On Tue, Jul 4, 2017 at 1:16 PM, Steve Coffman <***@stevecoffman.com> wrote:
(kind words snip!)
Overall I would say it’s "intelligible" and "complete" enough. I read the
material on Matchism.org a while back, so I already have a deeper
understanding of your project. I think what you’re presenting to this
competition will certainly get the overall idea across.
Thanks for the feedback!
As far as the level of “snark”…. I think you could probably leave out the
criticism of those running the competition as being a part of the problem,
and still achieve your objective.
I go back and forth on this: Is it better to just ignore the rules or
to point out how they would render any model that follows them
unworkable?
Overall, I find myself agreeing with many aspects of your project. I've been
working on a decision-making *process/structure* of my own and am surprised
to see how many parallels there are between the two. (At this point I don’t
know if I will be submitting mine to the GCF competition as I’m not sure
I'll have the time to complete it before the 9/30 deadline. And there is a
project developing closer to home that may be a more concrete and practical
application. We’ll see.)
I recommend you, and everyone else, take the time to do this: Nothing
like having to write everything down to see if you really understand
your own system. I find holes in mine every time I do this...
1. Is there a limit to the number of votes any one proxy can carry? (You may
have stated this on your website but I don’t remember seeing it, and it’s a
lot of material to reread.)
No. I believe it's essential to have this limit if people can pick
their own proxies, but as that architecture has been proven not to
work, there's really no reason to impose this limit if proxies are
automatically calculated: Better to have a few overloaded people that
really are representational of you than to match you with people who
aren't like you. The major risk here is corruption of those
individuals, but IMHO it's a low risk because it's hard to impossible
to predict where the inflection points are going to be for each
proposal (i.e., which personality/ideology "types" are going to have
the close 50/50 ratio where bribing a few of them will change the
outcome). Then again, if the proposals are well written most of them
will pass with sufficiently large majorities that it would be a
hopeless task to try to change the outcome by bribing a few (or even a
few thousand) individual voters.
2. Also, would proxies be aware of how many votes they are representing?
I don't currently show this, but it's easy enough to calculate, even
using the raw data files that are output at the close of each vote.
The problem now is that the numbers are small and so there's no point,
but it might be useful to show this in the future to encourage certain
personality types to vote more often. Many areas in the search space
are going to be pretty sparse because some personality types are just
rare, but also because there is a correlation between personality
types and participation rates (e.g., high conscientiousness
individuals (mostly conservatives) have significantly higher
participation rates than low-conscientiousness individuals). But this
problem is not unique to proxyfor.me: It's screwing up our existing
misrepresentative democracies already. The result of these unequal
participation rates is that elected officials are significantly more
conservative than the population as a whole.
3. I can see how being a proxy (particularly for a large number of people)
would be a huge responsibility that might encourage one in that position to
become more deeply familiar with the records of candidates, and the
complexity of proposals being voted on. I can also see how a large segment
of the population, through assigning their votes to a proxy, could
ultimately become disengaged from, or abdicate, their civic responsibility
by no longer paying as much attention to the issues that matter most. (Not
that the malaise of apathy and cynicism generated by our current system is
any better.) I’m curious though if you have a deeper sense of how this might
actually play out….in real life….in that regard?
All I can predict is that it'll be a *lot* less of a problem with DeMP
(Democracy by Matched Proxy) than with any other form of democracy.
The only potential problem I see that if the number of available
matches is really low for a significant portion of the population due
to low participation rates that region in the search space will become
a target for people who try to game the system by faking their profile
and then voting opposite to what people with real profiles would.
There are many ways of combating this, though, the most obvious of
which is to have the system automatically calculate where these weak
spots are and send out an engraved invitation to the people nearby to
increase their participation rates lest the bad guys end up jerking
them around. It's sort of the opposite of what you're proposing:
Rather than reward or encourage voters who are carrying a high burden,
we would try to scare non-voters into becoming direct voters by
pointing out that they're likely to be taken advantage of because, as
a group, their participation rates are too low. But note that low
participation rates *in general* are not a problem: I believe you
could run the whole world's government with a few thousand people. But
the key is that the personality/ideology search space of the active
participants needs to have the same distribution as the population as
a whole, which is something we're very, very far from in
misrepresentative democracy.
I also don't worry about the expertise/time investment issue nearly as
much as you (and probably most people working in this field)
apparently do. There are a lot better ways to deal with this than
chaining people to their desks by instilling fear of making a mistake.
The ability to redo a version-upgrade of a proposal being the most
important of these: How many changed votes do you think they'd see if
Brexit were voted on again? Having well-written proposals and
presenting the pro/con arguments right at decision time are two more.
What I really like about your Plan is how it would make the
one-size-fits-all dynamic of political parties obsolete.
Exactly. How much the zealots are going to resist this because they
don't have their "tribe" to fight for remains to be seen, but I am
encourages by the decreasing rates of party affiliation the world
over. I think most people are pretty sick of political parties and
will just override those tribalists when at some point they are given
the opportunity. Getting rid of politicians at the same time I also
figure will appeal to most people.
Regards,
Scott
All the best,
Steve