<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	>
<channel>
	<title>Comments on: Proportional Oscars</title>
	<atom:link href="http://democraticpiece.com/2007/12/17/proportional-oscars/feed/" rel="self" type="application/rss+xml" />
	<link>http://democraticpiece.com/2007/12/17/proportional-oscars/</link>
	<description>Tentative conclusions on democracy &#38; governance</description>
	<pubDate>Fri, 05 Dec 2008 04:49:34 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.6.3</generator>
		<item>
		<title>By: Jack</title>
		<link>http://democraticpiece.com/2007/12/17/proportional-oscars/#comment-266</link>
		<dc:creator>Jack</dc:creator>
		<pubDate>Tue, 18 Dec 2007 20:16:40 +0000</pubDate>
		<guid isPermaLink="false">http://democraticpiece.com/2007/12/17/proportional-oscars/#comment-266</guid>
		<description>Thanks, Bob. I'm not surprised that the rules aren't more specific. Lots of non-governmental STV elections seem to rely on the external authority of tabulators when it comes to tie-breaking and surplus transfers.</description>
		<content:encoded><![CDATA[<p>Thanks, Bob. I&#8217;m not surprised that the rules aren&#8217;t more specific. Lots of non-governmental STV elections seem to rely on the external authority of tabulators when it comes to tie-breaking and surplus transfers.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Bob Richard</title>
		<link>http://democraticpiece.com/2007/12/17/proportional-oscars/#comment-265</link>
		<dc:creator>Bob Richard</dc:creator>
		<pubDate>Tue, 18 Dec 2007 19:19:33 +0000</pubDate>
		<guid isPermaLink="false">http://democraticpiece.com/2007/12/17/proportional-oscars/#comment-265</guid>
		<description>The most interesting thing about the Academy Award story is that they use plurality for the final round -- with five candidates and usually more than two viable candidates.  This makes the choice of STV for selecting the panel of finalists seem odd to me.

On the STV procedure used, I've heard two stories.  One is the Los Angeles Times article (see below). The other is from someone who claims to know someone at the accounting firm, who says they use fractional transfers.  The Academy's rules say only that "In the nominations voting, the marking and tabulation of all ballots shall be according to the preferential or weighted average system."  The way I read this, it allows the accounting firm to do all kinds of things, include "bottoms up" and various flavors of "multi-seat IRV".

In my reading of newspaper article, it describes the random transfer method in which the randomization takes place before any ballots are counted. (Usually, this is done by shuffling them and then writing serial numbers on them in the randomized order.)  Once a candidate reaches the threshold, you stop transferring more ballots to that candidate and start giving them to the next-ranked candidate on each one.

Another main variation on random transfers is the Cincinnati-Cambridge method, where you continue assigning votes to candidates who have already reached the threshold.  After all ballots have been assigned, you calculate the fraction of the winner's ballots that need to be transferred, 1/N, and then transfer every Nth ballot.</description>
		<content:encoded><![CDATA[<p>The most interesting thing about the Academy Award story is that they use plurality for the final round &#8212; with five candidates and usually more than two viable candidates.  This makes the choice of STV for selecting the panel of finalists seem odd to me.</p>
<p>On the STV procedure used, I&#8217;ve heard two stories.  One is the Los Angeles Times article (see below). The other is from someone who claims to know someone at the accounting firm, who says they use fractional transfers.  The Academy&#8217;s rules say only that &#8220;In the nominations voting, the marking and tabulation of all ballots shall be according to the preferential or weighted average system.&#8221;  The way I read this, it allows the accounting firm to do all kinds of things, include &#8220;bottoms up&#8221; and various flavors of &#8220;multi-seat IRV&#8221;.</p>
<p>In my reading of newspaper article, it describes the random transfer method in which the randomization takes place before any ballots are counted. (Usually, this is done by shuffling them and then writing serial numbers on them in the randomized order.)  Once a candidate reaches the threshold, you stop transferring more ballots to that candidate and start giving them to the next-ranked candidate on each one.</p>
<p>Another main variation on random transfers is the Cincinnati-Cambridge method, where you continue assigning votes to candidates who have already reached the threshold.  After all ballots have been assigned, you calculate the fraction of the winner&#8217;s ballots that need to be transferred, 1/N, and then transfer every Nth ballot.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
