Potomac Tuesday prediction thread

[CultureKing]

Democrats:

VA -  Obama 55-60  %
MD - Obama  60-65 %
DC - Obama  65-70 %

GOP:

VA-   McCain 45-50,  Huckabee 40-45,  Paul 12
MD - McCain 50-55,  Huckabee  35-40, Paul 10
DC - McCain 60-65,  Huckabee 20-25,  Paul 8   

That looks right to me.

[Verily]

Their report seems extremely thorough: they interviewed 14,276 voters between February 7-8 and even projected delegates for all races.

Actually, getting really large sample sizes is, to me, a bad sign, not a good sign, as it indicates that the polling firm doesn't understand what makes a good poll.  The emphasis of a good polling firm should be on making your sample size representative, not large.  As was made so acutely apparent from the famous 1936 Literary Digest poll of some two million people, you can make your sample size as large as you like, but after you get above a thousand or so people, it really doesn't matter.  The Gallup poll for the same race surveyed less than a thousand people and got it bang on, in comparison to the Digest's absolutely laughable result.

What they appear to have done is try to get reasonable sample sizes in each CD. Which, of course, means an absurdly large sample size overall.

[Eraserhead]

God, I would certainly get a kick out of a Huckaupset in VA. That would be hilarious.

[Gabu]

What they appear to have done is try to get reasonable sample sizes in each CD. Which, of course, means an absurdly large sample size overall.

Oh, really?  Well, that's more reasonable then, I thought they were one of the one-off pollsters that just gets a huge boatload of respondents and is then like "LOOK AT OUR POLL WE'VE GOT SO MANY PEOPLE IN IT IT HAS TO BE RIGHT".

[Gabu]

Although, in thinking about it... if they got an equal number of people from each CD, wouldn't that make the overall statewide results skewed?  There won't be an equal number of people voting in each CD, which would mean that the ones with less people voting would be oversampled in the statewide results.

[Horus]

Maryland:
Obama - 61
Clinton 38

Virginia:
Obama - 55
Clinton - 44

DC:
Obama - 70
Clinton - 30

---

Maryland:
McCain - 57
Huckabee - 34
Paul - 9

Virginia:
McCain - 55
Huckabee - 37
Paul - 8

DC:
McCain - 66
Huckabee - 24
Paul - 10

[Verily]

Although, in thinking about it... if they got an equal number of people from each CD, wouldn't that make the overall statewide results skewed?  There won't be an equal number of people voting in each CD, which would mean that the ones with less people voting would be oversampled in the statewide results.

I didn't say they went for equal sizes. The more Republican CDs do tend to have smaller sample sizes than the more Democratic ones. The data's all in their PDF in unusually accessible form.

[Gabu]

Although, in thinking about it... if they got an equal number of people from each CD, wouldn't that make the overall statewide results skewed?  There won't be an equal number of people voting in each CD, which would mean that the ones with less people voting would be oversampled in the statewide results.

I didn't say they went for equal sizes. The more Republican CDs do tend to have smaller sample sizes than the more Democratic ones. The data's all in their PDF in unusually accessible form.

Oh yes, I suppose I should have looked at that, huh?

Well, in examining the data, it does actually look pretty good.  The samples from each CD seem to be fairly proportional to the number of delegates, which I assume has a fair correlation between the number of voters one could expect from each CD.

[Sam Spade]

I don't really feel like making any guesses this time around.  Isn't VA one of those potential Bradley Effect states, or is my memory just bad?

[True Democrat]

I don't really feel like making any guesses this time around.  Isn't VA one of those potential Bradley Effect states, or is my memory just bad?

Wilder experienced it big time, though it has been over 15 years since that happened.

[Josh/Devilman88]

All I have to say is, it isn't going to be a very good night for Hilliary.

[Keystone Phil]

MD - Obama between 60-65%

DC - Obama between 75-80%

VA - Hillary between 50-52%



McCain will be between 50-55% in MD and VA. He'll get between 60-65% in DC.

[perdedor]

Maryland - Obama 55-60%, McCain 50-55%
Washington D.C. - Obama 80-85%, McCain 50-55%
Virginia - Obama 50-55%, Huckabee 48-50%

[You are responsible]

Today, the ballot boxes are going to explode with vigor and decisiveness!


Well, I'm lazy... let me check my "official" uselectionatlas.org prediction.

Maryland:
Obama > 50% (Lean)
McCain > 60% (Strong)

Virignia:
Obama > 50% (Lean)
McCain > 50% (Strong)

D.C.:
Obama > 70% (Strong)
McCain > 60% (Strong)

[Adlai Stevenson]

An interesting new automated survey from Constituent Dynamics and RT Strategies uses a new methodology to model the allocation of delegates in today's Potomac primaries. Overall results:

Maryland: Obama 53%, Clinton 36%

Virginia:  Obama 51%, Clinton 34%

District of Columbia: Obama 63%, Clinton 27%

Bottom line: Based on individual performance within congressional districts, Obama will see a net increase of 32 delegates in his totals tonight.

http://politicalwire.com/

My prediction:

MARYLAND DEMOCRATIC PRESIDENTIAL PRIMARY -
62% Obama
33% Clinton
  2% Edwards
  2% Others


VIRGINIA DEMOCRATIC PRESIDENTIAL PRIMARY -
55% Obama
41% Clinton
  2% Edwards
  1% Others


DISTRICT OF COLUMBIA DEMOCRATIC PRESIDENTIAL PRIMARY -
71% Obama
26% Clinton
  2% Others

[Eraserhead]

I don't really feel like making any guesses this time around.  Isn't VA one of those potential Bradley Effect states, or is my memory just bad?

Wilder experienced it big time, though it has been over 15 years since that happened.

That was also a general election though.

[Alcon]

District of Columbia
Obama 67%
Clinton 30%
Edwards 2%

Maryland
Obama 62%
Clinton 34%
Edwards 3%

Virginia
Obama 55%
Clinton 40%
Edwards 4%

[HappyWarrior]

I'm going to my polling place at around 4.  So soon:-D

[Alcon]

I'm going to my polling place at around 4.  So soon:-D

First vote?

[Eraserhead]

Don't forget Democrats Abroad.

If Obama wins all 4 contests by a large margin, it's starting to look pretty much over for Clinton. That will be 9 large margin losses in a row. The best Clinton can do is hope they call New Mexico, and hope that everyone forgot that they actually voted a week ago.

Look for a "breaking news" bulletin on CNN when they call it.

[Sam Spade]

Personally, I suspect it won't happen (GE could be different), but who knows.

If the polling averages are correct, Obama should win VA by 17%-18%, which translates to 57%-58% in my book.  The polling averages also say he should win MD by 22%, which translates into 60%-61% to me.

[Meeker]

Virginia
Obama: 61%
Clinton: 37%
Edwards: 2%

McCain: 48%
Huckabee: 42%
Paul: 5%
Romney: 5%


DC
Obama: 68%
Clinton: 31%
Edwards: 1%

McCain: 61%
Huckabee: 27%
Paul: 8%
Romney: 4%


Maryland
Obama: 60%
Clinton: 38%
Edwards: 2%

McCain: 53%
Huckabee: 31%
Romney: 9%
Paul: 7%

[HappyWarrior]

I'm going to my polling place at around 4.  So soon:-D

First vote?

Yup.  Obama 08!

[falling apart like the ashes of American flags]

When Mason-Dixon, Survey USA and Rasmussen are in very tight agreement about Virginia, I think you should agree with them, too, J. J.

Hell, and Insider Advantage, too, for what they're worth.

Call it a hunch, but I think she'll do a bit better than predicted.

HA HA HA HA

J. J. = BRUTALLY F**KING OWNED

[Gabu]

When Mason-Dixon, Survey USA and Rasmussen are in very tight agreement about Virginia, I think you should agree with them, too, J. J.

Hell, and Insider Advantage, too, for what they're worth.

Call it a hunch, but I think she'll do a bit better than predicted.

HA HA HA HA

J. J. = BRUTALLY F**KING OWNED