballot method
for systems that do require a true majority, such as runoff voting,
which is the most advanced voting system in common use in the U.S.A.
Bucklin is much better than IRV in finding true majority approval, as
can be seen by study of actual election results.
The arguments against Bucklin generally are old, shallow, political
arguments that were used to kill reformed voting systems in the U.S.
The people apparently loved Bucklin, it was taken down through
machinations (such as in San Francisco, where the Election Board got
a measure passed to allow an experiment with voting machines, then
used this to kill Bucklin, since the machines couldn't handle it. The
public didn't know what hit them....) or an court obviously intent on
finding what arguments it could to kill the method (Brown v.
Smallwood in Minnesota).
Bucklin worked, it worked extremely well, electing a minor-party
candidate in Grand Junction when first used (the guy was a Socialist!
Can you imagine that?). That really worried some, and they had the
power to do something about it.
The flaws of Bucklin, and there are some, almost entirely disappear
if it's used as a primary method with runoff voting. If a majority in
the primary, done. If not, runoff. I've argued that write-in votes
should be allowed in the runoff. It's a way that an electorate can
fix bad strategy in the first round. If the runoff is Bucklin as
well, then at least first-level spoilers don't damage the results.
You can still vote for your favorite, then add a backup.
By the way, there were, in the early Bucklin elections, quite a few
voters who voted for their favorite, then skipped the second rank.
And then added a vote in third rank. That was adding to LNH
protection! They knew what they were doing. It indicates stronger preference.
And this led me to thinking of using a Range ballot for a
Bucklin-like method, and to the idea of doing Condorcet analysis on
the ballots. If the result isn't clear, then there is a runoff.
Under direct democratic election process, no result is accepted
without a majority approval. That's abandoned for public elections
because it is considered more important to get a result, and
*usually* the plurality winner would win a revote. However, there are
exceptions, and they can be doozies.
This is a red herring: Even now, unless every voter votes in contest, we
fail that flawed interpretation of "one person, one vote". The truth is that
as long as the voter has the *option* to vote, the ballot is fair.
If the voter *chooses* not to exercise the option, that's his or her
business, and it does not invalidate the voting method being discussed.
> Second, ADDING second choices to first choices means that my
> second preference counts against the election of my first preference.
Yup, too bad. Your first choice didn't garner 50% in stage one, so maybe you
should compromise.
> That failure to comply with "Later no harm" will likely lead voters to
> "bullet vote"
If you don't tell us your 2nd, then maybe the rest of us will put your
*last* place choice over the top. It's your choice, but don't complain about
the outcome if you bullet vote and lose.
If there's one voting criterion I could let go, it's this fanatical
attachment to the 1st place vote. What if I like two candidates almost
equally but hate everyone else? Can I become fanatically attached to the top
two? I demand a top-two system!
Oh wait, there are already a couple states that do primaries like that.
BTW, Additive RB looks a *lot* like approval voting. Question: What happens
if two candidates both break 50% at the same stage?
--
You received this message because you are subscribed to the Google Groups "Colorado Voter Choice" group.
To post to this group, send email to covote...@googlegroups.com.
To unsubscribe from this group, send email to covoterchoice+unsubscribe@googlegroups.com.