[lbo-talk] combined e-mail ;)

Bill Bartlett billbartlett at dodo.com.au
Tue Apr 11 19:00:47 PDT 2006


You wrote:


> > So, once again. What exactly is your problem with that simpler
>> system? Why do you believe the more complex matrix count produces a
>> better result?
>
>Nothing you have said so far indicates you are even reading what I say.
>You simply assert, over and over, that STV is simpler.
>
>How do you know this is the case?
>
>_WHY_ do you think STV is simpler?

It isn't polite to answer a question with another question. Especially since you have already granted that it is simpler. I could understand why you might give a rhetorical answer on the list, but why waste it in a private response?


>Because it doesn't use a matrix to count the votes?

Haven't you already granted that the matrix is more complex?.


>Could it, in fact, be _more_ complicated *for* *the* *voter* than any
>Condorcet, monotonic system?

Not at all. Its isn't hard to explain the notion of sequential preferences for your single vote. The American term Instant Runoff Vote explains it in three words. Though you do need to take care to use the actual words, instead of lapsing into the acronym. ;-)


>You said, over and over and over again, that this was not a math
>problem. It is really impossible for me to talk seriously with anyone
>who wants to talk about voting systems who doesn't want to talk about
>the Social Choice branch of Game Theory of Mathematics.
>
>You sound so ignorant, I won't embarass you further by responding on the
>list.

Please, think of others who might (unlike me) have some capacity to learn from your explanations. Don't worry unduly about my feelings. If it makes you feel better, I hereby absolve you of responsibility for any psychological damage you might do. Anyhow, I know I'm ignorant and everyone else on the list knows it too.


>What is monotonic? Once you understand what monotonic is, it should be
>instantly obvious that ANY MONOTONIC RANKED BALLOT METHOD IS EASIER FOR
>THE VOTERS than STV or IRV, both of which are non-monotonic.
>
>Do you want me to explain monotonic? You never asked for an explanation.

If you want. Just give an example of the circumstances in which this flaw in STV that might lead to inappropriate outcomes.


>Schulze is:
>Simpler for the Voter,
> (monotonic)
>More Expressive of the Popular Will,
> (all information on ballot is paid attention to, not just top)
>and Mathematically far closer to the actual value of Pi than STV or IRV

My problem with this is that I believe STV is more democratic, in that it is (as the name implies) a system which involves a single vote which is transferred according to each voters express wishes. Your preference is for a mathematical concoction which ignores the preferences of individual voters and basically averages them. It treats the numerical preferences the voter marks on the ballot as scores, rather than preferences.

It is also subject to manipulation, to some degree, as are all systems. The best way of dealing with manipulations is in regulation of other aspects of conducting a ballot. Rather than the count itself. Refinements such as rotation of names on the ballot paper, outlawing how-to-vote cards, preventing political parties from any direct control of preference allocation, that sort of basic thing, can make such manipulation statistically insignificant.

But the method you are so gung-ho about has a basic flaw that can't be got around so easily. It essentially gives more than one vote to every voter. If there are five candidates on the ballot paper, it essentially gives the voters five votes, each with five different values. It is not a preferential system (IRV in your parlance).

It is not designed to be mathematically more pure, as you imply, but to favour a compromise candidate, at the expense of a candidate that is the higher preference of the majority of voters. It has the same flaws as first-past-the-post in my view, lack of preferentiality. Except that it is more likely than not to elect the bland, motherhood campaigner, at the expense of candidates who are prepared to honestly state their position. Candidates who take care not to offend even those who won't vote for them and thus avoid being placed low down the ballot by anyone.

The whole idea makes me sick. You need to defend that, you refuse to even acknowledge it, you keep wriggling away from the issue. But patronising me won't work, I've been patronised by experts mate. ;-)

Bill Bartlett Bracknell Tas



More information about the lbo-talk mailing list