heef ([info]heef) wrote,
@ 2009-06-20 15:45:00
Previous Entry  Add to memories!  Tell a Friend  Next Entry
Mathematicians rule.
On the evidence that the Iran vote was rigged:

http://www.washingtonpost.com/wp-dyn/content/article/2009/06/20/AR2009062000004.html?hpid=opinionsbox1

(4 comments) - (Post a new comment)


[info]swagmonkey
2009-06-22 01:48 am UTC (link)
Eh. Not that I disagree with the conclusion the article reaches, but I hardly found its form of evidence convincing. It's statistical -- all based in probabilities, not facts -- and the odds aren't actually all that incredibly low. Sure, one in 200 isn't terribly likely, but then again...I'm sure there have been well over 200 elections in the world in the last 5 years or so, to pick a relatively arbitrary number, and has any other fit these characteristics? Chances are, after all, 63.3% that ONE of those 200 elections would fit those characteristics. Why shouldn't this be the one?

Okay, so to stop playing the skeptic game here with the article, I do believe the election was rigged. But I find the anecdotal evidence (consistency across provinces without the wide regional variations one would expect, winning in another candidate's home province and other places where he was understood to be unpopular prior to the election's release, etc.) significantly more convincing than this game of chance with number generation. Even though the probabilities do lend some additional credence to the fraud allegations.

(Reply to this) (Thread)

[info]heef
2009-06-22 02:01 am UTC (link)
Um, so, as long as the odds of something aren't exactly zero, statistics are effectively useless? Because that's what you're saying. You may as well look at Albert Pujols striking out in a single at-bat and claim he's a crappy hitter. Statistics are useful in the way that they used them. They're not using them as proof, they're using them as evidence. They're not suggesting (as per your red-herring second paragraph) that they're anything but another piece that suggests, strongly, that this evidence was rigged - along with the other improbable variations (after all, consistency where you'd expect variation is *the definition of* a statistical argument, as is the argument pertaining about a popular candidate losing in his own province.)

(Reply to this) (Parent)(Thread)

[info]swagmonkey
2009-06-22 02:19 am UTC (link)
I don't mean to say that statistics are "effectively useless" when they aren't all the way to zero or one...but in this case I don't really think they're close enough to remove reasonable doubt. Things like a candidate winning in a region where public opinion has been measured to be significantly against him are statistical, to an extent -- the "public opinion" is measured through polls which necessarily don't reach every person, or limit themselves sometimes to those who will exercise their right to vote -- but there seems to be much more meaning behind those statistics than just random or trying-to-be-random digits. Maybe it's just my bias, but I find that sort of context more convincing than what appears to me closer to pure statistics -- likelihoods divorced from the historical and political contexts that form the other objections.

(Reply to this) (Parent)(Thread)

[info]heef
2009-06-22 02:22 am UTC (link)
I think it's your bias. This is just statistics used to show something that you're not used to them showing.

(Reply to this) (Parent)

(4 comments) - (Post a new comment)

Create an Account
Forgot your login or password?
Login w/ OpenID
English • Español • Deutsch • Русский…