Re: [EM] Best winner

[Dave Ketchum]

On Jun 24, 2012, at 8:55 PM, Michael Ossipoff wrote:

To Democracy Chronicles, EM, and Dave Ketchum:
On Sun, Jun 24, 2012 at 12:19 PM, Dave Ketchum > wrote:


Quoting from today's Demoncracy Chronicles, 6/24/12:
The basic idea is avoid the situation faced today, where many  
candidates that are well liked do not get votes because voters  
choose the most likely to win candidate instead of their  
favorite.  Source: Democracy Chronicles (http://s.tt/1fy4W)




Dave comments:

Reads like a typo - that these voters would vote for the one they  
think is most likely to be voted for by other voters.


I reply:

The voters are certain that the winner will be either the Republican  
or the Democrat, and so they (nearly) all vote for the Republican or  
the Democrat. And so guess what?...The winner is therefore  
predictably always the Republican or the Democrat.


Assuming X is reported as likely to win, these voters would help this  
along by also voting for X, rather than voting for X or Y according to  
which they would prefer to have win.


Agreed that if X and Y are Rep and Dem, considering only among them as  
major candidates makes sense - but voting for the one reported as  
ahead fails as to being useful.


But thanks for your suggested wording-change, Dave.

Dave continues:

for Approval voters should:
. Start with their favorite.
. Add the best they see among possible winners - but not if this  
best likely will cause their favorite to lose.


[endquote]

Sounds about right. I like and agree with Dave's emphasis on  
avoiding helping an unliked compromise. You won't find any unliked  
compromises marked on my approval ballot. In Approval, one never  
approves an unacceptable candidate.


But I also refer Dave to the strategy suggestions in my Approval  
article at Democracy Chronicles, for voters who want to use strategy.


But my best suggestion for voting in Approval is: Just approve  
(only) the candidates whom you like, trusts, &/or consider deserving  
of your support.


If all you know is that you see X and Y as each deserving, you  
properly vote for both.


However, changing that to preferring X, and X and Y each being  
possible deserving winners, you need to consider:

 If your vote will likely not affect which one wins, vote for both.
 If voting for Y could cause Y to win over X, you think on this  
as part of deciding whether to also vote for Y.


It gets sticky.  If considering only Y, then whether Y may be  
deserving is all you need as to voting.  Add to this X being  
deserving, and you need to consider possibility of voting for both  
causing X to lose.


Mike Ossipoff


DWK
Election-Methods mailing list - see http://electorama.com/em for list info

Re: [EM] IIAC. Juho: Census re-districting instead of PR for allocating seats to districts.

[Michael Ossipoff]

There was a typo in my posting that I've just posted.
>
>
I'd said:

That means that it's putting all of the districts' seat-allocation as close
as possible to the same number.

And that means that it's putting all of the districts' seat allocations
quite close to eachother

[endquote]

I meant:

That means that it's putting all of the districts' representation per
person as close as possible to the same number.

And that means that it's putting the districts' representation per person
quite close to eachother.

No, it doesn't necessarily quite minimize the greatest difference in
representation per person. But it keeps it quite small, something that LR
doesn't even try to do, for the reason given above....as exemplified by
the transfer property that I described.

Mike Ossipoff

Election-Methods mailing list - see http://electorama.com/em for list info

Re: [EM] IIAC. Juho: Census re-districting instead of PR for allocating seats to districts.

[Michael Ossipoff]

On Sun, Jun 24, 2012 at 4:32 AM, Juho Laatu  wrote:

> On 24.6.2012, at 9.36, Michael Ossipoff wrote:
>
> > But remember that, to get one thing, you give up something else. What
> > are you giving up to get LR's optimization?
>
> You already know. Some well known paradoxes + non-monotonicity with
> respect number of seats.


[endquote]

No. Worse than that. LR makes random, often great, increases in the
greatest difference in the districts' representation per person.

You said:


> These may be problematic or may be what you want, depending on your needs.
>
> [endquote]
>
>
>


 It isn't problematic if the people in the district with drastically,
unnecessarily, less representation per person don't mind that :-)

>
> > So what are you getting?
>
> You already know this one too (maybe your questions are just rhetorical).
> You get an optimal method if your target is to minimize deviation from
> ideal proportionality in number of people.
>

[endquote]

I've already acknowledged that optimization that LR has. As I said before,
it's a a global property, aesthetically approving to us who look at it (but
who don't live in the drastically under-represented districts).


>
> >> Divisor methods focus on ratios of people and representatives. Why
> should that be the only approach that people should use?
> >
> > Because equal representation for all people is the goal.
>
> Do you have an exact formulation on what you think is the crucial property
> that makes SL optimal or best in "equal representation" that all should
> follow (at least when compared to LR)? You focus very much on optimization
> of seats per quota, although you also agree that not even SL does perfect
> job here.


So what if SL isn't optimal for equality of S/Q? It does dramatically
better than LR in that regard, and that's sufficient for this discussion.

You ask for an exact formulation of the crucial property that makes SL [not
optimal, but better than LR] in equal representation for people.

SL finds the allocation that results from dividing each district's
population by the same number, and then rounding off each district's
division result.

(that common divisor is chosen so as to result in the desired total number
of seats).

Thereby, SL puts each district's allocation as close as possible to its
ideal proportional share of the desired seat total.

That means that it's putting all of the districts' seat-allocation as close
as possible to the same number.

And that means that it's putting all of the districts' seat allocations
quite close to eachother.

That accomplishment shows in SL's transfer property:

Starting with an SL seat allocation, if you take a seat from one district,
and give it to another district, that will always put their S/Q farther
part than it initially was.

You said:


> I guess we have already agreed to disagree on the optimality of SL


[endquote]

 No we don't. I said that SL isn't optimal. But SL is  nevertheless better
than LR, when it comes to respecting people's right to equal
representation. Why? See above.


You said:

, but if you have an exact definition on what "SL's optimal
proportionality" is, that could provide some more material for discussion.

[endquote]

SL isn't optimal. See above, for why SL is better than LR in regards to
respecting people's right to equal representation.

Why use SL if it isn't optimal? Because it's simple, because it has much
precedent, and because it does very well a giving people equal
representation (see above).


> Yes. Every Hare quota in my district should have as much
> representation as do the Hare quotas in your district. But look at
> what you're doing: Again, you're fragmenting the situation. ...the
> Hare quotas this time. Looking at a particular piece of a Hare quota
> and saying "This fraction of a Hare quota has no representation."
You replied:

No. What I said was that a _full_ Hare quota of voters has no
> representation in the first district/party (1.05 quotas to be exact).



No? Yes. What you said was exactly the statement that i was referring to.
In your above-quoted sentence, you claim that a full Hare quota of voters
has no representation in district A (in my example). That statement is
incorrect. In district A, as in all 3 districts, no one is without
representation. You see, your above-quoted sentence again states that
fiction that I referred to. The fiction that there are voters with no
representation. There aren't. Everyone in district A has representation.
Each person's representation in district A equals the number of seats
possessed by district A, divided by the population of district A.

And that quantity, the representation per person, is the quantity that SL
does a very good job of making as equal as possible. As I said, SL does a
very good job of granting people's right to equal representation.  ...for
the reasons that I described above.

Mike Ossipoff

You asked:

Do you think that's how allocation should be done?
[endquote]

Yes.

Mike Ossipoff

Re: [EM] Best winner

[Michael Ossipoff]

To Democracy Chronicles, EM, and Dave Ketchum:


On Sun, Jun 24, 2012 at 12:19 PM, Dave Ketchum wrote:

>
> Quoting from today's Demoncracy Chronicles, 6/24/12:
>
> The basic idea is avoid the situation faced today, where many candidates
> that are well liked do not get votes because voters choose the most likely
> to win candidate instead of their favorite.  Source: Democracy
> Chronicles  (http://s.tt/1fy4W)
>
>
>
> Dave comments:


>
> Reads like a typo - that these voters would vote for the one they think is
> most likely to be voted for by other voters.
>
>
I reply:

The voters are certain that the winner will be either the Republican or the
Democrat, and so they (nearly) all vote for the Republican or the Democrat.
And so guess what?...The winner is therefore predictably always the
Republican or the Democrat.

But thanks for your suggested wording-change, Dave.

>
> Dave continues:
>


> for Approval voters should:
> . Start with their favorite.
> . Add the best they see among possible winners - but not if this best
> likely will cause their favorite to lose.
>
>
[endquote]

Sounds about right. I like and agree with Dave's emphasis on avoiding
helping an unliked compromise. You won't find any unliked compromises
marked on my approval ballot. In Approval, one never approves an
unacceptable candidate.

But I also refer Dave to the strategy suggestions in my Approval article at
Democracy Chronicles, for voters who want to use strategy.

But my best suggestion for voting in Approval is: Just approve (only) the
candidates whom you like, trusts, &/or consider deserving of your support.


> Mike Ossipoff
>
>

Election-Methods mailing list - see http://electorama.com/em for list info

[EM] Best winner

[Dave Ketchum]

Quoting from today's Demoncracy Chronicles, 6/24/12:
The basic idea is avoid the situation faced today, where many  
candidates that are well liked do not get votes because voters  
choose the most likely to win candidate instead of their  
favorite.  Source: Democracy Chronicles (http://s.tt/1fy4W)


Reads like a typo - that these voters would vote for the one they  
think is most likely to be voted for by other voters.


for Approval voters should:
. Start with their favorite.
. Add the best they see among possible winners - but not if this  
best likely will cause their favorite to lose.
Election-Methods mailing list - see http://electorama.com/em for list info

Re: [EM] IIAC. Juho: Census re-districting instead of PR for allocating seats to districts.

[Juho Laatu]

On 24.6.2012, at 9.36, Michael Ossipoff wrote:

> But remember that, to get one thing, you give up something else. What
> are you giving up to get LR's optimization?

You already know. Some well known paradoxes + non-monotonicity with respect 
number of seats. These may be problematic or may be what you want, depending on 
your needs.

> So what are you getting?

You already know this one too (maybe your questions are just rhetorical). You 
get an optimal method if your target is to minimize deviation from ideal 
proportionality in number of people.

>> Divisor methods focus on ratios of people and representatives. Why should 
>> that be the only approach that people should use?
> 
> Because equal representation for all people is the goal.

Do you have an exact formulation on what you think is the crucial property that 
makes SL optimal or best in "equal representation" that all should follow (at 
least when compared to LR)? You focus very much on optimization of seats per 
quota, although you also agree that not even SL does perfect job here. I guess 
we have already agreed to disagree on the optimality of SL, but if you have an 
exact definition on what "SL's optimal proportionality" is, that could provide 
some more material for discussion.

>>> Surely no one would deny that the number of representatives that a Hare 
>>> quota of people has is its "representation".
>> 
>> I note that although you wrote these words to support Saint-Laguë, they work 
>> also against it. Let's say we have proportions 61-13-13-13. SL allocates the 
>> seats 2-1-1-1. The number of quotas of each district/party has is 3.05 - 
>> 0.65 - 0.65 - 0.65. The third full quota of the largest district/party does 
>> not get its seat. Shouldn't all quotas get their representation?
> 
> Yes. Every Hare quota in my district should have as much
> representation as do the Hare quotas in your district. But look at
> what you're doing: Again, you're fragmenting the situation. ...the
> Hare quotas this time. Looking at a particular piece of a Hare quota
> and saying "This fraction of a Hare quota has no representation."

No. What I said was that a _full_ Hare quota of voters has no representation in 
the first district/party (1.05 quotas to be exact). Do you think that's how 
allocation should be done?

Juho





Election-Methods mailing list - see http://electorama.com/em for list info

[EM] Approval presidential poll at Democracy Chronicles

[Michael Ossipoff]

Democracy Chronicles, a publication with international readership,  is
conducting a 2012 U.S. presidential poll by Approval voting.

To vote in that Approval presidential election, go to
http://www.democracychronicles.com

The candidates are:

Barak Obama
Mitt Romney
Roseanne Barr (Green)
Dr. Jill Stein (Green)
Johnson (Libertarian)
A Constitution Party candidate whose name I don't remember

So you're invited to vote in the Approval poll by going to the website
specified above.

Mike Ossipoff

Election-Methods mailing list - see http://electorama.com/em for list info