Election Data

[CLAY S]

On Mon, Jul 28, 2008 at 12:58, Tom spot Callaway <tcallawa at redhat.com>
wrote:

> Alright. Lets start with:
> "1. Suppose the USA, by adopting range voting, lowers the risk of a
> 2-billion population crash in 50 years, by 5%. I consider this a
> conservative estimate."
>
> There is no evidence whatsoever to correlate the type of voting
> mechanism used with the risk of a "2-billion population crash in 50
> years".


Yes there is.  That multiple lines of evidence are presented that justify a
plausible derivation of that figure.


> I could just as reliably say:
>
> "1. Suppose the USA, by eating Cheezy Poofs, lowers the risk of a
> 2-billion population crash in 50 years, by 5%. I consider this a
> conservative estimate."


Only if you had evidence to back that up.  Do you?

The fact that such nonsense came out of the mouth (or keyboard) of
> someone with a PhD doesn't make it more (or to be fair, less)
> reasonable.


Do you have any evidence that it's "nonsense"? All I hear is adjectives, not
any actual factual analysis of the justifications for these figures.

If I wanted anyone to take me seriously, I would first have to show some
> kind of concrete proof that the consumption of Cheezy Poofs has any
> effect on a "2-billion population crash in 50 years", not to tell people
> that it just does (damnit!). Haven't you read anything published in the
> snack food academia in the last twenty years? It's obvious!


See points 2, 3, 4, and 5.  I would make the reasonable request that you
_read_ the piece you're criticizing before criticizing it.

I know that grandiose claims absolutely trigger the skepticism instinct, and
that these figures are highly approximate.  But the fact that similar
numbers can be obtained through several completely independent lines of
estimation is substantial.

The Bayesian regret figures are much less approximate, because they are the
combined result of millions of trials over a range of values among 5
fundamental parameters.  And they say scoring essentially _doubles_ the
effect of democracy.  If it's hard to translate that into lives saved, it's
still huge in the economic/happiness sense.  People constantly prove (via
e.g. Ron Paul and Barack Obama) that increased happiness with election
results is worth a huge amount of value to them - even more than the benefit
that could be gotten by donating that money to charities like the Red Cross
(that actually do save lives).

In either case, whether you think we're quacks or not, I don't think this
diminishes the potential value of releasing election data.  Even if it's
just a random sample, like 10% of the ballots, that's informative.  Even if
it's just a list of what percentage of the voters used intermediate values
vs. only extremes - that's valuable.

At the end of the day, there are a lot of people working hard to better the
world, and naively pursuing much less efficacious reforms.  But as long as
those reforms really are benificial and those people are trying to improve
humanity's lot, it seems nice to appreciate them rather than treat them as
lunatics on the basis of a very superficial view of their principles and
rationales.

Thanks,
Clay
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/fedora-advisory-board/attachments/20080728/8fbabdb6/attachment.htm>