<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
And if you let the voter choose how many different
<br>steps he wants you obtain a preferential ballot.
<p>If you really want all alternatives you should use a universal ballot.
<br>What is this universal ballot?
<p>...
<p>This came when I could not admit Mr. Ossipoff
<br>reasoning about winning votes being superior to margins.
<br>The case is:
<br>510: A > B
<br>490: B > A
<br>compared to
<br>200: B > C
<br>0: C > B
<br>800: B = C
<br>Winning votes would lock A>B before B>C.
<br>Margins would do the contrary.
<br>When I look at this my human analysis says that
<br>margins is the good way to order the locking procedure.
<p>But what if
<br>510: A > B
<br>490: B > A
<br>compared to
<br>19: B > C
<br>0: C > B
<br>981: B = C
<br>Would B>C be a more probabilistic ranking than A>B ?
<br>Margins locks A>B first. The question is what means the gesture
<br>of the 981 people would did not rank B and C. Did they simply truncate
<br>B and C because they were not interested by them and had not the time
<br>to look at this? Or did they have an opinon, which would be B and C
are
<br>clones, put anyone it is the same, I have no preference?
<br>So I think we need to differentiate these 2 cases:
<br>A ? B when I have no opinion and you can rely on other voters choice.
<br>A = B when my opinion is that both are equally good or bad.
<br>So if 981: B = C for real, then A > B should get locked
first.
<br>But if 981: B ? C, then B > C is more probable than A > B.
<p>You can check, when there is no "?" nor "=", winning votes,
<br>margins and relative margins produce the same result.
<br>Adding the equal (=) and undefinite (?) option for voters,
<br>relative margins leads to the most probabilistic locking
<br>procedure.
<p>Finally, another element comes from approval and is
<br>present with Demorep1's representation. Voters want
<br>to manifest who they approve and who they disapprove.
<br>This cannot be represented using only ">", "=" and "?" .
<br>Disapproving some candidates is equivalent to
<br>admit a standard replacement (because we need to have one
<br>representative) instead. Each voter can assess his own
<br>criteria for an average replacement politician.
<br>We will call him Z. We can now
<br>represent any ballot. Note we have gained some flexibility
<br>in ordering approved and disliked candidates.
<p>A FPTP ballot in favour of A:
<br>A > Z > B ? C ? D ? E
<p>An approval ballot for ABC:
<br>A = B = C > Z > D ? E
<p>A prefential ballot A1 B2 C3 D4 E5:
<br>A > B > C > D > E > Z
<p>A truncated preferential ballot , being unable to make its mind about
<br>its first choice:
<br>A ? B > C > D > Z > E
<p>We even have some freedom degrees with the use of "=" and "?" around
Z.
<p>As I said a couple of messages ago,
<br>then it is a matter of treatment of the information.
<p>----------
<br><i>Well, this is interesting.</i><i></i>
<p><i>For the record, Social Scientist types have found that a 7 bit system
is</i>
<br><i>most amenable to people's tastes.</i><i></i>
<p><i>Strong Approve</i>
<br><i>Approve</i>
<br><i>Weak Approve</i>
<br><i>Neutral</i>
<br><i>Weak Disapprove</i>
<br><i>Disapprove</i>
<br><i>Strong Disapprove</i><i></i>
<p><i>I'm not saying 7bit is better than 3bit, just that, when push comes
to</i>
<br><i>shove, there is an argument for it.</i><i></i>
<p><i>I suppose one should add a "No opinion" also.</i><i></i>
<p><i>8-bit, well, as a programmer, I like the sound of that :)</i>
<br><i></i>
<br><i></i> <i></i>
<p><i>-----Original Message-----</i>
<br><i>From: Alex Small [<A HREF="mailto:asmall@physics.ucsb.edu">mailto:asmall@physics.ucsb.edu</A>]</i>
<br><i>Sent: Saturday, April 27, 2002 8:45 PM</i>
<br><i>To: election-methods-list@eskimo.com</i>
<br><i>Subject: [EM] Forest's 3-bit approval method</i>
<br><i></i> <i></i>
<p><i>I like it!!!!!!!!</i><i></i>
<p><i>It would give greater security to voters afraid to "dilute" their
vote for</i>
<br><i>favorite, which many novices think is an advantage of IRVIt is more</i>
<br><i>conducive to majoritarian principles, which IRV promoters often
talk</i>
<br><i>about. Also, because voters would have less fear of "diluting"
their vote</i>
<br><i>for favorite it would encourage cross-over voting, which once again
helps</i>
<br><i>centrists.</i><i></i>
<p><i>I might actually like this better than Condorcet, due to its relative</i>
<br><i>simplicity. As Joe pointed out, if Condorcet were adopted
there would</i>
<br><i>still be an argument over the completion method, and although most
people</i>
<br><i>on this list are convinced that one or another method is ideal,
most of our</i>
<br><i>evaluations get technical. I don't know how well that would
go over in a</i>
<br><i>policy debate.</i><i></i>
<p><i>I would suggest a minor modification, however:</i><i></i>
<p><i>Let people indicate more than one preferred choice, so that in the
presence</i>
<br><i>of (perceived) clones voters could put both clones number 1.
Simply put</i>
<br><i>three ratings on the ballot: Preferred, Approved, Disapproved.
If more</i>
<br><i>than one person is rated "Preferred" by a majority the one with
the</i>
<br><i>most "preferred" votes wins. If nobody is "preferred" by a
majority then</i>
<br><i>do just as Forest suggested: Whoever has the fewest "disapproved"
votes</i>
<br><i>(or, equivalently, the most approved plus preferred) wins.</i><i></i>
<p><i>Finally, the increase in expense/complexity should be minimal.
In my area,</i>
<br><i>as long as the machines can read paper ballots with three circles
next to</i>
<br><i>each name there should be no need to buy new machines.</i><i></i>
<p><i>Alex</i><i></i>
<p><i>----</i>
<br><i>For more information about this list (subscribe, unsubscribe, FAQ,
etc),</i>
<br><i>please see <A HREF="http://www.eskimo.com/~robla/em">http://www.eskimo.com/~robla/em</A></i>
<br>
<br>
<br> </html>