Senate Election Overview -- Democrats Hanging On?
Sorry for the overly-provocative title, but I'm a little surprised at how all the big media election-predicting sites have apparently decided to just call the whole Senate for Republicans and clock out early. Because I just don't see it as quite the slam-dunk everyone else does, at this point.
Partly this is because I eschew the whole "percentage prediction" model everyone else seems so enamored of. My columns rely a lot more on state-by-state analysis than computer modeling, to put this another way. And I do listen to my gut feelings, which is a big no-no in the world of professional statisticians. But, hey, at least I admit it up front.
There are a few changes in my state rankings from last week's column, including one very important piece of good news for Democrats down South. We've got less than two weeks to go, the debates are in full swing, and early voting has started in many of these states already. One caveat is that I'm starting to pay closer attention to how old the polling data is -- a poll from ten days ago isn't going to capture anything which has happened in the meantime, to put this another way. And we've got several states which weren't on anyone's radar as a possible close contest, so in some places polling data is very thin on the ground.
Having said all that, let's get to my picks for this week. As always, share your thoughts and your picks (where you think I'm laughably wrong) in the comments section.
OK, that's enough intro, let's get on with this week's picks, shall we?
Safe Republican
The Safe Republican list didn't change from last week: Alabama, Idaho, Maine, Mississippi, Montana, Nebraska, Oklahoma (both seats), South Carolina (both seats), Tennessee, Texas, West Virginia, and Wyoming.
There are fourteen Senate seats in that list, and two of them (Montana and West Virginia) are pickups for Republicans.
Safe Democratic
Likewise, the list of Safe Democratic seats didn't change any from last week: Delaware, Hawai'i, Illinois, Oregon, Massachusetts, Michigan, Minnesota, New Jersey, New Mexico, Rhode Island, and Virginia.
These eleven safe seats, however, do not include a single pickup, leaving the Democrats' net score at zero.
Leaning Republican
One state (Kentucky) from last week's list of Leaning Republican states moved down into tossup status. This leaves three Leaning Republican states this week: Arkansas, Louisiana, and South Dakota.
Arkansas hasn't shown much movement in polling, as incumbent Democrat Mark Pryor seems to be steadily trailing. The Democrats have deployed their secret weapon (whose name is Clinton) here, but so far it hasn't had much effect in the polls.
One recent poll in Louisiana showed the race tightening a bit, with incumbent Democrat Mary Landrieu pulling within three points of Republican Bill Cassidy. But Republicans have a clear edge here. The real question is whether either candidate could pull in the majority needed (50 percent plus one vote) to avoid a runoff. Right now, that doesn't look likely, so no matter who wins on Election Day, the seat might not be decided for another month.
South Dakota is one of those states where we're all relying on polling data that is quite likely past its freshness expiration date. This is no surprise, really, considering that nobody had pegged it as a race to watch even as recently as a few weeks ago. Given the volatile nature of the three-way race there, a strong argument could be made for moving this state down to Too Close To Call, just on the lack of up-to-date polling alone. For the time being, we'll keep it as Leaning Republican, based on the last poll taken. Keep a close eye on any poll results which may appear here, though.
Every one of these Republican leaners is currently in Democratic hands. So taking all three states would mean a pickup of three seats for Republicans. Added to their Safe Republican pickups, this means Republicans are up a net five seats -- one short of taking control.
Leaning Democratic
Then again, maybe not. To last week's two states (New Hampshire and North Carolina) we have to add a surprise. Because Georgia has moved up from Too Close To Call to Leaning Democratic.
Georgia is, so far, the one bright spot for Democrats late in the election cycle. In this race between dynastic candidates, Democrat Michelle Nunn is showing herself to be a much better campaigner than Republican David Perdue. Nunn's been holding her own in the ad wars, and Perdue's "I'm proud to be a job outsourcer" stand is going over like a lead balloon with the voters. Since the gaffe was made public, Perdue has lost his lead over Nunn, and Nunn's numbers have climbed. We really could use some new polling here, but at this point it's a pretty good bet that the more recent the polling is, the better Nunn's numbers are going to look. The trendline is clear, so for the time being Georgia has to be considered Leaning Democratic. Georgia is another state we might have a late runoff contest, it is also worth mentioning.
It's looking like one favorable poll for Scott Brown in New Hampshire was nothing more than an outlier. Since that point, Democratic incumbent Jeanne Shaheen has regained her edge, and seems headed for victory.
The story is similar in North Carolina, where incumbent Democrat Kay Hagen also seems to be holding onto her slight edge over Thom Tillis. Hagen is looking confident, and she just turned down a fourth debate in a show of strength.
Holding New Hampshire and North Carolina won't represent any net gain for Democrats, but winning Georgia will put Democrats up one overall. This means the absolute lead for Republicans is cut from plus five to plus four -- meaning they will have to pick up an additional race to take control of the Senate. If Michelle Nunn wins and Democrats keep the Senate, they'd better give her whatever plum committee assignments she asks for, because she will have earned some sort of prize.
Too Close To Call
We have five states in the Too Close To Call category this week, but not the same five as last week. Kentucky moved down here from Leaning Republican this week, and Georgia moved up to Leaning Democratic.
Alaska's race seems to be getting closer in the polls (even though they're slightly outdated), although incumbent Democrat Mark Begich is still the underdog by a few points. However, I've always gone with gut feeling here, and not trusted the polls as much as in other states (polling in Alaska just isn't as accurate as elsewhere, for a variety of reasons). Begich has launched a monumental get-out-the-vote effort in Alaska -- the biggest the state has ever seen. He may very well be reaching precisely those voters who aren't registering in the polls. I may be going out on a limb, but I truly think Begich may surprise some poll-watchers on Election Day. I could always be wrong, though.
Colorado polls are tightening, and incumbent Democrat Mark Udall seems to be in a virtual tie right now with Cory Gardner. This race could go either way, and in addition the electorate might be different than what the statisticians expect, due to Colorado's new "all-mail" voting this year.
Iowa's polling is neck-and-neck, and Bruce Braley seems to have regained the ground he lost in September. The latest poll shows him up, but within the margin of error. This could be an outlier, and it could be the indications of a trend. At this point, it's impossible to say. A few more polls might answer the question, but probably not decisively. For months now, we've all known that the Iowa race is going to be one of the closest-watched races on Election Day, and that hasn't changed at all.
Kansas has settled down somewhat after the shock of the Democratic candidate pulling out of the race, but it has settled down into a virtual tie. Incumbent Republican Pat Roberts seems to be running a tiny bit behind Independent Greg Orman, but Orman seems to have lost the large edge he got immediately after the Democrat dropped out. Kansas is currently the front lines of the civil war raging within the Republican Party, both in this race and in the governor's race. Right now, the edge goes to Orman, but just barely.
Kentucky was the other surprisingly good news for Democrats this week, as Alison Lundergan Grimes employs a "throw everything at the wall and hope something sticks" last-minute take-no-prisoners ad effort. What surprised everyone is that it seems to be paying off for her. The latest poll shows her only down one point from Senate Minority Leader Mitch McConnell -- which, if accurate, would put her back into contention. The national Democrats pulled their money out of the state when it was looking like McConnell had the race sewn up, but they just announced that they're reversing their position and will be pouring some money in for last-minute ads. Whatever happens, the race cannot be seen as Leaning Republican any longer, and so belongs here in Too Close To Call.
Before taking into consideration any of these tight races, Republicans are up four seats and Democrats are down by the same four. In absolute terms, this means 47 Republican seats and 48 Democratic. Meaning Republicans need to pick up four of the remaining five states, while Democrats will retain control by winning only two of them. Three of these states (Alaska, Colorado, and Iowa) are currently in Democratic hands, while two (Kansas and Kentucky) are currently Republican. This doesn't really matter in the absolute math, though, because Republicans need a total of at least 51 to win control, while Democrats only need 50 to retain control (because they have Joe Biden's vote to break any tie).
Now, I realize that this is a lot more optimistic for Democrats than a lot of other wonky election-calling sites right now. And this scenario really hinges on Democrats holding on to North Carolina and taking Georgia -- both of which are still rather bold assumptions (especially Georgia). If Michelle Nunn's polling trendline continues, however, some of these other forecasts may change.
Even if my assumptions come true, Democrats could still indeed lose control -- that's worth mentioning too. And there's always the Independent wildcard in Kansas to consider -- if Greg Orman wins and becomes the pivotal vote, he may very well decide to caucus with the Republicans, and thus swing control of the chamber. In that case, Republicans would only need three wins from the other states to gain control. Plus, there's the uncertainty of possible runoff elections in Louisiana and Georgia.
But (at least from where I'm sitting) the election doesn't seem to be quite the slam dunk many others are now predicting for the Republicans. If Democrats can hang on, the media storyline will be to their benefit, as a result of pollster predictions right now. The headlines will be "Dems Hang On!" which isn't really completely deserved because they're still going to lose four or five seats overall. But my guess is it'll be seen as a huge victory, if it comes to pass.
[Program Note: I'll be running two more of these "call the Senate" columns, one next Wednesday, and then one on the Monday right before the election, where I will make my final picks for all of these races.]
-- Chris Weigant
Cross-posted at The Huffington Post
Follow Chris on Twitter: @ChrisWeigant
"The national Democrats pulled their money out of the state when it was looking like McConnell had the race sewn up, but they just announced that they're reversing their position and will be pouring some money in for last-minute ads."
Ugh. How many ads do these people think we need to see? It seems like there has to be a point at which more ads is just a waste of money. There is no shortage of TV ads. They need to focus on GOTV. That's what will win it. There are about half a million more registered Dems in KY than Republicans.
I got a large glossy mailer from Worker's Voice today. It says "Stop Mitch McConnell", but makes no mention of ALG at all. I'm not sure what I think about that strategy, but at least it does a good job of highlighting how evil he is.
The statistical professionals rely on their their gut feelings as much as anybody else....they just call them "underlying assumptions." The key one being "Past is prolog." Which it is, but in the sense that Mark Twain understood it...history doesn't repeat itself, it rhymes.
All the top dog modelers are using pretty much the same polling data....yet the range of predictions spans roughly 95% - 60% Rep takeover on any recent day. Simply change assumptions about how independently states behave and you can move the percentages about 5%. WAPO's near certainty apparently reflects their assumption that aggregated polls almost always predict Election Day outcomes this late in the game. We'll see. Stated probabilities are just a quantification of belief at any given outfit. CW.com's p = too close to call is pretty close to The Wonk Central Tendency and that grand call is mathematically consistent with CW's underlying state by state calls, at least according to the Stig model, which is in turn consistent with the big 4 state by state calls. The Stig sez states veer blue to red in packs, NYT sez no, they, don't.
It's all going to come down to who's gut is right. May the best gut win. In the public arena, you are only as good as your last prognosis. Tough crowd.
OK, in a previous comment I said I'd post my past record today. I didn't have time, but I'm putting it together for next week.
I've only put the Senate under the microscope in two past midterms, 2006 and 2010. Here are the columns:
http://www.chrisweigant.com/2006/11/06/here-are-my-election-picks-what-are-yours/
and:
http://www.chrisweigant.com/2010/11/01/my-final-midterm-election-picks/
I've got to go back and check the results, state by state and see where I got it wrong. I thought I had done this before, but it turns out I only did it for presidential years (where my record is pretty good).
I promise, I'll dig out all my stats for next time around.
-CW
Well, I guess this commentary can be summed up by one thing..
Hope Springs Eternal :D
I mean, come on.. Not one single poll gives Democrats any better than a 40% chance...
If the polls were reversed and Democrats had the advantage that the GOP has, Dems would be already handing out committee assignments.. :D
But I won't begrudge ya'all yer positive thinking.. Gods know I have been there and done that.. :D
Michale
And, who knows..
If all the reported voter fraud and machines that take a vote for GOP and turn it into a vote for a Dem pans out, Demcorats just might win it..
That's the motto of the Democrat Party...
By Hook Or By Crook....
Michale
They need to focus on GOTV. That's what will win it.
The problem with GOTV strategies is that they serve BOTH parties...
If Dems push a GOTV message, it might backfire on them and energize the GOP voters...
Of course, the converse is true as well..
Michale
I have to admit that I am somewhat torn on how the GOP should play it after they get control of the Senate..
The immature and childish side of me wants the GOP to stick it to the Democrats and stick it hard.. Make it hurt in any and every way imaginable.. Cause severe pain for Democrats as Democrats have caused severe pain for this country...
But the mature side of me realizes that THAT won't be good for the country as a whole.. The mature side wants the GOP to get things done. On the plus side of this option, there IS some red meat for the immature half in this...
Be the bigger man and show Democrats how an "effective" Senate is run... Shame Democrats by taking the moral and ethical high ground... Show Democrats that a Party doesn't need a SuperMajority to get things done...
Well, I have a week and half to reconcile the two sides of me.. :D
Michale
One thing that surprises me is that Democrats aren't pointing to gas prices..
In my area, gas has hit below $3 a gallon. We haven't seen prices that low in so long I have forgotten when....
Wonder why Dems aren't trumpeting that to the high heavens..
Michale
It's daylight, so I'm drinking a pot of coffee and firing up my CW.com dedicated XCL spreadsheet.
Coding Georgia....carry the one....rank order the scores... hoyven, count down 16 rows, the 16th score is zero = too close to call. Control of the Senate is Too Close too Call.
CWs state by state landscape is just a wee bit better for the Dems this week. Still 11 states genuinely in play (a lot compared to the big 4 number shops), but the Dems pick up a leaner from the Reps. The Dems have to run five of the 11 competitive states, Reps need to run 7. That slight asymmetry suggests to me the CW picks imply a slightly better than even chance of the Dems retaining control. I'm not sure CW would go that far out on the limb.
M - What "Not one single poll gives Democrats any better than a 40% chance" are you talking about? There is no national race to poll, just 36 individual state wide contests.
CW is more optimistic than most, but it's not cockeyed optimism. Senate polling is sparse compared to presidential polling, and its decidedly unreliable in some states, especially Alaska. Changes in personal communications are making it more and more difficult to construct unbiased samples. Getting out the vote may be the deciding factor, Dems may still hold an advantage here..or not. Pulling ad money may make sense if the TV market is already saturated. Candidates still have a chance to make a major blunder (dead girl, live boy or the Weinergram).
I'm still sticking with about a 40% chance of the Dems pulling off an upset. That's much better odds than I gave Romney on the eve before the presidential election. I think the numbers are still pretty fluid.
Gas is down to $2.80 in my area - $2.40 with discounts at my local grocery. I put a months worth in my tank yesterday.
Regardless of the outcome, the Congress will remain obstructionist. Republicans don't want to lead. It doesn't make political sense for them to lead. Their base won't let them lead. They will play Charlie Brown football with Obama a few times...and he'll fall for it a few times. This is likely to play out as Republicans trying to "improve Obama Care to Death." The veto is still powerful. The senate winds will shift in 2016. Romney will be the Presidential nominee again, and he will lose.
CW is more optimistic than most, but it's not cockeyed optimism.
It's still optimism, not reality...
Republicans don't want to lead.
Assumes facts not in evidence..
The senate winds will shift in 2016. Romney will be the Presidential nominee again, and he will lose.
Now THAT is cockeyed optimism... :D
Michale
Republicans don't want to lead.
And, by the admission of Weigantians, Democrats CAN'T "effectively" lead...
So, give me a reluctant leader over an incompetent leader any day of the week and twice on Sunday...
Michale
Michale [4] -
My point exactly. What you're quoting aren't "polls" they are "probability predictions." So, to state it properly, "none of the sites gives Dems over a 40% probability of winning."
OK, fine. But that's not the same as state-by-state polling. Check out RealClearPolitics and click on "Senate no toss ups". Today, they put it at 51R, 48D, and 1Ind(KS). Which means that if Dems manage to rally and win in ANY ONE of: AK, CO, IA, AR, LA, KY then they will retain control.
http://www.realclearpolitics.com/epolls/2014/senate/2014_elections_senate_map_no_toss_ups.html
Electoral-vote.com shows a similar picture:
http://electoral-vote.com/
Today's stats: 48D, 50R, 2 ties (KS, IA). If the independent wins in KS and Braley wins in IA, Dems control the Senate.
I'm not saying the odds aren't with Republicans -- they are. What I'm saying is that I don't deal in odds at all. I use a different lens, and the way I see it, this is no slam-dunk for Republicans. It's still very close, and there has been movement towards Dems.
Will Rs take control? Maybe. Maybe even "probably" -- but, again, I don't deal in probabilities.
The one thing that is for sure, this is not the "wave" election some Rs were predicting. At one point, Republicans thought they were going to take Oregon and Michigan. Now no sane R thinks that's possible. Now, they're about to lose Georgia and possibly Kansas and South Dakota. That is a big turnaround, no matter who else sees it. Reality-based, my friend, reality-based....
:-)
Michale [6] -
You might have a point about GOTV, so let me ask you something I haven't heard anyone else talk about: because of the all-mail voting in CO, which party do you think it will help? Again, I don't really have an answer to that (nobody else does, either, I think). But it's an interesting question to ponder.
Michale [7] -
I think it'll depend on how big their majority is, if it comes to pass. If they've only got 51, McConnell may be more cautious. But no matter how big their majority is, they will not be shy about claiming a "mandate" (something the Dems never seem to do, which I consider a tactical mistake). And I wrote earlier, I think for the first 100 days McConnell and Boehner will allow the Tea Party free rein, and that a whole bunch of extreme bills will pass, only to be vetoed. I think the real battles will be waged later in the year, though, after the grandstanding ebbs. That's my take on it, at any rate, but the GOP could always surprise me (they've done so before).
Michale [8] -
Good point, one I've wondered about too. Another tactical mistake by Dems, I have to admit.
TheStig -
Amazing what a difference the Georgia race makes, isn't it? The biggest flaw in my analysis could be that we're overstating Nunn's climb in the polls there, and that Dems won't turn out and actually vote. But if you put GA in the Dem column (another poll just confirmed Nunn's lead), then the rest of the map starts looking decidedly better for Dems.
If I was going with "wild-eyed optimism" (Michale, this is for you, too), I would call the races in IA, CO, and AK for Dems, just because that is what my gut tells me. I would also say that GA will not have a runoff, but that LA will, giving Landrieu a second chance (a slim one).
My gut, I have to say, is much more optimistic at times than my brain. If my gut was right about the three, but Landrieu still lost, the count would be: D51, R48, 1Ind(KS), and Orman would then caucus with Dems. 52-48 split outcome.
See, Michale? THAT is what unreasoning optimism looks like!
Heh.
-CW
You might have a point about GOTV, so let me ask you something I haven't heard anyone else talk about: because of the all-mail voting in CO, which party do you think it will help? Again, I don't really have an answer to that (nobody else does, either, I think). But it's an interesting question to ponder.
It's not too hard a question.
Since mail in voting makes it ton easier to cheat, Mail In Voting will help Democrats.. :D
If I was going with "wild-eyed optimism" (Michale, this is for you, too), I would call the races in IA, CO, and AK for Dems, just because that is what my gut tells me. I would also say that GA will not have a runoff, but that LA will, giving Landrieu a second chance (a slim one).
In a two-person match-up Landru is on the short end of a 60-40 split...
See, Michale? THAT is what unreasoning optimism looks like!
Agreed... :D
By the by, I watched the new DREDD yesterday... It comes close, but it's still not as good as Stallone's DREDD...
Michale runs and hides :D
Just kidding.. It was definitely better than JUDGE DREDD.. More gritty, much MUCH less comedy.... I like Karl Urban...
Not so much for his McCoy, although that is pretty good. But as a gritty Dredd-like cop in ALMOST HUMAN, he was awesome...
Michale
I forgot to mention something:
http://www.courier-journal.com/story/politics-blog/2014/10/17/realtors-say-mcconnells-for-sale/17461287/
If you look closely at the photo, you can see a thin black line under the words For Sale. That's because it had a flap. If you lifted it up, it said "Sold". Truth in advertising.
He's going to "fight to maintain the home mortgage interest tax deduction". ALG probably wants to take that away from hard working Americans to finance Obama's War on Coal.
Michale -
Man, you got me, there. I was flexing the fingers to blast your taste in movies, then found you were pulling my leg!
Yeah, the new one is TONS better. Reasons why:
* Actor never takes off helmet
* Cityblock much more realistic
* Dredd is total hardass, all the time
* Dredd is not played by Sly Stallone
* Dredd never takes off his helmet (wait, did I say that already?)
Heh. Actually, I had one minor complaint with the cityscape scenes -- the interiors of the cityblock were good, but the exteriors were wrong. There should be one of those skyscrapers on EVERY block! Not some wasted landscape with a tower here and a tower there. But like I said, that's a pretty minor complaint.
The Stallone movie tried something pretty ambitious, which I hope the new franchise will get to eventually, which was to take two actual comicbook storylines and make them the basis for the movie. Dredd heading out to the Cursed Earth and running into Mean Machine and Pa and the rest of the warped family (forgot their last name) is a great story arc (where most of the family gets killed at the end, but Mean Machine -- the guy with the dial on his forehead -- came back as a recurring villain because everyone thought he was pretty cool). Like I said, hopefully future Dredd movies will go back and do it again.
-CW
Michale -
Angel. Mean Machine Angel, part of the Angel Gang. Had to look it up...
-CW
"Another tactical mistake by Dems, I have to admit."
Chris, you're kidding, right? They should get up and brag about something that they really aren't responsible for? They'd look like Michele Bachmann and they'd risk that the price-fixing oil companies would jack the price right up out of spite.
In fact, the price went up about 35 cents a gallon here on Tuesday. Now, that was McConnell's fault.
CW,
If you liked Dredd, you would probably like ALMOST HUMAN with Karl Urban.. It lasted only a season, but it was pretty good.. You will see a lot of DREDD in Detective Kennex...
It's a good old fashioned kick ass cop show with androids.. :D
Michale
Chris, you're kidding, right? They should get up and brag about something that they really aren't responsible for?
And yet, they blamed Bush for gas price spike when he was POTUS...
In fact, the price went up about 35 cents a gallon here on Tuesday. Now, that was McConnell's fault.
Of course, it was...
Thank you for proving my point.. :D
Michale
John From Censornati -
You raise a valid point, but I have two gut reactions:
(1) All's fair in love, war, and election season. Whatever the public can be convinced to believe, they will believe in the voting booth.
(2) Obama's already taken a political hit (see: 2012 election) for high gas prices, so he might as well take some credit when the price eventually does go down.
Both of those are pretty cynical positions, I realize, but then hey, it's an election year.
I am actually a firm believer that presidents and politics are largely (not totally, but largely) independent of the buisiness cycle. Politicians can pass things to hinder or help the economy, but it's mostly tinkering around the edges, and it's mostly reactionary.
To put this another way: there simply is no big red lever in the Oval Office labelled "Jobs" where if you just give it a yank, the economy does better.
But this is the home stretch of an election, and gas prices are one of those things which affects almost EVERYbody. It's a direct pocketbook issue, and if people have more money to spend because they're spending less on gas, then Democrats should point it out: things are getting better!
But I realize it's political opportunism. Then again, if Democrats are so timid they won't point out any economic good news, then maybe they deserve to lose.
-CW
Then again, if Democrats are so timid they won't point out any economic good news, then maybe they deserve to lose.
As I mentioned before, Dems have cried wolf on so many occasions regarding the "recovery" that they would likely be laughed at if they tried it now..
Even if, in some ways, it happens to be true...
Michale
Check out Obama's poll numbers..
http://www.realclearpolitics.com/epolls/other/president_obama_job_approval-1044.html
They are really taking a hit...
Who knows??
I might pull off a two-fer..
GOP sweeps the Senate AND Obama's numbers go below 40% into Bush territory... :D
JL, you got that T-shirt ready?? :D
Remember, you can't tell anyone why you are wearing it until the day after.. :D
Michale
CW,
The one thing I liked about JUDGE DREDD was that Dredd was larger than life. He was THE bad-ass of the Judges and everyone knew it. He was special..
In DREDD, he was just a run of the mill every-day cop.. Put another way.. If you were to pull the name DREDD and change the uniforms, it would have been a simple kick-ass cop movie...
Urban's Dredd wasn't anything special...
Michale
Desperation in the ranks of Democrats???
Running Scared: Maryland Democrats Play The Race Card
http://dailycaller.com/2014/10/23/maryland-democrats-play-the-race-card-in-latest-mailer/
I saids it befores and it's still a fact..
If Democrats didn't have the race card, they couldn't play the game...
Michale
That right there is the worst form of fear-mongering..
Fear mongering by way of false accusations of racism.....
Michale
Has Joni Ernst disqualified herself? She won't answer any editorial board questions.
Barack Obama, bewildered bystander
The one scandal where you could credit the president with genuine anger and obliviousness involves the recent breaches of White House Secret Service protection. The Washington Post described the first lady and president as “angry and upset,” and no doubt they were. But the first Secret Service scandal — the hookers of Cartagena — evinced this from the president: “If it turns out that some of the allegations that have been made in the press are confirmed, then of course I’ll be angry.” An innovation in ostentatious distancing: future conditional indignation.
http://www.washingtonpost.com/opinions/barack-obama-bewildered-bystander/2014/10/23/335bd0e2-5aeb-11e4-b812-38518ae74c67_story.html
Ya gotta give it to Chuck... He doesn't pull any punches when laying out the reality. :D
Michale
Has Joni Ernst disqualified herself? She won't answer any editorial board questions.
"Really!!??? That's great!! And, of course, you can PROVE that, right?? Oh yea, that's right. I forgot. You were absent the day the taught LAW at Law School.."
-Tom Cruise, A FEW GOOD MEN
:D
Michale
My point exactly. What you're quoting aren't "polls" they are "probability predictions." So, to state it properly, "none of the sites gives Dems over a 40% probability of winning."
But the probabilities are based on the POLLS..
The polls are the data and the probability is based on the data..
Michale
M-29
"But the probabilities are based on the POLLS."
In the same sense that beer is based on water. There are a lot of ingredients, and some craftsmanship.
Raw polls don't predict elections very well. How do we know this? History. Truman:Dewey etc. All polls lack complete precision and all polls are biased. Polling flaws are measured and corrected (to some extent) by comparing raw poll data to actual voting patterns.
You can make some very good predictions without using any polls. Mississippi: the democrat is going to lose. Historical trends work very well in the majority of locales. There are lots of other fundamentals that work well...name recognition, fundraising, incombency, organization, it's a long list. Put them together and you get a model. The model may live on paper, in a computer, or in somebody's head, it's still just a model.
History isn't an experiment. You can't re-run Obama vs Romney 1000 times to see how accurate Nate Silver's probability was. Nate's stated probability is just a measure of his faith in his call. Track record gives you some clue about how good a forecaster is, but it's nothing close to controlled experiment.
"My gut, I have to say, is much more optimistic at times than my brain. If my gut was right about the three, but Landrieu still lost, the count would be: D51, R48, 1Ind(KS), and Orman would then caucus with Dems. 52-48 split outcome."
One of the useful things about running Monte Carlo simulations is that with 36 races going on, you get a sense that one or two "rare" events are fairly common in any given run. For example, assuming 26 "safe" states (p=.99 for the favorite)and complete independence among states, at least 1 bastion goes down about 20% of the time. You don't get this sense by just looking at tabulated columns.
Unless I run WAPO landscapes, which basically make all states pretty safe, the 52-48 split isn't mind bogglingly unexpected. Silver's 538 is putting the odds of Dems retaining senate control a 38%, but he rates the chance of getting 52 OR MORE states at about 14%. Events that occur with that frequency aren't commonly thought of as rare. Dems retaining what they have, yeah, that would be pretty close to zero...but less close to zero if state behaviors are highly correlated.
It's a complicated game, which is why it's so entertaining.
RE 31, make that 14% an 18%. Things are volatile, check web sites often!
Well, I guess we'll know in about 10 days, eh? :D
Michale
http://www.washingtonpost.com/blogs/monkey-cage/wp/2014/10/24/could-non-citizens-decide-the-november-election/
This is why Voter Photo ID is needed...
Because those who vote criminally will always vote for the Party Of Free Stuff... The Free Ride Party...
Michale
maybe this pew research article has a clue as to why our "gut" reactions may differ from polling:
http://www.people-press.org/2014/10/23/as-midterms-near-gop-leads-on-key-issues-democrats-have-a-more-positive-image/
http://nypost.com/2014/10/25/former-cbs-reporter-explains-how-the-liberal-media-protects-obama/
Oh yea...
There is no Obama bias in the media... None at all..
{/sarcasm}
:^/
Michale