[EM] IRV vs Plurality

[robert bristow-johnson]

On Jan 14, 2010, at 11:03 AM, Kathy Dopp wrote:

>> Date: Wed, 13 Jan 2010 22:24:53 -0500
>> From: Abd ul-Rahman Lomax <abd at lomaxdesign.com>
>> To: robert bristow-johnson <rbj at audioimagination.com>,  EM Methods
>>        <election-methods at lists.electorama.com>
>>
>> At 02:14 AM 1/13/2010, robert bristow-johnson wrote:
>>
>>
>>>> IRV/STV is fundamentally unfair because a large group of persons  
>>>> whose
>>>> first choice loses, never has their 2nd choice counted,
>>>
>>> only if they don't mark their 2nd choice.
>
> IRV promoters should do due diligence to understand how IRV works.

listen, Kathy, in case you haven't noticed, i've been pretty critical  
of IRV also (in case you're labeling me as an "IRV promoter").  and i  
understand exactly how IRV (as it had been enacted in Burlington)  
works.  your criticism notwithstanding.

> There are several scenarios where voters' marked 2nd choices are never
> counted, even when their first choice loses,

if their 1st choice loses at some time before the final round, their  
2nd choice is promoted to 1st choice and is not removed until after  
*it* loses.  but for it to lose, it is counted and, by the dumb IRV  
rules, is considered to come up short.

do you mean their 2nd choice is not counted because their first  
choice loses in the final round?  that goes without saying, but  
that's the dumb IRV rules.  that is an *arbitrary* threshold imposed  
upon IRV, that 1st choices count and *anything* below the 1st choice  
makes *no* difference until it gets bumped off and every other choice  
gets bumped up.

> and this is what makes
> IRV/STV such a fundamentally unfair system that tends to elect extreme
> right or left candidates while eliminating the majority favored
> candidates.

it tends to elect the candidate from the larger subgroup (in  
Burlington, Prog vs. Dem) of the larger group (left vs. right).  if  
the fringes were smaller than the center, it would elect the center  
candidate.  but, at least if you use the mayoral vote as a measure,  
in Burlington Vermont, there are more Progs than Dems.

but the elimination criterion is faulty in IRV, I KNOW THAT (next  
time you call me an "IRV proponent", i am going to remind you that  
you don't read).  considering the "big 3" (after write-in, Simpson,  
and Smith are out of the picture), IRV simple-mindedly identified  
Montroll as the biggest loser and promoted the Wright or Kiss votes  
on the Montroll 1st ballots.  there are more Dems in Burlington that  
align with Progs than GOPs.

but what would have happened if the rules were different?  what would  
have happened if the tabulation algorithm considered eliminating (for  
the time being) either Wright or Kiss?  we know what would happen.

but then, one might ask, by what reasonable measure would we  
eliminate Wright or Kiss over Montroll?  i heard one guy suggest that  
they should count *both* 1st and 2nd choices for evaluating which  
candidate is the weakest and eliminated.  but that's just another  
made-up threshold that someone pulled out of their butt.  in a sense,  
Condorcet considers *every* case of elimination, and draws an  
inference about who the winner is.  no arbitrary thresholds.

>
> Cases when voters' 2nd choices are never counted include:
>
> 1. 2nd and later choices eliminated prior to 1st choice, and the most
> important case

if their 1st choice survives and counts, how is it that Plurality  
supporters can complain?  with Plurality, the same vote survives to  
count and their 2nd-choices wouldn't have even been recorded.

> 2. the very large group of voters whose 1st choice makes it to the
> final counting round and then loses

so their 1st choice beats their 2nd choice.  why would they complain  
about that?

or, are you complaining that their 1st choice lost to someone who was  
worse than their 2nd choice?  we know about that.  sometimes it's a  
pathology and that pathology actually happened in Burlington in  
2009.  in case you didn't notice, i brought that up several times.   
the result is that the GOP Prog-haters in Burlington are gonna have  
to be considering strategic voting (compromising) in 2012, assuming  
IRV survives.

> The above will *always* happen in all IRV/STV elections. In
> particular, #1 above occurs anytime that there are a number of
> candidates that is greater by one (1) than the number of ballot
> positions.

that wasn't the case in Burlington with 5 candidates, one of which  
was completely inconsequential and did not campaign at all.

> This fundamental inequity is what causes nonmonotonicity,

Yawn.  non-monotonicity for the non-Condorcet candidate.  that guy  
should lose anyway.

> elimination
> of majority-favorite candidates, and the fact that commonly IRV/STV
> does not find majority winners because so many voters' ballots are
> exhausted prior to the final counting round, thus involuntarily
> excluding a large number of voters from participating in making the
> final decision as to who is elected.

but Kathy, the solution to that is Condorcet and you keep advocating  
the *worse* choice, plurality.  if we had Plurality in Burlington in  
2009, the Condorcet loser (of the big 3) would win the election and  
be mayor now.  to the dismay of 66% of the electorate.

> I truly cannot imagine a worse voting method than IRV/STV which fails
> more of Arrow's fairness criteria than plurality voting does.


your imagination seems to be deliberately limited.  Plurality is  
worse.  while IRV can and does fail to elect the *best* candidate  
(from the POV of the majority, the Condorcet winner), it won't elect  
the *worst* candidate (the Condorcet loser) and Plurality may very  
well do that and *would* have done that in Burlington in 2009.  the  
old law was not just simple Plurality, if the old rules were in  
effect in Burlington, there would have been a run-off, but with  
turnout reduced to 1/2 (and with GOPers apparently more motivated to  
get back to the polls on Runoff Day), it's likely that the Condorcet  
loser (of the big 3) would have won the election.

Kathy, you have strong opinions.  that's okay but you are very sure  
of yourself without a solid basis.  you also *grossly* underestimate  
people who don't walk lock-step with you.  that's just silly.

you continue to set aside the Condorcet solution to the IRV problem  
in favor of the old method (Plurality) that has *known* pathologies  
for decades or centuries.  IRV has screwed up in Burlington in 2009.   
it didn't do too bad in 2006.  but Plurality would have done *worse*  
in 2009 and no better in 2006.

labeling me as an "IRV promoter" (i doubt Terry B or Rob Richie would  
recognize me as an IRV promoter) and inferring that i haven't  
diligently studied or understood the mechanism (and the pathologies)  
only shows that you don't read.  at least you don't read what i have  
written on this forum.

--

r b-j                  rbj at audioimagination.com

"Imagination is more important than knowledge."