[quote]smh23 wrote:
[quote]countingbeans wrote:
[quote]smh23 wrote:
[quote]countingbeans wrote:
[quote]smh23 wrote:
or a forecast model like fivethirtyeight. [/quote]
Dude, I’m sorry but Silver is fast becoming a douche. Not only is the times an extension of the Obama campaign but he is just like Krugman…
Silver and Krugman could both have great careers NOT being souless shills for one party over the other (with the emergence of the tea party, ron paul bots, and even occupy to a degree, people like Krugman are scared, as his economic theories are proving to be causing massive problems in developed countries, and people are noticing.)
All I’m saying is Silver has gone out of his way to discount polls he doesn’t like, and he is in love with anything that says Obama +286,754. Although I haven’t gone back after he was the only source saying the debate didn’t change the race significantly, so he might have come back to earth since. [/quote]
First off, no argument on Krugman. He is a partisan hack.
Re; Silver, it’s interesting you feel that way, because I see him as one of the most objective guys in the field. I do assume that he’s liberal because he works at the NYT in politics and his last name isn’t Douthat. But every time I read his blog posts I’m pretty impressed at his generally dispassionate interpretations of the numbers. I can’t find the thing about Romney’s debate performance not changing the debate significantly. In fact, all I can find are headlines like “Romney maintains momentum” and etc. He IS cautious and it can at times seem like he’s discounting something, but in my view either side could see him as ignoring their best evidence.
But, that is more an argument about Silver and less an argument about his forecast model. When I say fivethirtyeight, I mean the model and not the blog. I’m not a mathematician and my understanding of these kinds of things is adequate but not by any means comprehensive, but I’ve been told by several people whose careers are built around this kind of thing that Silver’s forecast model is one of the most sophisticated out there. It does little things–downgrades Obama’s standing based upon economic data, compensates for post-convention bounces by watering down strong polls in their immediate aftermaths, evaluates national tracking polls against state-by-states–that many people believe make it preferable to a simple aggregator like realclear.
But, in the end, you can get lost in this stuff. It’s good to keep in mind that we may not be able to predict the actions of millions of separate individuals on a given day in November when we can’t even figure out whether or not it’s going to rain tomorrow.
[/quote]
Fair enough. I was hoping for a level headed responce.
I will give him a shot again, and see if I can find the peice I am reffering to.
In othr news PPP has Romney hitting 50% with the following tweet: “Impt to note on our national poll- 75% of interviews conducted within 48 hrs of debate. Would encourage Dems to wait a week before PANIC”
haha[/quote]
If you’re going to give Silver another shot, last night’s post is the place to do it. He makes some extremely good points:
http://fivethirtyeight.blogs.nytimes.com/2012/10/15/oct-15-distracted-by-polling-noise/#more-36116[/quote]
smh23,
Thanks for the heads up. I read this post and he is right on the money regarding the MoE of big N polls versus little N polls.
I expect the numbers to turn back to favoring Obama before the end of the weekend. I wonder if the polling after the first debate was less about loving Romney’s performance and more about hating Obama’s performance?
Either way, if the past few months of polling are worth anything, Obama’s numbers will hop up back to his 49 and Romney will land between 46 and 47.
jnd