[EM] Dave reply. ARLO and power-truncation.

[Michael Ossipoff]

First, I want to express my agreement with the statement that we’re here 
because public elections are important. We’re not here because of campus 
elections or organizational elections, etc. I suggest that campus 
governments and organizations be urged to use methods that are suitable for, 
and proposable for, public elections. To help get precedent, public 
experience and public exposure for better public voting systems.

Dave wrote:

While we can learn by participating in polls, we need to remember that our 
proper goal is assisting average voters in elections.

I reply:

Yes, and that’s all the more reason why we should have experience regarding 
what it’s like to use what we’re selling. Polls are the only way to get that 
experience.

Dave continued:

Perhaps we need to debate need for ranking all the candidates with 
Condorcet.

I reply:

I probably wouldn’t rank them all. In fact I’d power-truncate most of them 
if power-truncation were available. (I’ll define power-truncation later in 
this posting).

Dave continued:

There is no such need, and demanding more ranking than might be useful can 
even end up with false ranking if voters are forced to pretend to decide 
value of candidates that do not interest them positively.

I reply:

Faced with a ballot with 20 rank positions could discourage someone from 
voting. In one of our presidential polls, the nominations got sillier and 
sillier, till we had about 46 candidates. Only 7 people voted. Ranking all 
the candidates, as I did, was a bit of work. Rating them all was _a lot_ of 
work. Approval voting, however, was easy.

Dave continued:

Those voters who find they can fully express their desires in a race with 
Approval, should be allowed to express EXACTLY the same desires with 
Condorcet with the same effort.

I reply:

Quite so. And that’s why I’ve proposed power-truncation and the ARLO option. 
I’m not saying that Condorcet _needs_ them, but they’d reassure people who 
worry about strategy. And they’d please the principled voter who wants to 
show his opinion of his less-liked candidates. I’ll define them either at 
the end of this posting, or in a subsequent one later tonight or tomorrow 
morning when I next get a chance to get on the computer.

For me as a voter (as opposed to what I believe that others need) Approval 
would be fine. It’s other voters who, I feel, need Condorcet. I admit that 
that’s my subjective opinion, because other progressives disagree with me 
about what is acceptable and approvable.


I’d said:


Yes, I’d rather have SSD, but if it’s to be a handcount, or there are lots 
of candidates, and you might not want to ask people to rank that many 
candidates, then the much easier Approval offers a very good substitute for 
Condorcet’s expressiveness, Approval’s own kind of expressiveness. Approval 
would be my 2nd choice then.



Dave replies:

To me, depends on what kind of expressiveness I wish for in a particular 
race

I reply:

Well, one thing I’d enjoy about Approval would be showing my disgust for 
candidates by not voting for them. Likewise, for the same reason, I’d enjoy 
using power-truncation and ARLO, even though it’s very unlikely that they’d 
actually be needed in Condorcet.

About education: Polls in public places, such as outdoor fairs, or on the 
Internet, or use of better voting systems in campus elections and 
organizations would be good ways to show people how interesting and useful 
better voting systems are.

A few definitions:

Power truncation (PT):

If you indicate that you want to power truncate everyone below a certain 
rank position (In an EM poll, you’d do that by writing “PT” above those rank 
positions), then your ballot casts a pair-wise vote for every one of the 
other candidates (whether power-truncated or not) over every power truncated 
candidate. So, for example, say there are 20 candidates, and you 
power-truncate 15 of them. For each one of those 15 power-truncated 
candidates, your ballot casts a pair-wise vote for each of the other 19 
candidates over that candidate.

In general, no matter what the rank method, your ballot would treat each 
power truncated candidate as if you’d ranked every one of the other 
candidates over him/her.

ARLO (Automatic Rank Line Option):

ARLO uses up to 3 counts.

If you indicate that you want to use ARLO at a certain point in your ranking 
(In an EM poll you’d do that by writing ARLO at that point in the ranking), 
then your ballot truncates (power truncates, if power truncation is 
allowed--and it should be allowed) every candidate below ARLO. The 
candidates above ARLO are left in the order in which you ranked them.

If, in the 1st count, a candidate below at least one voter’s  ARLO line 
wins, then there will be a 2nd count.

If, in the 2nd count, a candidate below at least one voter’s ARLO line wins, 
there will be a 3rd count.

If a below-line candidate wins at least one of the first two counts, then 
your ballot promotes all the above-line candidates to equal 1st place, and 
they stay there.

If there’s a 2nd count, and it is won by  a below-line candidate, then your 
ballot un-truncates the below line candidates, and they’re treated as any 
ranking would treat them, according to where you have them ranked. In other 
words, they’re restored to where you’ve ranked them, instead of being 
truncated or power-truncated.

An example of a ballot with ARLO:

1. Ralph Nader
2. Peter Camejo
ARLO
3. Dennis Kucinich
4. Barak Obama
5. Hillary Clinton
6. John McCain

If you don’t want to restore the below-ARLO candidates in the event that a 
below-ARLO candidate wins, then that can be achieved by unconditionally 
power-truncating the below-ARLO candidates. Obviously then there’s no need 
to rank them:

1. Ralph Nader
2. Peter Camejo
ARLO
PT

But maybe you’d like to unconditionally power-truncate some, but not all, of 
the below ARLO candidates. That too can be easily done by how you place PT:

1. Ralph Nader
2. Peter Camejo
ARLO
3. Dennis Kucinich
PT
Of course you could rank the below-PT candidates if you wanted to, but 
there’s no reason to.

You don’t want a Republocrat, but if it appears that one is going to win, 
then you want to try to make it Kucinich.

I re-emphasize that it would be very unlikely for power truncation or ARLO 
to actually be needed in an SSD, CSSD or BeatpathWinner election. But it’s a 
way to make a statement,

Oh now look what I’ve gone and done! I’ve revealed how I’d vote in a 
Condorcet election between those candidates, if ARLO and PT are available, 
and how I’d vote in an EM presidential poll with those candidates. There 
goes the element of surprise. But you haven’t voted yet, so there’s still a 
reason to conduct an EM presidential poll.

Mike Ossipoff


----
election-methods mailing list - see http://electorama.com/em for list info