<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META content="text/html; charset=iso-8859-1" http-equiv=Content-Type>
<META name=GENERATOR content="MSHTML 8.00.6001.18812">
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT size=2 face=Arial>Jameson,</FONT></DIV>
<DIV><FONT size=2 face=Arial></FONT> </DIV>
<DIV><FONT size=2 face=Arial>You asked: "<FONT size=3 face="Times New Roman">The
part about other partisans not caring about utility seems strange to me. Why
not?"</FONT><BR></FONT></DIV>
<DIV><FONT size=2 face=Arial>I don't want to engage in a debate on this on this
list about the value of utility as a criterion, but can answer your
question...Political scientists and many other social scientists (who approach
election theory from a different angle than economists or social choice
theorists) generally dismiss utility or Bayesian regret as a meaningful election
assessment tool for several reasons. The quickest to understand is simply that
there could be a candidate who is preferred over all other candidates by a
large majority, yet another minority-favored candidate could be the utility
maximizer (average utility), simply because the minority of voters really hate
the majority winner and strongly like the other, while the majority think the
winner is better, but not great. Because the principle of Bayesian regret is in
direct conflict with the principle of majority rule, many people reject it.
Individuals can differ on the value of majority rule vs. utility, but it is a
reason often cited.</FONT></DIV>
<DIV><FONT size=2 face=Arial></FONT> </DIV>
<DIV><FONT size=2 face=Arial>There are of course many other considerations that
cause many scientists to reject utility as a meaningful measuring
system, such as the proven natural non-linear logarithmic scoring tendency
of the human brain, the unreliability of scoring compared to ranking and
either or comparisons (which is why eye doctors use Condorcet logic and ask
whether lens 1 or 2 is better repeatedly, rather than asking you to
score all the lens options), etc.</FONT></DIV>
<DIV><FONT size=2 face=Arial></FONT> </DIV>
<DIV><FONT size=2 face=Arial>Terry Bouricius</FONT></DIV>
<DIV><FONT size=2 face=Arial></FONT> </DIV>
<DIV>----- Original Message ----- </DIV>
<BLOCKQUOTE
style="BORDER-LEFT: #000000 2px solid; PADDING-LEFT: 5px; PADDING-RIGHT: 0px; MARGIN-LEFT: 5px; MARGIN-RIGHT: 0px">
<DIV
style="FONT: 10pt arial; BACKGROUND: #e4e4e4; font-color: black"><B>From:</B>
<A title=jameson.quinn@gmail.com href="mailto:jameson.quinn@gmail.com">Jameson
Quinn</A> </DIV>
<DIV style="FONT: 10pt arial"><B>To:</B> <A title=stepjak@yahoo.fr
href="mailto:stepjak@yahoo.fr">Kevin Venzke</A> </DIV>
<DIV style="FONT: 10pt arial"><B>Cc:</B> <A
title=election-methods@electorama.com
href="mailto:election-methods@electorama.com">election-methods@electorama.com</A>
</DIV>
<DIV style="FONT: 10pt arial"><B>Sent:</B> Sunday, August 30, 2009 12:30
AM</DIV>
<DIV style="FONT: 10pt arial"><B>Subject:</B> Re: [EM] Score DSV</DIV>
<DIV><BR></DIV><BR><BR>
<DIV class=gmail_quote>2009/8/29 Kevin Venzke <SPAN dir=ltr><<A
href="mailto:stepjak@yahoo.fr">stepjak@yahoo.fr</A>></SPAN><BR>
<BLOCKQUOTE
style="BORDER-LEFT: rgb(204,204,204) 1px solid; MARGIN: 0pt 0pt 0pt 0.8ex; PADDING-LEFT: 1ex"
class=gmail_quote>
<DIV class=im>Hello,<BR><BR>--- En date de : Sam 29.8.09, Jameson Quinn
<<A href="mailto:jameson.quinn@gmail.com">jameson.quinn@gmail.com</A>>
a écrit :<BR></DIV>
<DIV class=im>>> I don't see why you would guess that Score
DSV<BR>>> would have better Bayesian Regret than Range. It looks like
you tried<BR>>> to make a method that helps a voter get the best
result for himself,<BR>>> which isn't the same as<BR>>> getting
the best result overall.<BR>><BR>> I tried to make a method where
honesty was strategic. That<BR>> means allowing voters to usefully
distinguish<BR>> A>B>>C from A>>B>C or A=B>>C for
any<BR>> A, B, and C. This method does that, which removes any
need<BR>> for strategy at all in many cases, and gives defensive<BR>>
strategizers a chance to punish it in many more.<BR><BR></DIV>Yes. Making
honesty the best strategy is a common goal. But for BR it is<BR>a bad thing
with sincere votes.<BR>
<DIV class=im><BR>>> Warren defines BR in such a way that Range is
unbeatable<BR>>> given sincere votes.<BR>> Absolutely, which is why
I stated my BR challenge in terms<BR>> of rational voters where at least
half have an attainable<BR>> strategy threshold.<BR>>
<BR>>> If he measured your method, admitting strategic votes,
he<BR>>> would make<BR>>> strategy assumptions that would make
it look terrible.<BR>><BR>> Yep, which is why I (implicitly) offered
to do the<BR>> programming.<BR><BR></DIV>Warren makes his sim available.
I'm not sure if it can easily do this<BR>method, but probably.<BR>
<DIV class=im><BR>> My strategy assumption is that voters will
use<BR>> strategy iff it has an expected value greater than some<BR>>
threshold. This is a very easy bar to meet in the case of<BR>> Score
voting (approval-style strategy is a painless win) and<BR>> much harder
in the case of good Condorcet methods (where<BR>> "good method", in my
definition, means that they<BR>> reduce the cases in which strategy
works, and increase the<BR>> cases in which it backfires, to the point
where almost any<BR>> voter with less-than-perfect information has a
negative<BR>> expected value for strategy, and even under perfect<BR>>
information only a tiny fraction of voters can benefit from<BR>>
strategy). Therefore, *rational* strategic voters will be<BR>> more
strategic under Score than under a good Condorcet<BR>> method, giving the
Condorcet method a possible margin for<BR>> victory. Score DSV, because
it takes the actual utilities<BR>> into account sometimes, should have
the widest victory, if<BR>> the differences are
significant.<BR><BR></DIV>Issues:<BR>1. If you don't use Warren's
methodology and assumptions, it's not clear<BR>that your results will be
convincing to a Range crowd. (And other crowds<BR>don't care as
much.)</BLOCKQUOTE>
<DIV><BR>The part about Range partisans being wedded to Warren's assumptions I
understand, though I don't necessarily agree. The part about other partisans
not caring about utility seems stranger to me. Why not?<BR><BR>Anyway, I'm
proposing having each virtual voting group evaluate whether strategy will help
them, given different levels of true information. I think this is feasible
computationally, and I don't see how anybody in any camp could argue that
finding utility in this case is not relevant.<BR> </DIV>
<BLOCKQUOTE
style="BORDER-LEFT: rgb(204,204,204) 1px solid; MARGIN: 0pt 0pt 0pt 0.8ex; PADDING-LEFT: 1ex"
class=gmail_quote><BR>2. When Range voters vote approval-style and Condorcet
voters use<BR>reasonably sane strategies, Range/Approval is known to be
worse, as the<BR>number of viable candidates increases. So it won't be that
novel to show<BR>that your method is better than Range here.</BLOCKQUOTE>
<DIV><BR>Where are you getting this?<BR> </DIV>
<BLOCKQUOTE
style="BORDER-LEFT: rgb(204,204,204) 1px solid; MARGIN: 0pt 0pt 0pt 0.8ex; PADDING-LEFT: 1ex"
class=gmail_quote><BR>3. Given the nature of the differences between
Approval and Condorcet,<BR>it seems that Score DSV's consideration of
ratings is more likely to<BR>hurt it than help it here.</BLOCKQUOTE>
<DIV><BR>With honest votes, or considering strategy? I can't see why you'd say
this. Score DSV is more like Range than your average condorcet
system.<BR><BR> </DIV>
<BLOCKQUOTE
style="BORDER-LEFT: rgb(204,204,204) 1px solid; MARGIN: 0pt 0pt 0pt 0.8ex; PADDING-LEFT: 1ex"
class=gmail_quote><BR>
<DIV class=im><BR>> I realize this is all hot air until I actually
program<BR>> this. Yet it is at least falsifiable hot
air.<BR>><BR>>> Your wiki page seems to be lacking some<BR>>>
proofs.<BR>><BR>> As in, all of them? :)<BR>><BR>> Guilty as
charged. Which proofs would you like to see<BR>> first? I make about 25
provable/disprovable claims on the<BR>> page, that's a lot of work and it
would help if I knew<BR>> which ones y'all wanted me to start with. (I
already got<BR>> Marcus to disprove one of my claims for me by posting
here,<BR>> so my evil plot worked... thanks, Dr. Schulze
:)<BR><BR></DIV>Well, here are some comments going over the page
quickly.<BR><BR>"If there's a Condorcet winner, all voters' ideal strategy
will be to<BR>vote approval-style, and the Condorcet winner will win, thus
this method<BR>satisfies the Condorcet criterion."<BR><BR>I wrote out a
whole long thing here but eventually realized that you<BR>aren't ruling out
non-Smith candidates from winning. And that is why you<BR>are talking about
strategy above.<BR><BR>Fortunately or unfortunately depending on your
perspective, you have to<BR>evaluate Condorcet compliance based on cast
votes. If a voted CW doesn't<BR>necessarily win, then Score DSV isn't a
Condorcet method.<BR></BLOCKQUOTE>
<DIV><BR>Ouch. That passage is obviously unclear. I meant "strategy" in the
sense of "declared strategy". I was not considering any strategy at all from
the actual voters on the ballots they would input to Score DSV, but virtual
"declared" strategy on the output (imaginary) renormalized ballots, which are
intended to be equivalent to (the probabilistic average of) their strategic
Range ballots if their input ballots are honest and if they knew the true
Smith set but nothing else. In other words: if there is a condorcet winner,
the correct Range strategy for those who know that winner (and nothing else)
is to vote approval-style for that person and all better candidates, thus
Score DSV chooses the CW. It is a Condorcet method, even though it does not
satisfy the Smith criterion (if there is no CW, it could potentially elect the
condorcet loser, if that candidate had a high renormalized
utility).<BR><BR> </DIV>
<BLOCKQUOTE
style="BORDER-LEFT: rgb(204,204,204) 1px solid; MARGIN: 0pt 0pt 0pt 0.8ex; PADDING-LEFT: 1ex"
class=gmail_quote><BR>The fact that voters have a defensive counterstrategy
isn't remarkable<BR>or reassuring in itself; we would want to know what it
is and whether it<BR>is intuitive to use it. When we talk about the larger
group being a<BR>majority, I'm not sure we can design a Condorcet method
where there isn't<BR>a defensive counterstrategy.<BR><BR>It would be nice to
see reasoning as to why Score DSV would outperform<BR>Condorcet methods wrt
favorite betrayal incentive.<BR><BR>By the way, it's controversial to say
that favorite betrayal is a typical<BR>strategy in Condorcet methods.
Compared to other rank methods Condorcet<BR>is generally good at this, and
Schulze(wv) was nearly perfect when I<BR>tested it.</BLOCKQUOTE>
<DIV><BR>I was not aware of this.<BR><BR></DIV>
<BLOCKQUOTE
style="BORDER-LEFT: rgb(204,204,204) 1px solid; MARGIN: 0pt 0pt 0pt 0.8ex; PADDING-LEFT: 1ex"
class=gmail_quote><BR>I don't remember (and won't examine presently) the
precise wording of<BR>SFC (strategy-free criterion), but Score DSV doesn't
seem to satisfy<BR>the votes-only shortcut interpretation, because it can
elect B with<BR>these rankings:<BR>49 b (a and c rated zero)<BR>24
a>b<BR>27 c>a<BR><BR>The criticism is that the A>B voters can give
away victory to B, when<BR>assuming no order reversal, A might be the
"sincere CW" but B definitely<BR>is not.</BLOCKQUOTE>
<DIV><BR>This case has a CW, so Score DSV would choose that winner. There is
no condorcet cycle. You need at least 4 of the b voters to vote b>c for
your example to work. Then your example is no longer covered by the SFC, which
states: <I>"If a Condorcet candidate exists, and if a majority prefers this
candidate to another candidate, then the other candidate should not win if
that majority votes sincerely and no other voter falsifies</I> any
preferences.
<P>In a ranked method, it is nearly equivalent to say: <I>If more than half of
the voters rank </I>x<I> above </I>y<I>, and there is no candidate </I>z<I>
whom more than half of the voters rank above </I>x<I>, then </I>y<I> must not
be elected."</I> </P></DIV>
<BLOCKQUOTE
style="BORDER-LEFT: rgb(204,204,204) 1px solid; MARGIN: 0pt 0pt 0pt 0.8ex; PADDING-LEFT: 1ex"
class=gmail_quote><BR><BR>It doesn't satisfy the votes-only interpretation
of SDSC, because it<BR>can elect B with these rankings:<BR>49 b<BR>24
a<BR>27 c>a</BLOCKQUOTE>
<DIV><BR>Again, no, only if you change 28 total b and a voters to b>c and
a>b, respectively, which puts you out of the purview of SDSC.<BR><BR></DIV>
<BLOCKQUOTE
style="BORDER-LEFT: rgb(204,204,204) 1px solid; MARGIN: 0pt 0pt 0pt 0.8ex; PADDING-LEFT: 1ex"
class=gmail_quote><BR><BR>This is related to favorite
betrayal.<BR><BR>Again, it could be that it technically satisfies SDSC but
I'd have to<BR>reread it.<BR><BR>The "defensive participation" criterion I
would like clarification on.<BR>I don't see how it doesn't imply
Participation. It sounds like you are<BR>saying that if X wins, I can cast
any vote I want, and nobody I rate<BR>below X will become the
winner.</BLOCKQUOTE>
<DIV><BR>That's not what I was saying - but I've further looked at Schulze's
criticism and, more generally, the class of situations where the Smith set is
larger than 3, and I realize that what I was saying for both of these criteria
does not, in fact, hold.<BR><BR>So it's still a very good system in my
opinion, but I can now see little reason to favor it over other
approval-condorcet hybrids which allow ties (such as Llull voting), aside from
the greater expressivity of the ballot.<BR><BR>Jameson<BR></DIV></DIV><BR>
<P>
<HR>
<P></P>----<BR>Election-Methods mailing list - see http://electorama.com/em
for list info<BR></BLOCKQUOTE></BODY></HTML>