Stunningly Accurate Regression Model Predicts Outcome of House 2014 Elections

[tgards79]

It has the GOP gaining 4 seats in 2014 as of now. Check it out and other prediction methodologies:
http://www.borntorunthenumbers.com/2014/08/house-2014-election-update-dems-sure.html

[GeorgeGist]

1. Regression is right.

[bigdarryl]

2. If O signs an executive order on immigration

All bets are off plus with the rash number of police killing African American males I think the sleeper could be a higher number of African Americans coming out in the midterms than normally. Joe Madison is working hard on his show telling listeners to get out and vote in 2014

[former9thward]

4. AAs are concentrated in districts which are already Democratic.

Coming out in higher numbers will not help House races.

[tgards79]

7. Obama's team will plan Octpber...

I'm certain there will be as much aggressive activity as possible, executive orders, etc.

[konnie]

3. extrapolate blue

ok Smartypants, how about you use those numbers for the betterment, the good of the country
and tell us how many actual voters per district it would take to flip those districts blue?

not many I bet. they need to know there is a chance. they need to know that if they make
the effort they can make a difference even in a gerrymandered district.

[Liberal_Stalwart71]

5. Why are you sending this bad news out? Just fucking vote and get everyone you know

To do the same! We know that the House is lost. We need to retain the Senate and gain governorships!

[tgards79]

6. Senate update coming

Will be posting my Senate update on Monday. It will give every Dem a reason to start making calls to every swing state, if they needed another push.

[brooklynite]

8. Personally, I like to deal with reality...

I decided months ago that regaining the House was hopeless (supported by the opinions of the many House candidates I've talked to), that holding the Senate was essential, and that anything left over should go to winning back Governorships.

[BlueCaliDem]

10. That's where I am, too, brooklynite.

I know that there might be some districts that might flip (read somewhere that it's possible for Dems to win the House if they can register enough new voters and get out the vote) but as long as we keep the Senate, it will be a good 2014. Democratic voters, unlike RepublicanBots, are pretty fickle when it comes to midterms.

And yes, we need to take back governorships as well as State legislatures (where that gerrymandering has benefited Republicans, too - especially in Ohio). Let's hope that the American people are informed enough to do the right thing for the country on November 4th.

[Major Hogwash]

16. The Democrats will take back control of the House, and keep control of the Senate.

Much has happened since the last election, except in the Republican-led House of Representatives, where nothing of importance has been done.
Congress has the lowest rating in the history of ratings, and the President has bent over backwards trying to get them to work with him to solve all of the problems that the Bush administration, and the Republicans that were in control of Congress back then, have caused.

President Obama has been very successful for the last 6 years, even in the face of historic levels of obstruction by the Republicans in Congress.
Yet, imagine what the last 4 years could have been like if President Obama would have had some cooperation from the Republicans!

Well, a lot of other people have done just that . . and they are now ready to vote to give Democrats control of both houses of Congress.

[BlueCaliDem]

17. I hope so, MH. I really, really hope so.

It's vitally important that we flush the anti-We the People U.S. Reps, Governors, and State Officials OUT if we ever want a properous America again.

I don't know if we're able to win back the House (and dear god I hope you're right that we will) but I sincerely hope that Independents and Democrats come out and vote against Republicans come November 4th so we can give those lazy, government welfare queens in the House the boot and put people in who will actually govern for We the People and earn their generous salaries and government benefits.

[Liberal_Stalwart71]

13. Like I said, we know the reality already. Just GOTV!!!!

[demwing]

14. Did you read the article?

Because it says the Dems will do better than expected, and therefore is not bad news at all.

Relax.

[Liberal_Stalwart71]

18. Nope because politics is always moving, fluid and dynamic. Things change in a moment's notice.

I'd rather just act and get involved.

GOTV. Nuff said.

[tgards79]

19. Action is Good

The intent is to provide intelligence to guide the actions. That is, inform where to best allocate resources, like your time and energy.

[JaydenD]

9. How can any prediction be stunningly accurate?

[tgards79]

11. Reply

When you apply the equation to past elections, it is very accurate (see the chart in the article), within 2 seats of most actual outcomes. We'll see how well it does in 2014 but the equation is lining up well for now with all expert, district-by-district, analyses of the races.

[1StrongBlackMan]

12. I, so, love ...

this geeky political numbers stuff (which is amazing, because this dyslexic has never gotten higher than a "B" in any math class ... ever; but I digress).

I've book-marked the article, and the site, to look at more deeply; but, the President's Party Seat Change chart caught my eye ... particularly, those years where the model produced double digit inaccuracies (1950, 1954, 1958 ad 1966) and it begs the question of what happened in those years ... actually in the year (6 month) before those elections that threw the model off?

I'd hazard to bet there as some social "event/incident" that got people out to the polls.

[tgards79]

15. Answer

Actually it is a statistical/days thing. I could not find "generic ballot" information before 1970, and that is a powerful variable, so the absence of it skews the pre-1970 predictions. Undoubtably if I had such data it would eliminate those pre-1970 variances.