<div><br></div><div dir="auto">No, but some humans are. Count-fraud has been demonstrated to have occurred, in 2000 & 2004. </div><div dir="auto"><br><div class="gmail_quote" dir="auto"><div dir="ltr" class="gmail_attr">On Mon, May 27, 2024 at 01:39 Chris Benham <<a href="mailto:cbenhamau@yahoo.com.au">cbenhamau@yahoo.com.au</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-style:solid;padding-left:1ex;border-left-color:rgb(204,204,204)"><u></u>
<div>
<p>Are your computers really necessarily so untrustworthy?<br>
<br>
I think counting Hare would be somewhat easier. Probably very weak
candidates could be eliminated in batches (while ensuring the
correct one-at-a-time elimination result). One possible kludgey
solution could be to apply the Condorcet method to the Hare last X
candidates. X could be say 7.<br>
<br>
Something I've been meaning to point out: the Approval Sorted
Margins method I promoted in the poll but the other voters here
either ignored or claimed they couldn't understand (and then
failed to give it even majority approval) has the big practical
advantage over <br>
other Condorcet methods that it normally doesn't use (and
therefore doesn't have to know) the full pairwise matrix. The
more candidates there are I should say the greater would this
advantage be.<br>
<br>
This is because (as you all recall) it only needs to check if
adjacent pairs of candidates in the Approval-score order are in
"pairwise harmony" with that order. Probably most of the time
most of them will be.<br>
<br>
Chris B.<br>
<br>
</p>
<div>On 27/05/2024 4:28 pm, Michael Ossipoff
wrote:<br>
</div>
<blockquote type="cite">
<div dir="auto">…or of course you could, for the voting, just ask
each voter to write on hir (big) ballot, which member of each of
the 18090/2 candidate-pairs s/he prefers to the other. S/he
might be in the voting/booth for a while…</div>
<div><br>
<div class="gmail_quote">
<div dir="ltr" class="gmail_attr">On Sun, May 26, 2024 at
23:51 Michael Ossipoff <<a href="mailto:email9648742@gmail.com" target="_blank">email9648742@gmail.com</a>>
wrote:<br>
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-style:solid;padding-left:1ex;border-left-color:rgb(204,204,204)">
<div dir="auto">Suppose that there are 300 million voters.
Round 18090 off to 20000x</div>
<div dir="auto"><br>
</div>
<div dir="auto">3E8 X 2E4 = 6E12</div>
<div dir="auto"><br>
</div>
<div dir="auto">6 trillion.</div>
<div dir="auto"><br>
</div>
<div dir="auto">That’s roughly the number of miles in a
light-year.</div>
<div dir="auto"><br>
</div>
<div dir="auto">Whatever the number of voters, nationally,
is, suppose that every one of them participated in the
handcount.</div>
<div dir="auto"><br>
</div>
<div dir="auto">With the work divided among them all, each
has “only” an amount of work equal to that of determining
which member of each one of 18090/2 pairs of candidates is
ranked over the other on a ballot.</div>
<div dir="auto"><br>
</div>
<div dir="auto">135 candidates in each ranking. On the
average, s/he’d only have to look at half of the ranking’s
candidates to find each of the 2 members of each pair.</div>
<div dir="auto"><br>
</div>
<div dir="auto">…but s/he’d have to do it for both. So the
order-determination for each candidate-pair, on the
average will require looking at 135 candidates.</div>
<div dir="auto"><br>
</div>
<div dir="auto">Suppose that s/he can skim over 10 of them
in a second, when doing those searches.</div>
<div dir="auto"><br>
</div>
<div dir="auto">Then it would take (135)(18090/2)/10
seconds. That’s about 1.4 days. But say she only does it
for 8 hours per day (with no breaks). Now it’s more like
4.2 days.</div>
<div dir="auto"><br>
</div>
<div dir="auto">But supervision is the whole point. If
s/he’s working alone, she can say that the pairwise
vote-totals are whatever s/he wants them to be.</div>
<div dir="auto"><br>
</div>
<div dir="auto">So in reality, it would be counting *teams*
each wit representatives of several parties. Say
(optimistically) there are 10 parties.</div>
<div dir="auto"><br>
</div>
<div dir="auto">Now it will take 42 days. But the parties
don’t really have equal numbers of members. It’s going to
take longer.</div>
<div dir="auto"><br>
</div>
<div dir="auto">…&, realistically, they aren’t going to
scan 10 ranked candidates every second for 8 hours with no
breaks.</div>
<div dir="auto"><br>
</div>
<div dir="auto">It would obviously take months. Might it
not, in fact, be measured in years…with every one of the
nation’s voters participating in that handcount?</div>
<div dir="auto"><br>
</div>
<div dir="auto"> </div>
<div><br>
<div class="gmail_quote">
<div dir="ltr" class="gmail_attr">On Sun, May 26, 2024
at 22:11 Michael Ossipoff <<a href="mailto:email9648742@gmail.com" target="_blank">email9648742@gmail.com</a>>
wrote:<br>
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-style:solid;padding-left:1ex;border-left-color:rgb(204,204,204)">
<div dir="auto">…& of course multiply that 18090
by the number of ballots, to get an idea of what’s
involved in the 18090 order-determinations to be
done on each ballot, & recorded, & then
summed, to obtain each of the 18090 pairwise
vote-totals.</div>
<div dir="auto"><br>
</div>
<div dir="auto">…each of which then must be carried or
transmitted to where the central count is done.</div>
<div><br>
<div class="gmail_quote">
<div dir="ltr" class="gmail_attr">On Sun, May 26,
2024 at 22:03 Michael Ossipoff <<a href="mailto:email9648742@gmail.com" target="_blank">email9648742@gmail.com</a>>
wrote:<br>
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-style:solid;padding-left:1ex;border-left-color:rgb(204,204,204)">
<div><br>
</div>
<div><br>
<div class="gmail_quote">
<div dir="ltr" class="gmail_attr">On Sun,
May 26, 2024 at 12:56 Chris Benham <<a href="mailto:cbenhamau@yahoo.com.au" target="_blank">cbenhamau@yahoo.com.au</a>>
wrote:<br>
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-style:solid;padding-left:1ex;border-left-color:rgb(204,204,204)" dir="auto"><br>
In 2003 there was California
gubernatorial election with 135
candidates.</blockquote>
<div dir="auto"><br>
</div>
<div dir="auto">That’s 18090 pairwise
vote-totals to determine at each precinct
from the rankings, by examining each
ranking to determine which member of each
possible candidate-pair is ranked over the
other on that ballot.</div>
<div dir="auto"><br>
</div>
<div dir="auto">…& 18080 pairwise
vote-totals for the precincts to sum,
store, & transmit or carry to the
central count location.</div>
<div dir="auto"><br>
</div>
<div dir="auto">…& verify in an audit.</div>
</div>
</div>
<div>
<div class="gmail_quote">
<div dir="auto"><br>
</div>
<div dir="auto"><br>
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-style:solid;padding-left:1ex;border-left-color:rgb(204,204,204)" dir="auto"><br>
<br>
<a href="https://en.wikipedia.org/wiki/2003_California_gubernatorial_recall_election#Results" rel="noreferrer" target="_blank">https://en.wikipedia.org/wiki/2003_California_gubernatorial_recall_election#Results</a><br>
<br>
Chris B.<br>
<br>
On 26/05/2024 9:29 pm, Kristofer
Munsterhjelm wrote:<br>
> On 2024-05-26 07:28, Michael Ossipoff
wrote:<br>
>> Someone keeps repeating that the
voters shouldn’t have to vote <br>
>> strategically. He wants the
method to do it all for us, after we <br>
>> merely state our
sincere-rankings.<br>
>><br>
>> That’s of course a common
attitude:<br>
>><br>
>> …wanting a high-tech,
computation-intensive,computer-dependent
system <br>
>> to do it all for us, taking all
the actual choosing responsibility <br>
>> off of us.…sheltering &
isolating us from the choice.<br>
><br>
> I would prefer that you do not
attribute opinions that the proponents <br>
> have not expressed. Nowhere have I
said that Condorcet "[isolates] us <br>
> from the choice" we make.<br>
><br>
> What you call "sheltering" and
"isolating", I see as the method taking <br>
> proper responsibility - proper
responsibility to turn the voters' <br>
> unambiguous honest opinions into an
outcome without dumping the <br>
> algorithmic calculation upon te voter
themselves. See my "kick the can <br>
> down the road" post for more info.<br>
><br>
> Now, I could make a caricature of
Approval itself. Perhaps something <br>
> about a calculator that just says
"IDK, do the base conversion <br>
> yourself, I only accept input numbers
in factoradic". But caricatures <br>
> only make people angry. Let's not
stoop to them, shall we?<br>
><br>
>> I’ll ask this for the 3^rd time:<br>
>><br>
>> …<br>
>><br>
>> How would like you to
handount-audit a Condorcet count for a <br>
>> many-candidate national
presidential election?<br>
><br>
> I think your question assumes
something that won't hold. If you have a <br>
> 25-candidate presidential election,
you've already lost, because <br>
> nobody is going to rank 25
candidates, irrespective of whether the <br>
> method is Condorcet, IRV, or Borda.<br>
><br>
> I'm not familiar with minor parties
in the US. Has there ever been a <br>
> 25-candidate presidential election?<br>
><br>
> I don't think I can comment beyond
that: I don't know enough about <br>
> poll workers. I'll leave that to
someone with experience. (Although <br>
> roughly calculating: suppose 6
candidates like in Burlington. That's <br>
> 30 pairs. Five times the work if
counting a particular preference is <br>
> as before. So you'd either need 5x
the workers, or five times the <br>
> time, or some combination of the
two.)<br>
><br>
> On an aside, though, I would say that
I generally wouldn't want to get <br>
> computers anywhere near election
counting. However, if you absolutely <br>
> have to have them, there are ways of
making sure they don't cheat: <br>
> formal verification. You could also
create special-purpose tools that <br>
> say, only turn ranks into matrices
and nothing else: there's no reason <br>
> (apart from programmer convenience)
why an election tool should be a <br>
> general purpose computer that you
could hide all sorts of shenanigans in.<br>
><br>
>> I’ve discussed that at length in
previous posts, & it probably isn’t <br>
>> necessary to again post about
ways of choosing how to vote in Approval.<br>
>><br>
>> But, just summarize:It’s
easy.Whichever of the various ways you <br>
>> prefer to use, for choosing whom
to approve, it’s easy.…& no, it <br>
>> doesn’t require knowing your
objectively-optimal vote.<br>
><br>
>><br>
>> I’ve many times pointed out that
Approval’s Myerson-Weber equilibrium <br>
>> is the voter-median.<br>
>><br>
>> i.e. Approval soon homes in on
where the Condorcet-Winner is.<br>
><br>
> Not necessarily. See the following
paper: <br>
> <a href="https://arxiv.org/abs/2001.04216v2" rel="noreferrer" target="_blank">https://arxiv.org/abs/2001.04216v2</a><br>
><br>
> According to the authors, under their
model, if every voter follows a <br>
> particular thresholding rule, then
iterative approval generally <br>
> arrives at the Condorcet winner (but
not always). However, if voters <br>
> are left to choose any thresholding
rule they want (as you propose), <br>
> then anything is possible. The
arrangement might even elect a <br>
> Condorcet loser, and the outcome may
be slow to converge or who wins <br>
> may depend on how many polls you
hold.<br>
><br>
> I'm reminded of a quote about
distributed algorithms that I read <br>
> somewhere: "It's really easy to
design distributed algorithms that <br>
> suffer from deadlock, network floods
or widely unpredictable and <br>
> bizarre oscillations". That's in the
context of computer science - <br>
> deadlocks might not be applicable to
election methods. But it does <br>
> justify a starting position of
skepticisim when considering schemes <br>
> that offload more of the work to the
voters by turning a one-shot <br>
> method into a dynamical system.<br>
><br>
>> It seems to me that, in every one
of EM’s polls, including the recent <br>
>> one, Approval chose the CW.<br>
>><br>
>> Have we forgotten that?<br>
><br>
> I haven't, nor have I forgotten that
Approval wasn't actually the poll <br>
> winner.<br>
><br>
> But let's take this reasoning at face
value. I don't have the other <br>
> polls' ballot data available at the
moment, so let's consider the <br>
> latest one and pick... say, Borda.<br>
><br>
> As one can see by going to <br>
> <a href="https://munsterhjelm.no/km/rbvote/calc.html" rel="noreferrer" target="_blank">https://munsterhjelm.no/km/rbvote/calc.html</a>,
pasting in the data, and <br>
> clicking Borda, Ranked Pairs is also
the Borda winner.<br>
><br>
> But I don't think I'm going to start
advocating for Borda.<br>
><br>
><br>
> To not be accused of tu quoque, let
me clarify the point. As <br>
> Burlington shows, an election method
needs to handle the hard cases, <br>
> not just the easy ones, or there may
be a considerable uproar when the <br>
> method is faced with a hard case and
then stumbles. (In addition, <br>
> people trying to make sure a stumble
doesn't happen may start to do <br>
> mass compromising, further
entrenching two-party rule.)<br>
><br>
> So it's quite possible that our polls
are easy cases. But like <br>
> FairVote claiming that IRV gets the
Condorcet winner more than 90% of <br>
> the time, that doesn't by itself tell
us much, because the failures <br>
> have such a strong impact.<br>
><br>
> I illustrate the example above by
picking a method we know to be bad <br>
> (Borda being so extremely easy to
fool with cloning and burial), and <br>
> showing that the poll result comes
out right. If a bad method can get <br>
> a good result, then "getting a good
result" is less useful than it <br>
> might appear at first glance.<br>
><br>
> -km<br>
> ----<br>
> Election-Methods mailing list - see <a href="https://electorama.com/em" rel="noreferrer" target="_blank">https://electorama.com/em</a>
for list <br>
> info<br>
</blockquote>
</div>
</div>
</blockquote>
</div>
</div>
</blockquote>
</div>
</div>
</blockquote>
</div>
</div>
</blockquote>
</div>
</blockquote></div></div>