So, today is September 16, 2014 and the November mid-term elections are exactly 7 weeks (42 days) away.
Some things have changed since the OP was written.
First, KS-SEN looks like it may become a GOP loss and an IND pickup.
Going down the list of battlegrounds,
the four races moving toward the Democrats:
NC-SEN:
RealClearPolitics - Election 2014 - North Carolina Senate - Tillis vs. Hagan
Current aggregate: Hagan
+3.7
When this OP was written on Aug 11, 2014, in NC, the aggregate on that day was:
Tillis +1.3
So,
Kay Hagan has moved the needle +5 in her direction since then.
----------------------------------------------------------------------------------------------
In the COL-SEN, a similar story:
RealClearPolitics - Election 2014 - Colorado Senate - Gardner vs. Udall
Current aggregate: Udall +3.7.
On August 11, 2014, when the OP was published, the aggregate showed:
Udall +1.5.
So,
Mark Udall has moved the needle 2.2 points more in his direction.
-------------------------------------------------------------------------------------------------------
MI-SEN:
RealClearPolitics - Election 2014 - Michigan Senate - Land vs. Peters
Current aggregate:
Peters +5.2.
On August 11, 2014, as the OP was published, the aggregate was:
Peters +4
So,
Gary Peters has moved the needle 1.2 points farther in his direction.
-------------------------------------------------------------------------------------------------------
IA-SEN:
RealClearPolitics - Election 2014 - Iowa Senate - Ernst vs. Braley
Current aggregate:
Braley +1.4
On August 11, 2014, as the OP was published, the aggregate was:
Ernst +0.8.
So,
Bruce Braley has moved the needle +2.2 points in his direction during that time.
Those were the
four Senatorial races where the Democrats' chances have improved.
--------------------------------------------------------------------------------------------------------
Two races in no-man's land:
AK-SEN:
RealClearPolitics - Election 2014 - Alaska Senate - Sullivan vs. Begich
Current aggregate:
Sullivan +1.3
NOT in the statistic is the Harstadt poll that just came in, showing Begich +5 (45-40):
AK-Sen Harstad Research for Senate Majority PAC
With that poll in the mix, the aggregate would then be
Begich +0.25.
On August 11, 2014, as the OP was published, an AK aggregate was not really possible, for the AK primary had not yet taken place. That being said, conventional wisdom was that in a race against Sullivan, it would be a dogfight, and see, it is.
------------------------------------------------------------------------------------------------------
LA-SEN:
RealClearPolitics - 2014 Louisiana Senate Race
In Louisiana, we have a statistical mess, caused partially by LA state law:
Officially, the aggregate in a pure two-man race is:
Cassidy +1.3
But Lousiana essentially does an open jungle primary for it's SEN-GE and that aggregate is:
Landrieu +5.4.
BOTH of those values are very faulty. The latest 2-way data is 2 weeks old and the jungle primary data has a poll from the end of June, which is very, very cold coffee. There is no way in hell that that poll should be in the current mix.
Mary Landrieu has survived a number of razor-thin elections in her life, I just don't know if she will pull this one out.
On August 11, 2014, as the OP was published, the aggregate was:
Cassidy +1.
So,
Bill Cassidy has micro-moved the needle +0.3 in his direction. All way, way, way within the margin of error for both sides.
FLIP A COIN.
We really need fresh polling from Louisiana to get a better picture here.
--------------------------------------------------------------------------------------------------
Races where the GOP is prevailing:
AR-SEN:
RealClearPolitics - Election 2014 - Arkansas Senate - Cotton vs. Pryor
Current aggregate:
Cotton +2.5.
However, as was the case with Alaska, a more recent poll is missing:
Pryor Steel Lead New Statewide Poll Ross in Dead Heat Southern Progress
That poll shows Pryor +4 (46/42).
With that poll in the mix, the aggregate would be
Cotton +1.5.
On August 11, 2014, as the OP was published the aggregate was: Cotton +3.
So, actually, the statistic has moved between +0.5 to +1.5 toward
Mark Pryor. Either way, Cotton is still a nose ahead here.
--------------------------------------------------------------------------------------------------------
GA-SEN:
RealClearPolitics - Election 2014 - Georgia Senate - Perdue vs. Nunn
Aggregate:
Perdue +3.
On August 11, 2014, as the OP was published, the aggregate was:
Perdue +3.2
Technically, the needle has moved +0.2 toward
Michelle Nunn, but that is micromovement. Perdue is building a lean but stubborn lead in the Peach State.
---------------------------------------------------------------------------------------------------
KY-SEN:
RealClearPolitics - Election 2014 - Kentucky Senate - McConnell vs. Grimes
Current aggregate: McConnell +5.2.
However, as was the case in Alaska and Arkansas, at least one poll is missing here:
http://images.politico.com/global/2014/09/09/14mem908_ky_-_d11.html
showing Grimes +1.
The Mellmann Group is also a reputable pollster and had a
+0.60 Republican mathematical bias in the 2012 election. It also nailed Obama's national margin.
I warned that, with time, RCP would deliberately not include some polls, but that being said, McConnell is posting some substantial leads from very well known pollsters, and I do not doubt their results.
On August 11, 2014, as the OP was published, the aggregate was:
McConnell +2.5.
So, for now, Senate Minority Leader
Mitch McConnell has doubled his lead over Grimes.
-----------------------------------------------------------------------------------------------------------
The possible Independent barnburner:
KS-SEN:
RealClearPolitics - Election 2014 - Kansas Senate - Roberts vs. Taylor vs. Orman
I included the RCP text in this screenshot, because I think it is important.
It is still too early to figure out an accurate aggregate for this race, but most likely, it will become Orman's to lose, putting an Independent from KS in the US-Senate.
--------------------------------------------------------------------------------------------------------------
Odd man out:
NH-SEN:
RealClearPolitics - Election 2014 - New Hampshire Senate - Brown vs. Shaheen
Current aggregate:
Shaheen +3.5.
This is an EXCELLENT case study on a possible outlier and this one is especially interesting. All four of these polls were taken at about the same time. Two polls, one from a right-leaning pollster (Rasmussen) and one from a left-leaning pollster (CBS), both show Shaheen +6. The hometown pollster, if you will, WMUR (which has a good reputation but tends to the Right), shows Shaheen +2. Were it just those three polls, then the aggregate would be Shaheen +4.7. But the CNN poll shows a mathematical tie, 6 complete points off of two +6ers for Shaheen.
On August 11, 2014, as the OP was published, the aggregate was:
Shaheen +10.4.
So,
Scott Brown has gained +7.1 toward his side in the last 5 weeks.
-------------------------------------------------------------------------------------------------------------------------
Facit: not that much has changed since August 11th, 2014. The four senate seats that most were assuming would firm up for the Democrats are indeed firming up and the open senate seats in the South are leaning more Republican, making it easy for the GOP to pick up 6 seats, considering that it practically already has SD, MT and WV in it's pocket.
It's too early to tell, but the current signs are that the two senate seats where the Democrats have seen (and should still see) pick-up possibilities are currently moving toward the Republicans.
The curent polling shows no sign of a massive GOP wave, but it also shows no real enthusiasm for the Democrats, either.