Midterm elections prediction thread

[Nate tha\\\' Great]

[JoseGonzlez]

ill call up dan dorfmann and make my guess.

seriously though if you could provide a link to a good projection site it might be useful for people to bet off this baseline as opposed to randomly guessing.


i follow fairly closely but dont really feel qualified to guess. i would vote dems beat the consensus but id like to see what the consensus is.

[5thStreetHog]

Interesting poll,but only if Dem`s and Rep`s here answer with their MINDS not their HEARTS.I wont have any problem there lol,so ill give it a wing [img]/images/graemlins/smirk.gif[/img]

[Nate tha\\\' Great]

My answers, FWIW, are yes (>=25 seats), no, yes and no.

It's the House, not the Senate, where I think the national mood is going to come into play. Basically, if you look at the races as they were forecasted a couple of months ago, almost all of the "toss up" House races are polling for the Democrats, and most of the "leans Republican" House races are polling as toss-ups. And the Republicans have to be worried about races exactly like Jim Leach's IA-2 district, which wasn't even expected to be competitive but which is polling as a dead heat. The Democrats may even pick off a House race or two that hasn't been polled.

I expect something along the lines of a 35-seat swing to the Democrats; it might even be closer to 40 or 45 seats. I don't expect the "silent majority" of white evangelical protestants to turn out for the GOP in the same way it did in 2004. In fact, if you look at polls of those voters, their support for the Republicans has dropped 20% since the last election cycle (from something like 75% to 55%).

In spite of this, I think the math works against the Democrats in the Senate. The candidates in Senate races are much more identifiable, and the races are much more likely to come down to local politics. From a Democratic point of view, I worry about having to bat 1.000 in red state after red state, and I suspect they'll stumble somewhere along the way. I do expect Jim Webb to pick off Virginia, but Tennessee is a probably a lost cause, and I think the Democrats will ultimately lose heartbreakers in Arizona (by 2-3%) and Missouri (in a race so close that it may trigger recounts and the like).

[Nate tha\\\' Great]

[ QUOTE ]
ill call up dan dorfmann and make my guess.

seriously though if you could provide a link to a good projection site it might be useful for people to bet off this baseline as opposed to randomly guessing.

[/ QUOTE ]

www.realclearpolitics.com

www.electoral-vote.com

www.tradesports.com

[Wynton]

Here's a summary of some numbers on the Arizona race:

http://pollster.com/polls/?state=AZ&...ce=senate_race

[disjunction]

Yes, no, I don't know (but whatever the consensus is, I think Leach will do better), and no.

Has there ever been an election where Republicans do worse than people thought? I think it was Berge who noted that when it rains it usually poors, but isn't an alternate hypothesis that Republicans just always beat the polls?

[Uglyowl]

[ QUOTE ]
And the Republicans have to be worried about races exactly like Jim Leach's IA-2 district, which wasn't even expected to be competitive

[/ QUOTE ]

I fell asleep on this one after reading the initial analysis of this race. This is great that this one is in play!

http://www.realclearpolitics.com/epo...owa_2-150.html

[Nate tha\\\' Great]

[ QUOTE ]
Yes, no, I don't know (but whatever the consensus is, I think Leach will do better), and no.

Has there ever been an election where Republicans do worse than people thought? I think it was Berge who noted that when it rains it usually poors, but isn't an alternate hypothesis that Republicans just always beat the polls?

[/ QUOTE ]

We should distinguish between exit polls and pre-election polls. One reason that Democrats felt very disappointed in 2004 is that throughout the course of election day, there were a lot of exit polls leaked that showed Kerry with leads in virtually every close race. Those polls proved to be inaccurate. But the consensus of the pre-election polls showed Bush with a 1.5% lead heading into election day, and his actual margin was 2.4%. Similarly, the results in 2000 were extremely close to the pre-election polls (remember that Gore actually won the popular vote that year).

What was a bit unusual about 2004 was that the Republicans won in a high-turnout election, something that was generally taken to favor the Democrats. This is what triggered a lot of the reverence for Rove's get-out-the-vote operations. However, many of these votes came from white evangelical protestants, and as I indicated above, support for the GOP amongst this group has declined tremendously.

I do expect that the GOTV operations will slightly favor the GOP in high-profile races like Missouri and Tennessee, which is why the Democrats face an uphill battle in taking the Senate. On the other hand, I expect the opposite dynamic to hold in the hundreds of House races across the country, which may not have been subject to heavy advertising and exposure.

[disjunction]

[ QUOTE ]

I do expect that the GOTV operations will slightly favor the GOP in high-profile races like Missouri and Tennessee, which is why the Democrats face an uphill battle in taking the Senate. On the other hand, I expect the opposite dynamic to hold in the hundreds of House races across the country, which may not have been subject to heavy advertising and exposure.

[/ QUOTE ]

I could see this. Everything is so murky to me, including the extent of the murkiness. Iowa Electronic Markets and Tradesports only have the House at 70/30? But the media experts make it sound like near 100% Democratic takeover. I don't quite understand where the difference is coming from.

On the other hand, I have no clue how someone can conduct an accurate poll nowadays when everybody uses their cell phone, and if you don't it clearly says something about the way you vote. I guess you can model and stuff, but I'm not sure if that would be called a poll, so I don't quite know what they're doing and how they come up with these numbers.

Finally, I see a lot of people on 2+2 that have changed their party, but I only know a couple in real life. Maybe that's significant, I dunno.