[EM] Approval equivalent to Condorcet-random :-)

Russ Paielli 6049awj02 at sneakemail.com
Mon Feb 21 04:12:12 PST 2005


A couple of weeks ago I described a small simulation I wrote of an 
idealized DSV Approval election with pre-election polls. With random 
inputs, I found that it sometimes fails to converge. I hypothesized that 
the cases of non-convergence may have something to do with a Condorcet 
cycle. My hypothesis was corroborated, with a few minor "adjustments," 
by Rob LeGrande.

I wrote:

"... it seems to me that Approval may be roughly equivalent to Condorcet 
with random selection of the winner from the Smith set. Do you agree 
with that?"

Mike Ossipoff then angrily replied by pointing out that Approval 
converges to the "sincere Smith set". That is only true under certain 
assumptions, which Mike did not state, but even if we grant that it is 
true, how does it contradict what I wrote? Mike replied that "Approval 
meets FBC & WDSC, while pairwise-count//random meets neither." Perhaps I 
should have stipulated Condorcet with sincere voting. Is that the 
problem here, and is that why Mike saw fit to cite my "confusion"?

If Approval "converges to the sincere Smith set" but randomly cycles 
through the set, then it seems to me that "Approval may be roughly
equivalent to Condorcet with random selection of the winner from the
Smith set." In no way did my statement warrant the derision that Mike 
poured on it. Indeed, Rob LeGrand replied at length without questioning 
my statement.

Perhaps the problem here is that Mike does not understand the meaning of 
convergence and random cycling because he never saw it for himself. 
Remember, Mike is completely incapable of simulating even the simplest 
model for himself. All he can do is read the results that others have 
obtained and parrot them.

I copied his reply below, because it baffles me.

By the way, I just spent a significant amount of time responding to 
Mighty Mike's bullshit. I can't help but wonder how much of the valuable 
time of others he has wasted on this forum over the years. And I can't 
help but wonder if his "contributions" (if any) outweigh the sheer waste 
he is responsible for. You're a zero, Mike. You're a worthless, pedantic 
cretin.

Here's Mighty Mike's reply:

Russ said:

If that is true, then it seems to me that Approval may be roughly
equivalent to Condorcet with random selection of the winner from the
Smith set. Do you agree with that?

Mike wrote:

First of all, the Smith set that Russ is presumably referring to is the
smallest set of candidates such that each candidate in that set
pairwise-beats each candidates outside that set. But the "Smith set" that
Approval converges to is what is called the sincere Smith set, the set of
candidates who are all publicly preferred to every candidate outside that
set--where X is publicly preferred to Y if more voters prefer X to Y than
vice-versa.

Russ´s confusion about that elementary difference shows that he doesn´t
understand the material that he copied into his own website. That
distinction was clearly made there, in the articles of mine that Russ had
there. You can easily undestand why Russ´s website was an embarrassment,
with the sloppiness and befuddlement exhibited in Russ´s Smith-set
confusion, and his sloppified rewordings.

Very obviously, the fact that Approval, when voters vote in their best
interest based on information from previous elections, converges to the
Smith set does not imply that Approval is equivalent to Condorcet (by which
Russ presumably means pairwis-count) with random selection from the Smith
set. For instance, just to cite a few differences by criteria that Russ had
copied to his own website, Approval meets FBC & WDSC, while
pairwise-count//random meets neither.

Russ desesves credit for trying very hard to sound as if he´s doing some
original investigations of voting systems. Sometimes he discovers things
that have been known for a long time, things which have been previously
explained to him. Sometimes he discovers things that are very obviously not
true.

Russ´s pretetentious posing and continual ignorant pronouncements suggest
that instead of calling someone else an intellectual midget, perhaps it
would be better if Russ would try not be be one. Unless, of course, he 
can´t
help it.





More information about the Election-Methods mailing list