My model will flag internal inconsistencies between state by state predictions vs the overall national prediction. I don't need all the data, or all the process for that. I wouldn't have time to wade through it.

The three examples I analyzed where All internally consistent, two gave a grand prediction of "too close to call" and one gave a prediction "weak (=slight?)Rep favorite." All were well within the mainstream of prognostication. Which I currently read as trending 60% chance of Republican control, subject to change without much notice. I haven't even bothered to look today for updates.

The main advantages of defining categories in terms of probabilities is greater clarity and more precise comparisons between different prognosticators. I think the exercise of assigning perceived probabilities to their subjective classes would be illuminating.

So, say, if this owner is reticent to give you his data parameters, you could discern that he might be hiding something fishy...

I completely agree.... :D

I guess we'll know in a little over 2 weeks how right he is, eh?? :D

Michale

I don't want all of his data and process, just his belief of what his categories signify in terms of probabilities. Given this, my QDM model spits out a probability range of overall victory.

Example:

Strong D = 100% P Dems Win State (PDWS)

With this data,

Solid D = 99% - 80% PDWS

Moderate D = 79% - 60% PDWS

Weak D =59% - 50% PDWS

Weak R 50% - 41% PDWS

Moderate R 40% - 21% PDWS

Solid R 20%-1% PDWS

Strong R 0% PDWS

With this hypothetical set of author generated probability ranges, the somewhat nebulous QDM projection of "weak Rep victory" becomes 50%-41% chance of Dems retaining the senate and 50% - 59% of Reps taking over.

This allows a more meaningful comparison with quantitative forecasts like 538, HUFFPOL, WAPO and NYT, or gambling odds. Moreover, it forces the proprietor of a qualitative site to seriously reconsider the implications of how his or her categories are structured. Especially if categories are based on aggregated leads in the polls. Aggregated poll leads plus or minus something are not equivalent to probabilities of winning a state, plus or minus something.

I would hope qualitative pollsters like the redoubtable CW would latch on, so I throw the notion out to the blogosphere.

Interesting that you would say this..

Why, exactly would you want his data and process??

I put this analysis in to keep Michael from exploding, but I'm not sure it will do that.

Oh, I never explode... I just either gloat or sulk.. :D

But I do like the Election Projection website.. It's neat and concise and really lays it out.. The fact that it also re-enforces my own projections is simply icing on the cake.. :D

Michale

What have I been saying for over a year now???

That the 2014 Midterms are going to make the Great Democrat Shellacking Of 2010 seem like a family picnic by comparison..

Looks like things are shaping up to be proven dead on ballz accurate...

I promise I won't gloat... TOO much... :D

Michale

Just in case you were wondering about the link.. Ya know.. LINKS.. Something YOU never provide...

http://www.washingtonpost.com/blogs/the-fix/wp/2014/10/09/40-painful-seconds-of-alison-lundergan-grimes-refusing-to-say-whether-she-voted-for-president-obama/

How can Kentuckians trust a Senator who can't even answer a simple question??? Who won't take a stand on anything?? Who says whatever she has to say to get elected???

Fact is, they can't....

That is why ALG is going to lose...

If you think I am wrong... Make a wager. Put yer money where yer considerable mouth is....

No??? Didna think so....

Michale

I'm on the ropes, so I won't have any FTP suggestions this weekend.

Yea, sorry about that... But yer just too damn easy of a target.. :D

Michale

as to your part should the dems retain control of the senate, how about you wear a shirt in support of democrats?

Sounds fair to me. I'll even let you pick out the shirt.. :D

Michale

Once again, JFC proves he has no idea what he is talking about.. :D

Surprised to see it's ONE comment he hasn't played with his race cards...

FYI y'all - there have been exactly zero debates in the KY senate race, but there will be one on Monday.

So concentrated on the minituea to salvage yer decimated reputation, eh son?? :D

So it wasn't a debate per se, but rather an appearance before an editorial board..

Big whoopdee doo...

The simple fact is, she refused to own up that she voted for Obama. A very rookie mistake..

Chuck Todd called it... Stick a fork in ALG. She is done..

You want to make a wager on the Kentucky race, son?? Are you that sure of your facts???

Jeeze, look who I am asking about facts..

Still waiting to see that McConnell mailing you claimed you received...

:D

Michale

I'm on the ropes, so I won't have any FTP suggestions this weekend.

FYI y'all - there have been exactly zero debates in the KY senate race, but there will be one on Monday.

In addition, nobody outside of the cancervative bubble has heard of or cares about James O'Keefe's video. Coal will not be a deciding factor. Everybody knows that Dems are in favor of environmental protection and the GOP are union busters. Coal miners have also heard of fracking and natural gas. Hillbillies are not quite as stupid as the GOP propagandists seem to think they are.

Nor will not answering questions be decisive since neither of them will answer questions (thus far).

This will be decided by turnout which does not favor ALG. Bill Clinton is still popular in KY and
he's a friend of ALG's daddy. We'll see how many voters he can turn out for her and how many Rant Paul (R) can turn out for the King of Crony Capitalism. Rant and Jesse Benton will surely be holding their noses while they campaign.

if you show up in NYC, i will absolutely buy the beer, and craft beer at that, regardless of the outcome of our wager. as to your part should the dems retain control of the senate, how about you wear a shirt in support of democrats?

JL

Election projection uses 8 qualitative victory probabilities.

Solid D, Strong D, Moderate D, Weak D, Weak R, Mod R,Strong R, Solid D R.

No Toss Ups! NOW THAT'S WHAT I CALL FORECASTIN' said in my best Scots accent!

The implied qualitative probability falls in the weak Republican Victory class, the implied qualitative probability of winning at least 21 states according to the QandD model. I'm sure what fraction equates with "Weak" but I would hope the owner of the Election Projection could give us his perceived range of values. If not, I'd ask him why not?

I put this analysis in to keep Michael from exploding, but I'm not sure it will do that.

safe D, likely D, lean D, toss up, lean R, likely R safe R.

RCP is pretty noncommittal, lots of toss ups, which I equate with a fair coin throw, p =0.5

RCP Implied Probability of DEMS retaining control is a coin toss. In fact, the Implied Probability of DEMS winning 19 states is also 0.5. The latter is much at odds with the more fine grained models I track. That's the price of a coarse grained qualitative analysis.

Looking at things from the Republican side, the implied probability of winning at least 24 states is also 0.5 Also very much at odds with fine grained forecasts.

Once again, I suggest that map driven pundits would gain a lot by stating perceived probability ranges for all their categories, but especially the toss up, which should probably be considered as at a minimum spread of 47%-53%, based on pure polling sampling error considerations.

Between what her staffers were caught on tape saying and her dismal performance in this recent debate??

Stick a fork in ALG.... She's done....

Kentucky goes to McConnell..

Michale

I am also constrained to point out that Kansas has sent a Republican to the Senate since the 1930s... There is really no reason to think that will change..

Orman tried to wear the clothing of an Independent, but it's clear from his answers in the debate and from his past that he is an Obama Democrat.....

The people of Kansas are now realizing this, which is why Roberts is now favored in the race...

Michale

You buyin' the beer?? :D

AND the republicans would still have to actually win

Don't think that's gonna be a problem.. :D

should the democrats retain the senate, what will you do for your part of the wager?

I am open to suggestions...

Michale

for that you'd have to come to new york, AND the republicans would still have to actually win. should the democrats retain the senate, what will you do for your part of the wager?

JL

http://www.washingtonpost.com/wp-tran/politics/election-lab-2014

Of course, this may change before election day.. At one point, it was down to 51% chance for the GOP...

But so high a percentage so close to the mid-term???

I don't think anyone here REALLY believes that the Dems will keep control of the Senate...

Hay CW, you going to be doing live blogging on Election Day??? :D

Michale

Looks like Lundergan Grimes stepped on her wee-wee again...

How ya'all can even THINK to hope that Dems will take Kentucky is beyond me....

How the hell can the people of Kentucky think that ALG can be a competent leader when she can't even answer a simple question honestly and truthfully???

Michale

Awww, I was hoping to see ya in it!!

OK, that sounded weird.. :D

i'm not quite as bold as you were in '12, so i'd be forthright about it being a lost wager.

If only more rank and file Weigantians would have your gumption.. I would be a lot less frustrated around here.. :D

Michale

All I am saying is that you make a fascinating, compelling and accurate sounding case...

It's just not in keeping with the reality that ALL polls and predictions are making...

Having said that, you COULD be right.. That's more of concession than I have ever gotten from rank and file Weigantians... :D

Michale

Nope. My evaluation of the evidence leads me to believe the probability of Dems retaining senate control is somewhere around 0.4. Even odds is a bad bet. Betting is one thing, fundraising another. Mixing them cheapens both. Like rum and coke. I'm in for reasoned debate, not fraternity house chest bumping.

Intricate? You find notebook paper, a pencil , a column of states, and a column of probabilities intricate? All my model does is quickly generate the probability of winning elections, given somebody else's state by state estimates, and without the need of computer simulations. My quick and dirty model usually agrees quite closely with well known senate forecasters - NYT, 538, WAPO, HUFFPO. Across the range of optimists to pessimists. Even on-line bookmakers (2012 pres election).

You don't like the result? Then blame CW, it derives from his considered state by state probabilities, evaluated without fear or favor by a fixed procedure, presented transparently. That and the fact too close to call is generally interpreted in an election context as 50:50 , coin toss, p=0.5. Given the CW breakdown, the model says the probability of Dems winning at least 16 seats is Too Close to Call = 0.5.

Still don't like my model? Write your own, implement it, populate it, benchmark it, post it and defend it.

Nobody calls the the senate race 50:50? I think CW more or less did, even if he wasn't, explicit about it. I'm sure he'll correct me if
I'm wrong.

Actually, it would be interesting to know what probability of victory range CW would assign to Safe. To Lean, and to Too Close to Call.

"i'm your huckleberry"
~val kilmer - Tombstone

i will get a pro-republican t-shirt from my father in law (staunch republican), wear it to work if the dems lose the senate, and send CW a photo of it on me. i'm not quite as bold as you were in '12, so i'd be forthright about it being a lost wager.

JL

Do I smell concession in the air?? :D

Michale

Is ANYONE so sure of their ideology that they are ready to state for the record that the Demcorats will keep the Senate and to back up that claim with some fundraising...

*I* am sure enough of my facts to do so...

Is anyone else??

That's all I am saying...

Michale

Do you think the Democrats will retain control of the Senate???

As to Orman and Kansas???

Orman is quoted as saying he would relax the Dodd-Frank regulations.....

How do you like Orman now???

http://nypost.com/2014/10/07/why-democrats-may-keep-the-senate/

Yes... I'll readily admit.. The Democrat **MAY** keep the Senate... But the odds are against it...

But I am willing to put my wallet where my mouth is... :D

Michale

I suppose I should have written an actual comment, but mainly I just thought the fivethirtyeight forecast should also be included in the discussion.

Here's another one, from a total right-winger:

http://nypost.com/2014/10/07/why-democrats-may-keep-the-senate/

-CW

Good point about OK and SC, but isn't one of those SC races the one where Dems didn't even field a candidate? Anyway, I'll include them next time, mea culpa for the omission.

Michale -

I'll bet that the Republican who is running without a Dem challenger is going to win his race. How's that?

:-)

As for who is getting worried, take a look at this:

http://www.washingtonpost.com/politics/late-season-surprises-shake-gop-confidence-in-senate-elections/2014/10/08/66499c1c-4ef4-11e4-babe-e91da079cb8a_story.html

-CW

Still at 94% as far as I can see...

But if you are so sure that Democrats have a better than even chance of retaining control of the Senate, surely you can make a wager without any special considerations???

Straight up or down shot...

Dems keep the Senate...

Dems lose the Senate...

Any takers?? :D

Michale

You sound like Karl Rove saying that Romney will win the 2012 POTUS Election.. :D

There is no poll on the planet that gives Democrats a 50% chance of keeping the Senate...

While your process is intricate and chock full of stats and explanations, the simple fact is there is nothing to substantiate the claims..

It's like me explaining to you that I actually have 11 fingers... My explanation is compelling and it SOUNDS great...

But it's readily apparent that I don't really have more than 10 fingers.. :D

Michale

As always, I appreciate your connoisseurship in these matters. One, actually two minor quibbles...there are TWO South Carolina and TWO Oklahoma races. Granted all four are VERY safe Republican wins, but you need to include them all for scoring purposes.

As a helpful memory device, I always picture Steve Martin running around shouting Oklahoma, Oklahoma, Oklahoma in the movie "Dirty Rotten Scoundrels".
WAPO Election Lab doesn't include the extra SC and OK in their summary tabulation page, which as anybody who bothers to read my crap knows, drives me nuts. By the way Election lab now calls the race decided at greater than .99 probability of a Republican senate. That's going way out on a rotten limb IMHO, even if it doesn't exactly make them dirty rotten scandals. Misguided? I guess the "greater than" sign makes them look less of a prat should the race break Democratic, but.....

Just for fun, I've run your picks through my obscure Quick and Dirty Rank Ordered Model. Put the CW picks in order, starting with the Safe DEM, then the Lean Dem, then Too Close to Call, Lean REP and finally Safe REP. Since the QUICK and Dirty model evaluates the overall likelihood of senate control in terms of BLUE probabilities (RED are just BLUE-1) I've converted CW categories into the following QUALITATIVE BLUE probabilities.

HIGHLY PROBABLE, BETTER THAN EVEN, TOSS UP, LESS THAN EVEN, NEAR IMPOSSIBLE.

Fair enough?

Putting the state races in descending order of categorical probability, there are a dozen states in the HIGHLY PROBABLE category which means the DEMS are highly likely to win at least 12 states. There are 3 states in the less likely category, so DEMS have a better than even chance of winning 3 plus twelve equals 15 states, which is still a loss of senate control. There are 4 states in the TOSS UP Category, meaning DEMS have a roughly 50/50 chance of winning at least 4+3+12=19 states, which gives allows them to retain senate control (with caucus assumptions). DEMS have a less than even chance of winning at least 26 seats, which would actually amount to a blowout victory, but less than even is a pretty vague term. The chances of winning all the races is clearly nearly impossible.

All in all, the Quick and Dirty summary of the CW rankings imply that DEMS have a somewhat better than even chance of retaining senate control. An optimistic outlier, but not by that much. Not bad for chunked data, but it's very knife edged. Change things a little, and things look a lot different.

Read 'em and weep, people... :D

I know, I know.. It's an outlier...

But if ya'all put all your eggs in the Kentucky outlier, I can gloat about this 94% chance outlier... :D

Speaking of Kentucky, ya'alls own HuffPoop gives McConnel a 63% chance of beating the proven liar Grimes...

It's gonna be a rout!!!!! :D

Michale

Obamacare Cancellations: Blame 'AV Drift'
realclearpolicy.com/blog/2014/10/08/obamacare_cancellations_blame_av_drift_1094.html

If You Like Your Plan, You Still Can't Keep It
There's another wave of Obamacare health plan cancellations on the way.
usnews.com/opinion/economic-intelligence/2014/09/22/under-obamacare-americans-will-continue-to-lose-coverage

Face the facts, people...

TrainWreckCare is an Edsel... Just like it's namesake creator...

Michale

"If you like your doctor, you will be able to keep your doctor. Period.If you like your health care plan, you will be able to keep your health care plan. Period.No one will take it away. No matter what."
-President Barack Obama, 2009

Now, ya'all seem to think lying was a big deal back when ya'all accused Bush of lying even though Bush didn't really even lie...

I guess if the liar has a '-D' after the name, lying is perfectly acceptable...

Go figger... :D

Michale

Virginia is going to be the hardest hit with a quarter of a million policies being cancelled..

Bringing TrainWreckCare and Obama's LIE OF THE YEAR back to the forefront right before the midterm elections....

Again, save wishful thinking, there isn't ANY reason to think that the Demcorats will retain control of the Senate...

Sorry to be the bearer of bad news...

Well, no.. Not really.. :D

Michale

TRANSLATION: "It never really happened except in my wet dream..."

:D

Just like JFC's non-existent quote from DHS....

I have finally learned how to tell EXACTLY when JFC is totally bullshitting....

He posts comments. :D

Considering all the bad news that is coming in that make Obama look like a weak fool and considering that no Democrat wants to be in the same city with Obama and considering that Obama has definitively stated that HIS "policies are all on the ballot".....

Considering all these FACTS I really have to wonder how ya'all can believe that control of the Senate is in doubt...

I guess Hope Springs Eternal, eh?? :D

Michale

I put it in the recycling bin and it's already gone, but I will see if I can scare up another in my neighborhood this afternoon.

Hey, can you send me an image of that?!?

I'll contact you via email, just in case you don't already have an address you can reach me at...

:-)

I'm thinking... Friday... I can fit that in SOMEwhere...

YoYoTheAssyrian -

56.4% is awfully close to "a coin toss would give you just as accurate information"... I'm just saying'...

:-)

-CW