[EM] Let's clear up some confusion

[Juho Laatu]

On 5.10.2012, at 6.45, Michael Ossipoff wrote:

> On Thu, Oct 4, 2012 at 6:51 PM, Juho Laatu <juho4880 at yahoo.co.uk> wrote:
>> On 4.10.2012, at 23.53, Michael Ossipoff wrote:
>> 
>>>> I think you recommended Symmetrical ICT for informational polling. I guess you like and trust it within that framework.
>>> 
>>> I like and trust Symmetrical ICT within every framework.
>>> 
>>> In official public elections, I like and trust Symmetrical ICT.
>>> 
>>> What I don't trust, in official public elections is the people who own
>>> and operate the machines that do the machine balloting, and the
>>> computerized counting. That's the "trust" reason why I don't propose
>>> any rank-balloting method for official public elections.
>> 
>> We went through this already once.
> 
> Yes.
> 
>> My opinion was that machine balloting can be avoided if needed. Computerized counting is not a problem if the (securely recorded) ballots are public, >or if many parties can double-check the results.
> 
> As you said, we've already covered that topic.  I refer you to my
> postings in the earlier discussion. So you want 150 million ballots to
> be "public".

Yes, or alternatively available to few neutral parties. In many Condorcet methods this could also mean availability of the pairwise preference matrices of each polling station / vote recording station.

> What, you mean copies of the electronic recording are
> made public?

Yes, electronic versions.

> You have great faith in the honesty of the recording.

Not necessarily (you tell me, since I guess you assume your home country here), but I think that recording of rankings or ratings is not much more dishonest than recording bullet votes or approvals.

> ...the process between the voting and this 150 million-ballot record.
> 
> As I said before, an Approval count can be publicly watched. Not just
> the making of an allegedly-honest electronic recording of rankings,
> but the actual final approval tallies in an Approval election, with
> marking-pen on paper. When the actual result can be arrived at, via
> simple tallying, in public, in the open, in front of observers from
> the various parties, and recorded and televised by cameras belonging
> to each party, Approval is incomparably, qualitatively, more
> fraud-secure than any Condorcet method could be.

I wonder where the difference is. Simple bullet votes are easy in the sense that one could collect them into piles for each candidate and then count the number of ballots in each pile. In Approval the process could be to tick the marked candidates of each ballot on a computer screen. In Condorcet the process could be to tick the marked candidates of each ballot on a computer screen, in the given order. The results could be double-checked by another counter and continuously monitored by others, physically and electronically.

Juho


> 
> Mike Ossipoff
> ----
> Election-Methods mailing list - see http://electorama.com/em for list info