<html>
<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">C: But above you are suggesting that
U/P somehow uses a both a 2-slot ballot and a 3-slot ballot.
Which is it?<br>
<div><br>
</div>
<div>J:3 slots. Where and how do I suggest otherwise?<br>
<br>
C: I pasted in where you wrote (in a message you said I was free
to send to EM) this:<br>
<br>
<blockquote type="cite">J: "On the 3-slot ballots, they vote
A>B. On the 2-slot ballots, they vote A. These are
perfectly consistent."</blockquote>
<br>
C: That was in response to me asking you how, in an example you
gave, how some voters were able to vote one<br>
"unacceptable" candidate above another (they voted
A>>B>C).<br>
<br>
J: In MTA, if no candidate is majority preferred and several are
majority approved/acceptable, the most approved wins.<br>
<br>
C: No, that is MCA. <br>
<br>
MTA says that if the most top-rated candidate is top-rated by a
majority then s/he wins, otherwise if more than one candidate <br>
is approved (voted above bottom) on a majority of ballots then
the one of them that is most top-rated wins, otherwise the most
<br>
approved candidate wins.<br>
<br>
And I now notice that is that last clause that makes it
different from U/P. When no candidate is majority approved it
can give<br>
a different result.<br>
<br>
I would say usually the Approval winner pairwise-beats the
Top-Ratings winner, and of course is more "broadly supported".<br>
<br>
MTA (with default rating bottom) I think is my favourite of the
methods that fail Irrelevant Ballots. <br>
<br>
<br>
Chris Benham<br>
<br>
</div>
<br>
<br>
<br>
On 9/9/2016 4:58 AM, Jameson Quinn wrote:<br>
</div>
<blockquote
cite="mid:CAO82iZwhf_fuvfS=MWe8_MazBQBjbh+36=ATdTbw1Ae2hUknXQ@mail.gmail.com"
type="cite">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div bgcolor="#FFFFFF" text="#000000">
<div><br>
<br>
<blockquote type="cite"><span>
<blockquote class="gmail_quote" style="margin:0 0
0 .8ex;border-left:1px #ccc
solid;padding-left:1ex"> C: Again, I'd be
interested in seeing a plausible example of when
U/P doesn't elect the Approval winner.<br>
<br>
Easy.<br>
20: A>>B>C<br>
35: B>A>>C<br>
45: C>>A=B<br>
<br>
Threshold in approval is >>. In U/P,
voters are as expressive as possible.<br>
</blockquote>
<br>
</span> C: On 3-slot ratings ballots, how are the 20
A supporters able to vote one unapproved candidate
above the other?</blockquote>
<br>
<blockquote type="cite">On the 3-slot ballots, they
vote A>B. On the 2-slot ballots, they vote A.
These are perfectly consistent.</blockquote>
<br>
C: But above you are suggesting that U/P somehow uses
a both a 2-slot ballot and a 3-slot ballot. Which is
it?<br>
</div>
</div>
</blockquote>
<div><br>
</div>
<div>3 slots. Where and how do I suggest otherwise?</div>
<div> </div>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div bgcolor="#FFFFFF" text="#000000">
<div> <br>
Actually it seems to me that the stripped-down 3-slot
version (if default rating is "Unacceptable") is
actually the same method<br>
as MTA. "Unacceptable" is just the inverse of
"Approved". Any candidate who doesn't get a majority
"Unacceptable" score must<br>
get a majority Approval score. <br>
<br>
</div>
</div>
</blockquote>
<div><br>
</div>
<div>Not the same. In MTA, if no candidate is majority
preferred and several are majority approved/acceptable,
the most approved wins; in U/P, the most preferred wins.
This is the only difference, aside from secondary issues
like ballot design. I believe U/P is better in this case
as it makes a chicken strategy harder to pull off
successfully; a clean cliff rather than a slippery slope.</div>
<div> </div>
</div>
</div>
</div>
<p class="" avgcert""="" color="#000000" align="left">No virus
found in this message.<br>
Checked by AVG - <a moz-do-not-send="true"
href="http://www.avg.com">www.avg.com</a><br>
Version: 2016.0.7752 / Virus Database: 4649/12972 - Release
Date: 09/08/16</p>
</blockquote>
<p><br>
</p>
</body>
</html>