Senate Forecast - September 1 update

the chart (0.00 / 0)

The chart seems to assume that the Democrats currently have 60 seats.  I'm sure it's just a typo somewhere.  My guess is that when you moved WV into the chart you forgot to subtract it from 12/14.

---

by: pretzalz @ Thu Sep 02, 2010 at 20:41:48 PM UTC

Fixed (0.00 / 0)

thanks for catching it. Yes, it was caused by adding WV. The overall average was correct - just the numbers in the chart itself were off.

---

by: Matt @ Fri Sep 03, 2010 at 01:38:56 AM UTC

[ Parent ]

Alaska (0.00 / 0)

I think Alaska should be considered RL now, but no one in the above chart has moved it yet.

It looks like Murkowski can't run as a Libertarian, so the odds are increasing that she'll stay out of it.

And the most recent poll from Rasmussen has a lot of good news for McAndrews: only down by 6%, and a 22 point (!) lead with independents. If Murkowski does not endorse Miller, this could get very interesting very quickly.

---

by: SarahLawrence Scott @ Fri Sep 03, 2010 at 03:31:51 AM UTC

Changing a State v Changing a Forecast (0.00 / 0)

Putting "the forecast" together is often done once a month. It's more time consuming than you might think. Most sites do it that way, and tweak individual races throughout the month.

As elections get closer, the time frame shortens -- so this is Labour Day weekend, and people will start polling more and polls will change more -- "excitement factor" rises.

Just saying -- up until now, it was easier and more expedient to look at all the races once a month.  

---

by: DocJess @ Fri Sep 03, 2010 at 12:36:31 PM UTC

[ Parent ]

NC also (0.00 / 0)

I'm surprised the bug 3 (Cook, Rothenberg, CQ) still have this as an R, even though the polls are showing a closer race - and the poll based forecasts, 538 and DK both have it as RL. I would love for Rothenberg and CQ to justify AR as RL and NC as R.

---

by: Matt @ Fri Sep 03, 2010 at 11:34:56 AM UTC

Matt (0.00 / 0)

I just want to point out that I pegged that as a toss-up in November of 2008.  

---

by: DocJess @ Fri Sep 03, 2010 at 12:34:11 PM UTC

[ Parent ]

Absoloutely (0.00 / 0)

And the race has a lot of parallels to 2008, where Hagen came from behind to win easily. Yes it's a Republican year, but it's also an anti-incumbent year, and the polls are showing a single-digit race.

---

by: Matt @ Fri Sep 03, 2010 at 21:44:32 PM UTC

[ Parent ]

Updates (0.00 / 0)

Cook has moved WV D->DL, AR and OH T->RL. He moves from 54.7 to 53.9.

CQ has also moved OH to RL. CQ goes to 54.2.

Bowers has also updated, but his rankings are so sensitive to polling changes that they just jump around too much.

Overall average goes to 54.5

---

by: Matt @ Sat Sep 04, 2010 at 22:12:52 PM UTC

More updates (0.00 / 0)

Crystal Ball has made a number of changes, all trending red:

WV: D->DL
WA, WI: DL->T
CO, OH, PA: T->RL
IN, LA: RL->R

CB drops from 54.2 to 52.2, and the overall average drops to 54.2.

---

by: Matt @ Mon Sep 06, 2010 at 03:34:09 AM UTC

[ Parent ]

Rothenberg update (0.00 / 0)

AR: RL-> R
IN: T->R

Rothenberg goes from 54.6 to 53.9, and the overall average drops to 54.1

---

by: Matt @ Mon Sep 06, 2010 at 19:53:07 PM UTC

[ Parent ]

two questions (0.00 / 0)

http://www.cnn.com/2010/POLITI...

CNN Poll: GOP's midterm advantage is growing
By the CNN Wire Staff
September 6, 2010 7:58 a.m. EDT
"Independents and voters who dislike both parties are starting to break toward the GOP," says CNN's Keating Holland.

(CNN) -- With November's midterm elections less than two months away, a new national poll indicates that the Republicans' advantage over the Democrats in the battle for Congress is on the rise.

According to a CNN/Opinion Research Corporation survey released Monday, the GOP leads the Democrats by 7 points on the "generic ballot" question, 52 percent to 45 percent. That 7-point advantage is up from a 3-point margin last month.

The generic ballot question asks respondents if they would vote for a Democrat or Republican in their congressional district, without naming any specific candidates.

does the generic ballot tell us anything? and can this, or even the more specific trends posted on demconwatch be reversed in 60 days?

---

by: UplandPoet @ Mon Sep 06, 2010 at 13:31:40 PM UTC

Here's a good analysis (0.00 / 0)

http://www.pollster.com/blogs/...

To summarize, a 3% lead leads to a GOP takeover. A 7% lead would give them 231 seats. This is based on a formula that takes other factors into account, like how many open seats there are. The error on the regression is 5-9 seats, so a 7% lead pretty much secures a GOP majority.

And the CNN survey is for registered voters. If a likely voter sampling gave us 10%, things will be really bad.

Now can the trends be reversed? That's a very different question. Probably not significantly at this point. But can the Democrats target specific districts to try to save the majority? That's the focus at this point.

---

by: Matt @ Mon Sep 06, 2010 at 18:33:18 PM UTC

[ Parent ]