THE MATH Weekly – For the Week Ending Sunday, August 31 – Obama 323, McCain 215 (Link)

[phrigndumass]

Now posted in the Editorials & Other Articles forum. Link:

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=103x380099

Please help keep this link thread kicked in GD: Palin for easier access. Thanks!

:hi:

[Kdillard]

1. kick and thank you

[phrigndumass]

2. .

:kick:

[phrigndumass]

3. ..

♫ Slip-Slidin' Away ♫

[phrigndumass]

4. - - - ... - - -

[azmouse]

5. K&R

[cliffordu]

6. Thanks, Phrig -

Say, what happened between you and grantcart, anyway??

Tell him I asked....


:hi:

[Sugarcoated]

7. Loooking Goood

[Coexist]

8. thanks so much

[Awsi Dooger]

9. One paragraph stood out, far above the rest

"The national daily tracking polls are a good indicator of the projected popular vote, and there is a correlation between the lead in popular vote and the probability of winning the electoral vote. A three-point lead in a national poll would correlate to a 97% probability of winning the electoral vote, according to FiveThirtyEight.com."

That's exactly right, the popular vote has overwhelming correlation to the electoral college. I know bettors here in Las Vegas who are willing to wager virtually any amount that there will not be a split in the popular vote and electoral vote outcomes.

The pundits who have been asserting Obama is under performing and the lead should be larger are clueless. A 3 point gap is hellish to overcome, during a general election in a polarized era. It's hardly equivalent to a small lead in the primaries. Likeminded individuals have far more potential to swing wildly one way or the other late in a race. When it's already lock steady at 45% minimum for either side, a 3 point bulge is a massive percentage of what is remaining and what is susceptible to collapse.

I realize you guys love state polls. IMO, that's foolish, based on 12 years of wagering on this stuff. But I can understand it because it follows suit with the typical tendencies, here and elsewhere, politics and sports and you name it, a thrill to scrutinize recency and dozens of variables, instead of evaluating from afar with one focus in mind.

My approach is comparatively boring. I know how the states break down in liberal/conservative percentage, and how they compare to the nation in that regard. At that point, the national number is everything. The only value in state polls are stupid ones, like the polling in Indiana this year which allowed me to steal Indiana in the market sites at 60, or giving only 3/2. It should have been closer to 80 or 85. That's where it's headed now.

Same thing in the primaries when idiotic state polls understated Hillary in California and Ohio, ignoring demographic realities. California traded at a joke 50/50 and Ohio at severe bargain of 60/40. God bless state polls if they continue to provide nonsense and profit like that.