Condorcet(Tobin) (fwd)

[Mike Ossipoff]

Mike Ossipoff writes:
> From dfb Sun Sep 29 16:33:26 1996

I've just replied to Tobin's posting, using the regular "reply" option,
which ordinarily results in a reply being posted to the list. But
it didn't. It only went to Tobin, the person whose post I was 
replying to. So I suggest that, when replying to a posting on EM,
it's best to use the "group reply" option, which automatically sends the
reply to everyone who received the message to which you're replying.

In any case, it would also be a good idea to put your own address
on the Cc: line of your reply, so that your copy of the reply will
show, in its header, who else got a copy of it. Had I not done that,
I wouldn't have found out that my message didn't post. But, as I
said, I suggest using the "group reply" option.

So my reply follows, after the usual extra header:

> Subject: Condorcet(Tobin)
> To: htobin at ccom.net
> Date: Sun, 29 Sep 96 16:33:24 PDT
> From: Mike Ossipoff <dfb at bbs.cruzio.com>
> Cc: dfb at bbs.cruzio.com
> In-Reply-To: <324E1B35.2CEC at ccom.net>; from "Hugh R. Tobin" at Sep 28, 96 11:46 pm
> X-Mailer: ELM [version 2.3 PL0]
> Message-ID:  <9609291633.aa19764 at cruzio.com>
> 
> As I understand it, Tobin's modification of Condorcet's method, or 
> Smith//Condorcet, was: If you leave A & B unranked, then you'd be counted as 
> casting half of a vote for A over B, & half of a vote for B over A.
> 
> One reason why I oppose that modification is because, with it, Condorcet
> & Smith//Condorcet would lose the properties that make them better than 
> the other pairwise methodds. Their unique & important majority rule 
> & lesser-of-2-evils properties.
> 
> For instance, as I said, if a majority indicate that they'd rather have
> A than B, then, if we choose A or B, it should be A.
> 
> The desirability of not unnecessarily violating that basic democratic 
> principle is obvious.
> 
> ***
> 
> I've objected to Condorcet(Tobin) on other grounds as well. For instance,
> I've pointed out that it counts the voter as saying something that he/she
> hasn't said. It falsifies your ranking.
> 
> The desire to give extra "bad-points" to the alternatives A & B which
> you didn't rank can be satisfied by using a point system, if that's what 
> you really want, and if you're willing to give up Condorcet's advantages 
> & accept the disadvantages of a point system. Condorcet(Tobin) is already
> modified at least halfway into a point system.
> 
> As I've pointed out before, however, the fact that you've ranked
> a whole set of other alternatives over A & B, by not raking those 2,
> and presumably a number of other voters have too, having everything else
> ranked over them by sufficient people is enough to ensure that they
> won't win. Of the alternatives that you've ranked over A & B, it's 
> only necessary that _one_ of those is ranked over them by lots of
> other people. So it isn't really necessary for the method to falsify
> your ballot & count you as ranking each over the other. Of what 
> importance is that half-vote against each of A & B, compared to the
> _whole_ vote that you've cast against them in their pairwise comparisons
> with everything else that you've ranked over them??!
> 
> So Condorcet(Tobin) would be giving up Condorcet's advantages, &
> falsifying voters' ballots, for no reason.
> 
> Aside from all this, there's something very undemocratic about counting
> you as casting a vote that you never expressed on your ballot.
> 
> The discussion on EM has been recorded in an archive, and Rob
> (robla at eskimo.com) can tell you how to gain access to it. The 
> complete discussion of Condorcet(Tobin) can be found there.
> 
> ***
> 
> But I have nothing against adding Tobin's definition of his
> modified Smith//Condorcet to Steve's list of definitions, as long
> as it's clear to the reader that it's the definition of a different
> method, a modified method different from the Smith//Condorcet &
> Condorcet that won the election.
> 
> ***
> 
> Mike
> 
> 
>  
> 
> -- 
> .-
> 


--