h a l f b a k e r yNormal isn't your first language, is it?
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Please log in.
Before you can vote, you need to register.
Please log in or create an account.
|
Gaussian Party
Simple, basic, and backed by 200 years of scientific and mathematical enquiry | |
The Gaussian party polls the entire population quietly in the
background, and establishes the
median values, along with the standard deviation of various views
on different political topics.
It then campaigns for the median, whatever that may be, using
facts and information to back
up the
arguments.
All this would be enshrined in good scientific practice, polls and
data collection would be
performed under strict best-practices, and methods published as
part of the manifesto.
Members of the party would remain anonymous, and funding of
the party made only through
transparent donation, on the basis of zero bias intended or
requested.
In cases where public perceptions have been skewed by fast-
talking fringe career politicians
funded by self-serving media organisations, I'm not entirely sure
what should happen, but it
ought to be a darn sight better than the status quo. Perhaps if the
Gaussian party offered some
kind of media service, whose output was bound by contract to be
fact-based,
unbiased, and truthful, that might stand a chance.
This way, normal, sensible centrist ideas would naturally hold
sway over short-term, populist
ones, and we'll all sleep easy in the knowledge that our children's
futures wont be dictated by
the growing trend of hate and fear politics of today.
Wikipedia: Multivariate Normal Distribution
https://en.wikipedi...normal_distribution
Looks a bit like a pimple [zen_tom, Jun 13 2016]
Wolfram Alpha: Binom(8,3/8)
https://www.wolfram...%7B%22Formula%22%7D
Binomial Distribution of Users with Numbers referenced in their user names, based on the sample collected in this idea [zen_tom, Jun 15 2016]
Exploring Virtual Reality Data Visualization with Gear VR
https://civisanalyt...n-with-the-gear-vr/
Here you go [Pertinax]! [zen_tom, Jun 15 2016]
[link]
|
| |
Far, far too sensible. Congratulations on a truly half-baked idea. |
|
| |
// the growing trend of hate and fear politics of today. // |
|
| |
Actually, no, it's not growing at all. It's just exactly the same as it's always been. Doubtless, uncooperative Athenian children were told "Eat your vegetables or the Spartans/Persians/Trojans will come and get you". |
|
| |
Spartans/Persian/Trojans didn't have nukes, energy weapons, and tailored diseases at their disposal. |
|
| |
The selected bogeyman only needs to have equivalent capability, and be recognisably "other". Napoleon didn't have chemical weapons, the Carthaginians did have elephants but they weren't very practical... the threat isn't illusory, but doesn't have to be overwhelming, just credible. |
|
| |
//Perhaps if the Gaussian party offered some kind of media
service, whose output was bound by contract to be fact-
based, unbiased, and truthful// |
|
| |
I can already hear the lawyers salivating. It's not pleasant. |
|
| |
Could there also be an opposition de-Gaussian Party which reduced
the effects of the fields of expertise in order to destroy accurate
information about the issues? |
|
| |
Will the Gaussian Party have splinter groups who will
campaign for policies based on the mode and mean of
public opinion? |
|
| |
That could simply be one person carrying an extremely large induction coil ... |
|
| |
[8th] you're right in that little has changed since classical times - one key difference we have today is statistical technique. |
|
| |
[2fries] qualitatively, having your city is sacked by a horde of angry Spartans, or nuked by a bomb are different, though it is safe to say that both can be considered "bad days". |
|
| |
[8th] I think the idea here is to try and do away with Bogeymen. That might be a stretch too far for some. |
|
| |
[pertinax] if we enshrine scientific technique, we can all hide behind transparency and slavish devotion to method to preserve us from the sound of lawyer's hand-driers spinning into operation. |
|
| |
[nineteenthly] might be worth keeping a small de-Gaussian contingent, if only to allow future generations to experience that <click>-<boinggzzzzz> that is sadly lost to those who've never used a CRT monitor. Less sensationally, a de-Gaussian party is how I see most politics today - irrespective of position on the Political complex-plane, most arguments are made emotionally, without the slightest requirement to reconcile against fact. So yes, that could continue, and I'd prefer not to wipe anything out, just to offer an alternative based on providing a verifiably unbiased set of policies built only upon facts. |
|
| |
[hippo] Only in cases where the distributions are skewed, under technically "normal" distributions, we'd naturally expect the median, mode and mean positions to be exactly the same. |
|
| |
// slavish devotion to method // |
|
| |
Much as I respect slavish devotion to method, the method
in this case seems to require a universe of political topics
each of which is one-dimensional; unless, that is, you can
have a Gaussian distribution over multiple axes. |
|
| |
Also (and I think this problem has come up in relation to
other mathematical-politics ideas here), it assumes that
these unidimensional political topics are clearly distinct
from and in some sense orthogonal to one another. |
|
| |
So... on day one, we discover that the median preference
for the overall level of public spending is vastly exceeded
by the sum of the median preferences for public spending
in each individual field. At this point we throw up our
hands and dither, at the mercy of some awful populist of
the Rotterjack party, who is at least able to make a
decision, even if it proves to be a bad one. |
|
| |
// you can have a Gaussian distribution over multiple axes. // |
|
| |
At least two; rotation about the x axis would describe a discus-like structure. |
|
| |
There doesn't seem to be anything in topological theory that would preclude a multidimensional extrapolation of a Gaussian curve; the probem might come with representing it visually on a planar surface, like a tesseract. |
|
| |
//the median preference for the overall level of public spending// |
|
| |
Aha, on day-one the data-set assumes all the preferences established to date using the earlier, if flawed, democratic model - so if there is an existing value, that is assumed correct. |
|
| |
All we need then to do is define the median preferences for increasing or decreasing the level of public spending. |
|
| |
It's a bit pedestrian a start, but it is at least verifiable. |
|
| |
Later, we can conduct preferential analysis amongst different spending priorities - both surfacing any human paradoxes such as wanting to reducing gun-crime, and at the same time wanting more guns, or having a national health-service, actually paying for it. Any such contradictory policy decisions can be piped into a log of "things we need to empirically study" so we can generate the data to help resolve the conflicting interests. |
|
| |
Agreed that the multi-dimensionality becomes difficult, but given a suitable model, we ought to be able to simulate (within suitable confidence intervals) outcomes based on possible choices - and, if some principle components become evident such that we can project them from a higher-dimensionality onto a lower one, that can be used to help present the conclusions to people who don't want to get too deep into the details. So if for example, the ability to get a job were down to 30% 1/bureaucracy, 30% economic stability, and a bunch of other factors to complex or multifaceted to call on, we can still point to 60% being explainable by those two factors, and act accordingly - or at least, publish the results for people to have a think about. |
|
| |
// the multi-dimensionality becomes difficult, // |
|
| |
The solution to that is to ensure that the overwhelming majority of voters are provided with 3D TVs, shirley ? |
|
| |
There are serious discussions afoot for using 3d-virtual reality headsets for data-analysts for just this purpose. |
|
| |
<sigh> Yes, we know … please, do try to keep up … |
|
| |
I love the spirit of this idea, but I have a few issues. |
|
| |
1) If you're going to say it is scientifically based on public
opinion then don't attempt to influence public opinion. If
public perception has been influenced by fringe elements,
so be it. Take that into account. |
|
| |
2) I don't disagree that there should be some good
unbiased news sources, but that must be completely
independent from this political party. Otherwise it
looses all credibility. |
|
| |
3) I like the idea of using the median, not the mean.
Currently is seems that people support the most extreme
view in the hopes of swaying the system slightly in one
direction. If this existed and people understood that
they could influence things just as much with an opinion
just slightly left or right of center as with an extreme
view, then the debates would be much more civil and
productive. |
|
| |
4) I think for this to be really effective it needs to get
away from the one-dimensional view of most issues. As
seen in politics today, the two extremes on each topic
are driven from opposite ends of the political spectrum,
leading many to believe that it's a straight line between
them. In reality there is almost always at least some
small degree of orthogonality to the logic behind most
opposing views. For example, those favoring legalizing
marijuana, the primary reason seems to be the money
wasted and increased crime due to the "war on drugs".
The opposition primarily wants to reduce drug use. Most
of those that want to reduce drug use don't actually want
jails full of people. Most of those that support
legalization don't have a dream of a world with 3/4 of
the people walking around stoned. The tricky part is to
get survey questions that identify the public opinion on
the root issues that are slightly orthogonal and propose
policies to take advantage of that. That can result in
compromises where more than half of the people feel
they got more than half of what they want. The problem
is that it requires smart people who can look objectively
at both sides of an issue. |
|
| |
So, if you want to implement this (at least in the USA),
NOW is the time to jump. You've got a large number of
democrats and republicans dissatisfied with their current
nominee. If you could find someone qualified yet not
seen as a lifetime politician who is highly respected and
holds moderate views on most issues, this is probably the
one election where a 3rd party candidate might have a
fighting chance. This candidate would only need to
secure 1/3rd of the Democrats, 1/3rd of the republicans,
and would be pushed to victory by independents. |
|
| |
Is there any mileage in researching into why so many
annotators to this idea have numbers in our names? How
well-supported would the evidence be for that? |
|
| |
Not sure, based on the sample to-date, I make it 8 annotators, 3 of whom reference numeric in their user names. We might assume this to be spread across a binomial distribution ~Binom(8,3/8). I've attached a link that describes the function but I'm not entirely sure how helpful that is. You could say that on an idea annotated by 8 different users, the probability of none of those users having referenced a numeric in their username is around 0.025, but again, the utility of that particular statement is questionable. |
|
| |
It's not useful because of the small sample size. You would need to include all the annotations to all the HB ideas, or at least a significant proportion thereof. The starting point could be to determine the average number of annotations per idea, then identify a representative sample within one standard deviation. |
|
| |
// leaving no evil to be found now // |
|
| |
//the data-set assumes all the preferences established to
date// |
|
| |
Nice try. Are you assuming that the current spending
level is correct, or that the current deficit level is
correct, or that the current debt level is correct, and is it
correct adjusted for inflation, adjusted for GDP growth or
not adjusted at all? If the status quo is correct for any
one of these, it is likely wrong for most of the others. |
|
| |
Regarding the general principle, I think Sartre would like
a word. So would Kierkegaard, but he doesn't have a
polo-neck. |
|
| |
Oh, and [8th of 7], look behind you. |
|
| |
Oh, and meanwhile, can either of you provide links to those
// serious discussions afoot for using 3d-virtual reality
headsets for data-analysts // ? |
|
| |
Eeeek ! Evil ! Who let that cat in ? |
|
| |
// If the status quo is correct for any one of these, it is likely
wrong for most of the others // |
|
| |
In this case, as with all other problems in economics, it's just a
matter of trimming the data until it fits the theory. |
|
| |
// it's just a matter of trimming the data until it fits the theory // |
|
| |
Thank you for the link [zen_tom]. I have some quarter-
baked thoughts of my own in that direction ... |
|
| |
I thought, momentarily, that this might be nifty way to locate refreshments at a social gathering party to distributionally distribute the guests, as the tropism for hanging out near the food is well known |
|
| |
unhorrifyingly, the idea of motor voter, where every state ID is also voters registration has similarities except for the showing up to vote part (rather than being measured) |
|
| |
4d gaussian a disk, or s-orbital are 2 or 3d gaussians, a 4d gaussian at a social party would just be suggesting the party starts at different times, so everyone can be fashionably late, yet with a nice distributional peak, then they could adjust the music to gradually get people on their way. |
|
| |
actually the actual idea though, is not that like regression to the mean, kind of well known for being um, unimpressive, and responsive to television advertising. so er, -1 |
|
| |