New Ranges, shift, and cause for movement.

House: D+4 - R+8 - [Shift: <-] (MI-11, IA-03)
Senate: R+0 - R+7 [Shift: <-] (No range changes, but more individual races are moving left than right)
Governors: D+6 - R+1 - [Shift: <-] (IL-GOV)

House: This leftward shift is probably temporary: accidental Representative Republican Kerry Bentivolio is considering running a write-in campaign after being trounced in a primary election earlier this year. If Bentivolio pulled the trigger it would almost ensure a Democratic pickup of this marginally Republican seat, but this is likely just one man who will never return to Congress again trying to keep the door from closing behind him. The Democrats are also showing strength in IA-03, currently held by retiring Republican Tom Latham, after their nominee Staci Appel posted a lead in a couple of reputable House polls.

Governors: Democratic incumbent Pat Quinn is alive again, after trailing badly in polling for the better part of a year (though mostly because the only polling that was being done in Illinois regularly was by GOP shill WeAskAmerica). Quinn posted a lead in a DGA survey and even posted an outrageous 11-point lead in an outlier from the Chicago Tribune. There's no chance that Quinn has a double-digit lead but the outlier wouldn't have even popped up had Quinn not made this race competitive again.

Senate: No change in the Senate ranges since the battlegrounds have been pretty much set (absent one massive change mentioned earlier) for the better part of a year. The only thing that's happening now is individual races moving towards one party or the other. Strong polls from PPP for the Democrats in North Carolina and Kansas are pushing those races towards the Democrats (in addition to New Hampshire, Michigan, Colorado, and Iowa) while Louisiana and Kentucky have moved towards the GOP. Georgia is still completely up for grabs, and a rare 4-way race in South Dakota is making that state a huge question mark as well.

Edit: A note on election forecasters. Most of the notable models are starting to move towards one another. Previously some (Washington Post's Monkey Cage and Nate Silver's 538 to name two) were being ridiculously charitable to the GOP's prospects of a narrow Senate majority but now most models are pretty bunched up near one another:

538: 54.7% of GOP majority.
Upshot: 51% of GOP majority.
Monkey Cage: 51% of Democratic majority.
HuffPo Pollster: 53% of Democratic majority.
Princeton Election Consortium: 70% of Democratic majority.

This post was edited by Pollster on Sep 16 2014 11:48am

The Kansas Supreme Court may have just dealt a crushing blow to Republican hopes of winning the Senate [Court-watcher Rick Hasen has the early coverage: http://electionlawblog.org/?p=65645]. Democrat Chad Taylor consulted with the Kansas Secretary of State's office (led by Republican Kris Kobach of Birther fame) on how to draft a letter requesting to be removed from the ballot pursuant to an important Kansas statute. The SoS office helped Taylor but after receiving his letter Kobach (who supports Pat Roberts) intervened and ruled that Taylor must stay on the ballot. Taylor took it to the Kansas Supreme Court who ruled unanimously that he was off the ballot after they skewered Kobach's lawyer during oral arguments. Kobach, who is facing a stiff challenge of his own, is left with only the option of suing to force the Democrats to name another nominee. Originally ballots were to be printed and mailed tomorrow, but Kobach is now going back on what he said in court about the deadline and has decided to delay printing for 8 days.

Kobach is basically left holding the bag here. His entire party was counting on him to be a good shill and to abuse his position, but now he has to decide whether to stick with a lawsuit that could cost him his own job or not. The electoral importance of this mess can't be overstated: Orman has led Roberts straight-up in every recent poll. If the GOP loses a seat in Kansas they can kiss any prayer of winning a majority goodbye.

I enjoy reading fiction, but i prefer to keep it separate from political analysis.

Kansas was already a toss up and the Republicans hopes of winning the senate do not rely solely on winning that seat.

They have 47 seats safe or leaning significantly in Republican favor. (Yes even Kentucky, I told you so)
Orman has said he will likely caucus with the majority party if he wins.
That means they have to win 3-4 other seats for a majority.

Of the other "Toss ups", Georgia, Arkansas and Alaska are favored to go to the Republicans. (78, 67 and 56%)
That makes 50 right there...

The 538 model also gives Republicans a 48% and 49% chance in Colorado and Iowa respectively.

A Republican majority is not a fading pipedream or a "prayer", its a probability.
There is a very real chance they win the majority even without Kansas.

This post was edited by cambovenzi on Sep 18 2014 05:49pm

Oh, it's not fiction. The Republicans are simply not currently in a place where they can afford to lose Kansas but yet be very confident that they can cobble together a majority. The same is true for all their seats because the additional climb to +7 is steep at this point. If there wasn't a noticeable difference then the models wouldn't be settling around a 50%-50% proposition.



Not a chance that is greater than microscopic, and fleeting. There is a degree of a universal swing in elections that can't be ignored. The GOP can not reasonably expect to lose Kansas but win a state like Colorado, Iowa, or Michigan. The odds are that they either win KS but lose the others, lose them all, or win them all. The GOP openly acknowledges that their odds in those states are low and shrinking and more importantly they know that triage time is around the corner. The Democrats have already prioritized their money and the GOP will have to soon as well. They won't cut off Ernst, Gardner, Tillis or even Brown because they all could theoretically win but not even the GOP's billionaire bankrollers can justify putting good money after bad.

Regarding your math: the models don't agree. And something to keep in mind about them and particularly 538: they're easily compromised by outliers. It happened today when they incorporated Quinnipiac's significant outliers in both Iowa and Colorado and gave outsized significance to the survey, causing the projections to adjust substantially. The models include the surveys in good faith but they still hurt the numbers. One single outlier in Colorado from Quinnipiac with a sample not even remotely within the realm of possibility can shift races towards an outcome that will be different the next poll. After knowing of Quinnipiac's well-established history of blowing Colorado, one would be crazy to look at an aggregate like this [See: http://www.realclearpolitics.com/epolls/2014/senate/co/colorado_senate_gardner_vs_udall-3845.html] and believe that the GOP can sub in CO for KS. They don't believe that and no one else should, either.

So their chances being wiped out completely is the same as a 50/50 proposition? Obviously not.
You are backtracking while not admitting you grossly overstated the position.



A democrat getting thrown off the ballot leading to more votes swinging towards an independent is not representative of what is happening in other states in the slightest.
A technicality leading to a single seat possibly being lost to a guy who might caucus with the Republicans anyways is not a sign that they'll lose all other close states.

As my post above showed, they don't need to go anywhere near Michigan or Kansas to have a real shot at a majority.

Bill Maher's infamous talking snake whipping boy Mark Pryor is likely finally on the way out in Arkansas and a Republican lead is building in Alaska and Georgia.
(MT, WV, SD are locks to swing Red and LA is going that way too)
Thats the big 50 right there.




..The models DO agree, you just believe the outlier is the reason.

I am aware of the Quinnipiac outliers.
I am also aware of the very recent USA today poll that showed Gardner up a single point and the litany of polls that have Ernst and Braley tied or within 1-2 points.
Both states are definitely in play for the Republicans at this point, who likely won't even need them to hit 50.

This post was edited by cambovenzi on Sep 19 2014 12:23am

No one said they were. You're actually just confusing (or conflating) two different set of assumptions: 1) where things stands now and 2) where things would stand, if the GOP lost KS. The odds of a GOP majority in the first case have been ~50% for months. The odds in the second case are much, much lower. As far as "backtracking" goes thanks for the laugh. One can follow the sequence of events pretty easily because I've noted several times how crucial the seat was and beyond that just how vulnerable it was to slipping away: first here, then two weeks later, then one week after that and finally yesterday. The commentary expressed in these posts is, as always, very consistent.



Like I said there's a degree of universal swing in elections that can't be ignored. A Republican loss in KS would be representative of what's happening in other states to that very degree. Senate elections are particularly isolated events but most of the races do tend to move in the same direction, together, and in predictable ways.

Nothing about this is a technicality. Orman might caucus with the GOP but he'd do so on the basis that they have a pre-existing majority. A pre-existing majority without Orman is not a particularly likely outcome. Once more: a majority is no better than 50%-50% even if they have Roberts. Him losing would mean something. Republicans will not lose KS but win more hostile states barring race-defining events such as a Todd Akin-style gaffe occurring that fundamentally changes the races.



Most importantly: There is no "big 50." Orman could caucus with the Democrats to retain their majority even if the GOP reaches 50 seats. The GOP needs 51 seats either WITH Roberts or WITHOUT Orman. Now that that's out of the way, everything else you wrote is either flat-out incorrect or is incredibly far from certain.

There's not a "lead building" in Alaska at all, it's debatable whether the Republicans have a lead period. The only "polls" that show them ahead aren't polls at all because they weigh for party ID. Actual polls have Begich leading by narrow margins. This is very apparent in the aggregate [See: http://elections.huffingtonpost.com/pollster/2014-alaska-senate-sullivan-vs-begich]. There is a lead in Georgia but it is within the MoE and it actually shrunk; on top of that runoff dynamic keeps the race as a tossup. There's nothing about the four-way race in South Dakota that's a lock. Louisiana isn't a lock to swing either way, let alone to the GOP. I myself noted how the race had moved in that direction in this post but the race won't be decided until December and therefore will be heavily influenced by November.



They do not agree with your math or your characterization of the races. They don't even agree with one another about hardly anything. You can review what they recently projected at the bottom of this post. I'll save you some time: they disagree with what you wrote and in some cases they disagree entirely. In regards to the outliers' impact on the models: they are the reason for the shift. The one model you cited explicitly stated that they were the cause for the noted shift [See: http://fivethirtyeight.com/datalab/senate-update-optimistic-numbers-for-gop-in-colorado-and-iowa/]. They and other outliers make up almost the entire difference between the new projections and those in the linked post.

The models do not agree with your claims. They don't agree with how many states are "leaning significantly" in the GOP's favor (especially when adjusting for said outliers) and they don't agree re: the GOP's odds if they lose KS. This is quite easy to demonstrate by going to HuffPo's model [See: http://elections.huffingtonpost.com/2014/senate-outlook] and moving the percentage sliders around in each race. After moving both CO and IA back to their pre-Quinnipiac outlier standing watch what happens as you slide KS towards the Democrats in the hypothetical situation.



The models don't agree with this either. One does (538) but literally every other model disagrees about Alaska. WaPo (70% Dem), UpShot (54% Dem), HuffPo (52% Dem), and the PEC (54% Dem) all have it going the other way, and that's on top of the fact that the race is being artificially skewed towards the GOP on account of the party ID weighing by Suckmussen and YouGov mentioned earlier. That's the danger in making your case on the basis of what one model says, especially when outliers or junk "polls" are presently having an outsized influence on the data. The models also disagree about how likely the GOP is to win Arkansas as well (HuffPo even thinks it's a tie) and the complications in Georgia still stand.

The Republicans simply cannot afford to lose Kansas unless they want to put their faith in a Black Swan. Of all the contested races there are only two that are more critical to the GOP trying to gain a majority than Kansas and they are West Virginia and Montana. Losing Kansas or any of the other contested races would really hurt their chances.

Also, just a quick range change:



House: This shift is not owed to any one particular race, but just to reflect some new House polling and newsworthy events that occurred.

CA-52: SurveyUSA has Democratic incumbent Scott Peters up 47%-46%.
NH-01: New NEC poll has Democratic incumbent Carol Shea-Porter up 46%-42%
NH-02: NEC also has Democratic incumbent Annie Kuster up 50%-37%.
NJ-03: Stockton College has the race between Aimee Belgard and Tom MacArthur tied at 42%-42%. A survey for the DCCC shows MacArthur leading 46%-43%.
NY-01: New Siena poll has Democratic incumbent Tim Bishop up 51%-41%.
NY-18: Siena also has Democratic incumbent Sean Patrick Maloney up 50%-42%.
NY-21: Siena also has Republican Elise Stefanik up by 13% over Democratic opponent Aaron Woolf.

Notable events outside of polling: Endangered Republican Steve Southerland in FL-02 caused a stir when he held an "all-male" fundraiser, and when pressed on it told the media to ask his female opponent if she had ever been "to a lingerie shower" as a way to insinuate that he's more in touch with women than she is. The NRA endorsed Democratic incumbent Nick Rahall in WV-03 while the Chamber of Commerce is backing Democratic incumbents Kyrsten Sinema (AZ-09), John Barrow in (GA-12) and Peters (CA-52). When Peters' opponent tried to scold the CoC's decision it publicly rebuked him by announcing that he too worked hard for the endorsement. The NRCC unexpectedly made a $1 million ad buy in MI-01 to help incumbent Dave Benishek. Republican Marilinda Garcia (NH-02) also got into some hot water when she fumbled an interview question about her own personal healthcare situation.

Taken into totality some important races shifted ever so slightly towards the Democrats.

NRA endorsement of the Donkey in WV-03 isn't surprising. You don't keep a highly conservative (and highly Democrat-gerrymandered) seat by being anti-gun.

50/50 does certainly not turn to 0 on the back of 1 seat that was already a toss up when so many states are in play.
You were wrong. Admit it.



Except not and I clearly told you why. A judge throwing someone off the ballot is not indicative of a swing in the countries electorate.


It isnt a technicality? Google what that means please.

I already explained to you in clear and unwavering terms why that could easily be the case.



COULD. Not "WILL".
The republicans can certainly take 51 without Kansas as well. The Math is right in front of you.

Montana, WV and SD turn red whether you stomp your feet or not. They need 4 more seats after that.
R are more than 2-1 favorites to take Arkansas and Louisiana according to some models. Less so in other models, but a lead nonetheless. (61% and 52% on huffpo)
Thats 6.

That means they need to win 1 measly seat in any of the tossups which include Alaska, Iowa and Colorado as the most likely suspects.
In those states both parties are withing a single point of each other according to the huffpo model.



I know you'd certainly like to throw out all discouraging results and pretend the democrats are winning or even in these states, but its simply not the case.
SD is extremely likely Republican. 90% on 538. Its an overwhelmingly Republican state with double digit leads in the polls.

I didnt say LA was a lock. I said it was going in that direction. Cassidy is crushing the democrat heads up in the most recent polls and it is very significantly more likely to turn red than not.




So some do and some don't agree with the exact percentages obviously, as I used the percentages from 1 model...

You cannot rightly say "hey these outliers are skewing the models thats why they show this" while simultaneously saying the model doesn't show it.

The fact of the matter is states like Alaska, Iowa, Colorado and Arkansas are very much in play for Republicans whether you want to link Democrat Senate Majority PAC saying its slightly in favor of democrats or the other way around.
A disagreement on the exact percentages in close races does not make what you said right.
There is a real possibility the Republicans win some of those states, and mathematically it would allow them to win a majority without Kansas.



and WV and MT are locks.

You're back into your fairyland rhetoric of black swans and prayers.
Their hopes of a majority do not vanish completely without Kansas, no matter how many times you say it.
Their chances worsening without Kansas is NOT the same thing as being wiped out completely.

This post was edited by cambovenzi on Sep 19 2014 09:37am

as an outsider looking in much of these numbers, claims don't make sense

can someone give me a very quick run down of the political system there and what these elections are about? i only remember the stuff about presidential elections and electoral colleges