The Jmhx Call: Dems Win Congress

[JMHX]

I've been keeping a long, ungainly spreadsheet database of polling data for House and Senate races since January of 2006, and at last I feel confident to make a relatively measured prediction on the November Mid-Term Elections.

United States House of Representatives: With 95% confidence, Democrats will gain between 18 - 23 seats, netting them control of the House of Representatives and a powerful check on the Bush Administration.

United States Senate: With 95% confidence, Democrats will gain 5-6 Senate seats, narrowly keeping control in Republican hands unless special circumstances arise, but with moderate Republicans and wavering independents, the working majority of Republicans will be essentially destroyed.

Why is this? Various reasons: Lobbying scandals, Page scandals, poor Presidential approval, an unpopular war, a generally do-nothing 109th Congress, all-time low levels of voters who are willing to call themselves Republicans - but most of all, it appears voters are ready to change leadership gears. They are tired of hegemonic domination of the government by Republicans, and far and away feel like their incumbents are not doing their jobs, but are rather focusing on self-enrichment. This is comparable to 1994, when Republicans swept into power on a wave of Democrat-targeted discontent.

Senate Democratic Gains
-----
Montana - Jon Tester
Ohio - Sherrod Brown
Rhode Island - Sheldon Whitehouse
Missouri - Claire McCaskill
Tennessee - Harold Ford

Senate Tossups
----
Virginia - Jim Webb v. George Allen

When you see states like conservative Indiana (which has not gone Democratic nationally since 1964) giving three Democrats House seats where previously untouchable incumbent Republicans sat for decades, it is hard not to imagine the beginnings of an underpolled landslide. Perhaps the tried-and-true GOP message is not working this time; perhaps voters have decided they will not follow the unfulfilled promises of incumbents for another term, and cast their votes in November as protest voters.

In time, 18-23 seats may seem like small potatoes.

Go vote this November.

[IllustriousPotentate]

Bah. At this point, it's like trading in a porta potty at a stadium for a toilet at a 5 star hotel. The hotel toilet Democrats may be cleaner than the stadium PortaPotty Republicans, but when it boils down to it, they're both full of it.

[200monkeys]

Democrats have it laid out for them, but they have a history of not doing anything......

[SEXY-FETUS]

IF the dems control congress it will be a republican in the white house.
This has been a FETUS prediction.

[Demosthenez]

At 10/17/06 09:11 PM, JMHX wrote: it is hard not to imagine the beginnings of an underpolled landslide.

Thats just it, the polls are still close in alot of places. And the Republicans have their machine and in many cases, incumbency. I dont think the Dems will gain the House but I wouldnt be very suprised either. I think they will gain back alot of seats but I doubt they will get the House back.

You said 18 seats may just seem like small potatoes but. . . its still 18 seats. And the economy is doing real good and gas prices are down (relatively). People will remember those things when they go to vote. 95% confidence seems a bit much to me but whatever.

[200monkeys]

You wonder why gas prices are down and the economy is doing well? THE ELECTIONS! :O!!!!!

[JMHX]

Incumbency is more of a curse than a blessing if you are an incumbent Republican - there is a palpable anti-incumbent feeling in the electorate this year.

[fli]

I feel that the Republicans will take control...
Gut feeling, I mean... Think about it.

Twice elected... one of the most unpopular president so far.

[Imperator]

At 10/17/06 09:34 PM, IllustriousPotentate wrote: Bah. At this point, it's like trading in a porta potty at a stadium for a toilet at a 5 star hotel. The hotel toilet Democrats may be cleaner than the stadium PortaPotty Republicans, but when it boils down to it, they're both full of it.

Best line ever.......

Game, set, match.

[JMHX]

If I'm wrong, I'm going to be eating fucking crow for two years on here.

If I'm right, I want some oral.

[EnragedSephiroth]

At 10/18/06 03:17 AM, JMHX wrote: If I'm right, I want some oral.

Tch I would probably demand some anal too. Anyway what's your raw data? Can you plot it on a spreadsheet and make a graphical representation for us? That would be most welcome :)

[Jose]

You missed the Minnesota senate election.
Star tribune reports a healthy lead of the democratic upstart over the republican incumbent.

[JMHX]

At 10/18/06 03:30 AM, EnragedSephiroth wrote:

At 10/18/06 03:17 AM, JMHX wrote: If I'm right, I want some oral.

Tch I would probably demand some anal too. Anyway what's your raw data? Can you plot it on a spreadsheet and make a graphical representation for us? That would be most welcome :)

My spreadsheet covers columns A - T, with about 650 lines of data. It's a bit big to post here.

[Buffalow]

You are not Colbert, and you are not funny.

[Gunter45]

At 10/17/06 09:36 PM, 200monkeys wrote: Democrats have it laid out for them, but they have a history of not doing anything......

Right, because the Civil Rights movement wasn't really anything big.

At 10/18/06 11:39 AM, Bull-Hound wrote: You are not Colbert, and you are not funny.

That made about as much sense as the aformentioned quote and was even less intelligent.

[RydiaLockheart]

What's your take on Pennsylvania?

[JMHX]

At 10/18/06 02:57 PM, RydiaLockheart wrote: What's your take on Pennsylvania?

Bob Casey has it, no contest. I'm expecting between 7-10 point victory on the conservative side of estimates. Either way, Casey's ousting Santorum.

[EnragedSephiroth]

At 10/18/06 11:00 AM, JMHX wrote: My spreadsheet covers columns A - T, with about 650 lines of data. It's a bit big to post here.

That's cool, could you use data analysis though? Any way to measure correlation, regression, confidence intervals? I saw you had a confidence interval of 95% leaving an alpha/error of 5%. It should be a normal bell-shaped distribution then no?

[MortifiedPenguins]

At 10/18/06 03:34 PM, JMHX wrote:

At 10/18/06 02:57 PM, RydiaLockheart wrote:

Bob Casey has it, no contest. I'm expecting between 7-10 point victory on the conservative side of estimates. Either way, Casey's ousting Santorum.

Why do I feel so sad knowing that the same two senators will be elected in my state, even if thier large fuckups.

[JMHX]

Kerry? A douchebag?

Shock.

[Jose]

At 10/19/06 02:04 AM, JMHX wrote: Kerry? A douchebag?

Shock.

What is your say on Minnesota, JMHX?

[JMHX]

At 10/19/06 02:11 AM, Jose wrote:
What is your say on Minnesota, JMHX?

Senate - Klobuchar by a large margin.

Governor - I'm giving this one even odds, though Pawlenty is currently up by 1.1% in my trends. However, he's been headed down since late August.

District 6 House Race I've had going GOP since late July, but I have recently switched it to Democratic gain over the past week and a half, given new data and general trends. Patty Wetterling has a fair shot at the seat.

I'm still really, really torn about the Governor's Race.

[Begoner]

I'm a bit more conservative in my estimates. At a 90% confidence interval, my call is:

+4 Senate seats for the Democrats.
+17-21 House seats for the Democrats.

Although the polls currently show a larger Democratic lead, many of the races are quite close, and could swing over to the Republican side quite easily.

[JMHX]

At 10/19/06 04:35 PM, Begoner wrote: I'm a bit more conservative in my estimates. At a 90% confidence interval, my call is:

+4 Senate seats for the Democrats.
+17-21 House seats for the Democrats.

What is your methodology to pull a 90% CI instead of 95%?

[Begoner]

What is your methodology to pull a 90% CI instead of 95%?

A lot of the races are quite close, as in "a couple of percentage points" close. Those could swing either way, so the best I can do is a rough estimate. If I went for a confidence interval of 95%, I'd have to say my guesses were +/- 10 seats, which is too big a selection.

[JMHX]

I meant statistically, what standard deviation are you using to get a 90% confidence interval as opposed to the standard 95% confidence interval? That curve doesn't add up.

[Begoner]

At 10/19/06 06:55 PM, JMHX wrote: I meant statistically, what standard deviation are you using to get a 90% confidence interval as opposed to the standard 95% confidence interval? That curve doesn't add up.

None. I pull predictions out of my ass -- what do you expect?

[Memorize]

Dems... slowing progress down 4 years at a time.

[JMHX]

At 10/19/06 08:01 PM, Begoner wrote:

At 10/19/06 06:55 PM, JMHX wrote:

None. I pull predictions out of my ass -- what do you expect?

For a minute you had me thinking you knew what you were talking about.

[Begoner]

For a minute you had me thinking you knew what you were talking about.

Indeed. Citing random statistics has that effect on many people. Nonetheless, I consider my estimate an educated guess that should not be too far off from the true result, and there is nothing mathematically wrong with it (what the hell do you mean by "the curve is off"?). I don't have the time to compile massive amounts of data to more accurately predict the results, so I conveyed that fact via a wider confidence interval. Maybe I'll put some energy into computing the statistics for a couple of key races before the election, though -- SurveyUSA has a wide selection of polls from a variety of congressional districts which are quite useful.