Political statistics

[asterroc]

If you like politics and math, check out this site with electoral college projections.

Example:

Comments

[jrtom.livejournal.com]

Ah, yes. The site I refer to as "election statistics porn". :)

[galbinus-caeli.livejournal.com]

So are the numbers as airbrushed as your typical "men's" magazine?

[marquiswildbill.livejournal.com]

I wouldn't be surprised. 350-187 just seems to be too wide of a spread to me.

[galbinus-caeli.livejournal.com]

The "those can't possibly be real" argument?

[marquiswildbill.livejournal.com]

Pretty much. I don't think any election has been that one sided. Other sites I've seen with similar numbers are calling states that are barely leaning left for Obama, and I would bet that a few of them are within the margin of error but are still being called. I'm just a cynic like that.

[jrtom.livejournal.com]

I don't think so. They're very clear about their methods and sources. You can argue with their models (heh) but at least there's no question about what they are. Really, I should have said "election statistical modeling porn", which would have been funnier anyway. :)

FWIW, their spread is pretty close to what a few other sites are showing. I wouldn't mind seeing a confidence interval, though.

[galbinus-caeli.livejournal.com]

Like I have said, these numbers are mostly "interesting". It would be interesting if an objective team of political scientists and mathematicians were to analyze the process.

Oh, and the most presidential elections in my lifetime have been very one sided.

2004	286 Baby Bush	252 Kerry
2000	271 Bush the Selected	266 Gore
1996	379 Clinton	159 Kemp
1992	301 Clinton	191 for Daddy Bush	0 for Perot
1988	426 for Bush the Elected	111 for Dukakis
1984	Reagan 525	Mondale 13
1980	489 Reagan	49 Carter
1976	297 Carter	240 Ford
1972	520 Nixon	17 McGovern
1968	301 Nixon	191 Humphrey	46 Wallace

[calieber.livejournal.com]

I was thinking of '84. Mondale won Minnesota and ... that's it.