Thread: Election Day!
View Single Post
Old 11-06-2012, 12:53 PM   #101
ColoradoDarin
Not Too Shabby Poster
 
ColoradoDarin's Avatar
 

Join Date: Aug 2005
Location: Wilson, NC
Posts: 7,851

Adopt-a-Bronco:
T J Ward
Default

Quote:
Originally Posted by Hulamau View Post
Not true about skewed momentum factors in the better analysis of aggregate poll averages. Those more complete and reliable models account for likely momentum as much as it's possible.

And a very critical factor in this case is that the only sign of momentum direction that gives us any real indication at this point beyond wishful thinking or fantasy, is the fact that in a dramatic majority of polls over the last week, and increasingly so with each passing day up through this morning, there has been a consistent and steady, even if small and incremental, shift in obvious momentum toward Obama.

That is just an objective fact based on the published polls at this point.

Romney's ONLY snowball's chance in hell to win at this point is if there has been a collective statistical bias in nearly all the polls, both individually and when taken in aggregate which, while statistically possible, is 'EXCEEDINGLY' unlikely to happen ...

The polls, of course are not infallible, however they have been, by in large, far more accurate since the mid to late 90's than from the eighties and earlier due to vastly improved understanding and technical improvements in how to model and structure polls.

If the majority of polls over the last week had started to break toward Romney, then your wishful thinking might well have a lot more merit.

But in fact, its been the exact opposite this past week with Obama gaining in momentum almost across the board in swing states as well as moving up in the national polls as well. That is a very hard reality to overcome for anyone trying to find some solace or comfort in the projection of a Romney win at this point.

In essence, Romney's only hope is now down to a Hail Mary TD from Romney's own 20 yard line with 1 second on the clock with Von, Doom and Wolfe crashing the line and with Champ, Darrell Revis and Cromartie locking down the republican receivers. It has happened, it can happen, but I sure wouldn't bet one red cent on it happening tonight at this point!

Regardless of whom nerdy Nate Silver is pulling for personally, I can find no inherent bias or flaw in his methodology or mathematics. Read his book 'The Signal and the Noise', as I did, and you are bound to be impressed at the level of detail and cross-checking he adopts to argue against what his own models might seem to be suggesting at each step of the way. Silver does this in order to continually refine those models and to give more accurate and unbiased prognostications .. his rep depends on it.

Silver isn't running any of these polls himself, his analysis is looking at what the combination of all the polls are saying in total which is a dramatically more powerful and reliable view on what is likely to happen than depending on any one or a handful of individual polls regardless if their 'house effect' leans toward Obama or Romney.

And it is certainly more solid than any amount of wishful spin from either the right or left about 'their guy'.

Those polls with a known consistent bias for either side are adjusted for in his models, such that PPP, Pew, Democracy Corp. and a couple others that tend to skew a few points toward Obama/Democrats are weighted as such, just as Rasmussen, Gallup, Gravis, Zogby and a couple of others that consistently overrate Romney/Republican chances are adjusted and weighted accordingly.

Thus, if Romney wins at this point it will truly be a large shock and upset even though the race is technically close... and I very much doubt that's going to happen. Kind of like the odds of the Broncos coming from behind to win at SD on the road in the second half with a 24-0 half time deficit!

That kind of win happens, obviously, but only once in a very blue moon. Consider that awesome Bronco win was tied, or something like that, for the largest comeback in a 665 game Monday night football history ... and also was tied in all of Bronco history!? That gives you a rough idea of what the approximate odds are. Romney's might be slightly better at this point but not much and the analogy is a good one.

In Silver's last model posted today, prior to us all getting the real results, he ran the total of all polls through his model for 100 repeat elections .. Obama won 91 times and Romney 9 times. That is not no chance for Romney, but I really wouldn't want to be in his shoes right now.

Put it this way, if Obama wins Virginia whose polls close at 7pm EST we can all go to bed early as the fat lady will have already sung.

Its sure to be an interesting evening regardless.
We'll take Ohio, because really, whomever wins Ohio is President (though Romney could win WI and NH, lose OH and still win, but that scenario isn't as likely as just winning OH if you look at historical EC/state trends).

The problem I see with the OH polls, especially the final polls is that they are claiming that 35-47% of people have already voted, when the number is closer to 31-32% (32% using 2004 turnout, 31% using 2008 turnout). And those early voters in the polls skew more Democrat, but then the OH Sec of St releases final EV numbers and it comes out with this, "While total early voting in Ohio only increased by 2.44% from 2008, early voting in counties that voted heavily for Kerry/Obama declined 4.1% while counties that voted heavily Bush/McCain increased their early voting by a shocking 14.39%."

Also (same link), "The next important piece of data is what the polls consistently report: Obama leads by huge margins among early voters but trails Romney among those who say they will vote on election day. This inverse in voting segments is why the proportion of early votes in the total votes — and that virtually every poll overestimated this proportion — is so tantamount. In most polls (which usually only have Obama leading by a small margin, although some give him a more comfortable ~+5%), lowering the percentage of early votes in the polling sample means lowering Obama’s lead drastically. And when Obama’s lead is only one or two percentage points, that can mean handing the election to Mitt Romney."

So if the polls show an OH tie, but are wrong by showing more early voters, and those early voters skew Dem, then Romney is winning OH and the presidency.
ColoradoDarin is offline   Reply With Quote