Mason-Dixon CA, GA, MO, NJ and AZ polls

[Tender Branson]

DEMOCRATIC RACES

ARIZONA

Clinton, 43 percent
Obama, 41 percent
undecided, 13 percent

CALIFORNIA

Clinton, 45 percent
Obama, 36 percent
undecided, 16 percent

GEORGIA

Obama, 47 percent
Clinton, 41 percent
undecided, 10 percent

MISSOURI

Clinton, 47 percent
Obama, 41 percent
undecided, 10 percent

NEW JERSEY

Clinton, 46 percent
Obama, 39 percent
undecided, 12 percent

REPUBLICAN RACES

CALIFORNIA

McCain, 40 percent
Romney, 31 percent
Huckabee, 13 percent
Ron Paul, 3 percent
undecided, 11 percent

GEORGIA

McCain, 33 percent
Romney, 27 percent
Huckabee, 18 percent
Paul, 4 percent
undecided, 17 percent

MISSOURI

McCain, 37 percent
Huckabee, 27 percent
Romney, 24 percent
Paul, 1 percent
undecided, 11 percent

NEW JERSEY

McCain, 46 percent
Romney, 31 percent
Huckabee, 5 percent
Paul, 4 percent
undecided, 12 percent

The McClatchy-MSNBC polls of 400 likely Democratic and Republican primary voters each in California, Georgia, Missouri and New Jersey -- and 400 likely Democratic primary voters in Arizona -- was conducted by telephone Jan. 30-Feb. 1. The margin of error was plus or minus five percentage points.

http://www.newsobserver.com/politics/story/920176.html

[Alcon]

I give up.

Sleeping through Monday.

[Reaganfan]

I added all of them to the polls section, thanks.

[Eraserhead]

MD has built quite the record of underestemating Obama this year.

[Ben.]

How come MD always seem to have such high undecided tallies?

[Keystone Phil]

I'm satisfied with the numbers for the Dems and absolutely thrilled about the results on the GOP side.

[Eraserhead]

MD totally blew Iowa and South Carolina for the Dems. They did pretty well in NV, although they overestimated Clinton's margin of victory slightly.

[Filuwaúrdjan]

Isn't it great that the Obamahacks seem to trust Zoggy more than Mason-Dixon? Why? The answer is obvious and doth not reflect so well...

[Ben.]

Isn't it great that the Obamahacks seem to trust Zoggy more than Mason-Dixon? Why? The answer is obvious and doth not reflect so well...

I'd trust MD more than Zogby but it doesnt detract from the fact that for some bizzare reason Zogby has by and large been better so far than MD (that could of course change).

[Filuwaúrdjan]

Isn't it great that the Obamahacks seem to trust Zoggy more than Mason-Dixon? Why? The answer is obvious and doth not reflect so well...

I'd trust MD more than Zogby but it doesnt detract from the fact that for some bizzare reason Zogby has by and large been better so far than MD (that could of course change).

Has it? Looking over the results so far and comparing them to the polls, I'm not quite sure how this can be concluded without a fair degree of wishful thinking (especially when caucuses are ignored; Zoggy seems to do better when there be no secret ballot).

[Adlai Stevenson]

If Obama can hold Clinton to narrow victories in Arizona, California, New Jersey etc. then I don't think thats too bad for him.  If they about half the delegates then its still more of a victory for him than it is for her.  And he will win Georgia I think.

[Gabu]

Isn't it great that the Obamahacks seem to trust Zoggy more than Mason-Dixon? Why? The answer is obvious and doth not reflect so well...

I'd trust MD more than Zogby but it doesnt detract from the fact that for some bizzare reason Zogby has by and large been better so far than MD (that could of course change).

Has it? Looking over the results so far and comparing them to the polls, I'm not quite sure how this can be concluded without a fair degree of wishful thinking (especially when caucuses are ignored; Zoggy seems to do better when there be no secret ballot).

You know, people would probably find it easier to talk to you if you'd stop talking in oblique terms that assume people already know what you mean.   No offense, but the number of times I've heard "it's obvious" or something along those lines is kinda large, and it gets kind of annoying because there's a good chance that people do not, in fact, know what you mean.

[Filuwaúrdjan]

You know, people would probably find it easier to talk to you if you'd stop talking in oblique terms that assume people already know what you mean.

It's a perfectly logical assumption . In this case all people need to do is look at the polls and results were can be found elsewhere on this site.

[Gabu]

You know, people would probably find it easier to talk to you if you'd stop talking in oblique terms that assume people already know what you mean.

It's a perfectly logical assumption . In this case all people need to do is look at the polls and results were can be found elsewhere on this site.

I did look up the polls.  And honestly, from what I could see, I think people have a point.  Mason-Dixon had no real relevant polls for Iowa and New Hampshire, their polls for those states having come around a week before the actual vote.  Even then, their poll from January 4 still showed Obama winning New Hampshire by a margin similar to other polls released at that time period - in fact, Zogby did a poll released on the same day that was better than Mason-Dixon's in terms of the actual result.

They did release polls for Nevada and South Carolina that were close to the vote there, however... and as far as I could see, their results were, in fact, worse than Zogby's.  In both cases, Zogby came closer to the actual outcome than Mason-Dixon.

Zogby did definitely blow New Hampshire; I'm not disputing that, as they had the worst results for that state.  Although they certainly were not alone; there was not a single poll that came out before the election that actually predicted a Clinton victory.  But they basically completely nailed Iowa, were the closest pollster in terms of margin and Edwards vote in Nevada, and came closest out of basically everyone in South Carolina.  Mason-Dixon did passably well in Nevada (though the error in margin of victory was outside of the margin of error), but seriously underestimated Obama's victory in South Carolina.

So when you say that you're "not quite sure how this can be concluded without a fair degree of wishful thinking," I certainly think it's fair to say that more explanation is needed than simply saying "it's obvious".

[Filuwaúrdjan]

I did look up the polls.  And honestly, from what I could see, I think people have a point. 

Only because you're looking for them to have a point.

You must be logged in to read this quote.

In the case of Nevada all the silly caucus rules mean that the polls aren't much use (and both company's were pretty close anyway. See also earlier point about Zoggy and caucuses). In the case of SC, all companies got it "wrong" by rather a lot.

You must be logged in to read this quote.

That Zoggy was the worst, by a reasonable amount as well, is significant though. Frankly I don't see why anyone would ever trust any of his polls (when there's a secret ballot anyhow). If he gets it right it might as well be a fluke.

You must be logged in to read this quote.

I already mentioned that Zoggy tends to do well when there is no secret ballot. I'm not sure if "basically completely nailed it" is accurate, but hey, that hardly matters.

You must be logged in to read this quote.

Zoggy's last poll: Obama 41, Clinton 26, Edwards 19. Actual result: Obama 55, Clinton 26, Edwards 17. "Closer" than most, but still way out.

You must be logged in to read this quote.

So did Zoggy.

I should note that you should also look at the polls and results for the Republican primaries. Some important differences between the two sets o/c, but that shouldn't matter as far as polling goes.

EDIT: I should *also* add that as my main concern is for an interesting set of primaries an Obama "lead" of 2pts in a CA poll would be a good thing (as, taking early voters into considerations, it would mean that the state would be extremely close). I'm just sick of the Obamahack Hive Mind which has become so dominant here. Any poll that's good for their candidate is drooled over as if it were political pornography, any poll that's not good for their candidate is dismissed out of hand. In no case does it matter who the pollster is (though at least no one trusts ARG. That's something) and a total lack of consistency is easy to see.

[exopolitician]

Zogby was right on the ball with Republicans in Florida, and they actually underestimated Obama in South Carolina. They were good with predicting Clinton's numbers in SC but not in Florida, cause it actually was closer than expected. I probably take Zogby and Rasmussen more seriously than any other poll really.

[Eraserhead]

Isn't it great that the Obamahacks seem to trust Zoggy more than Mason-Dixon? Why? The answer is obvious and doth not reflect so well...

Please dig a hole and enter it.

[© tweed]

Isn't it great that the Obamahacks seem to trust Zoggy more than Mason-Dixon? Why? The answer is obvious and doth not reflect so well...

[Alcon]

While I think Gabu has a point, that chart's methodology is quite flawed.

[Mr. Morden]

Agreed.  As I pointed out before, SUSA's chart includes polls that were taken two weeks before the primary in question.  Obviously,  if you keep polling up to election day, you're more likely to be right.

[Filuwaúrdjan]

While I think Gabu has a point,

Of course he does. I'm exaggerating a few things a little in order to illustrate something that I find to be extremely irritating.

You must be logged in to read this quote.

Disraeli quote anyone?

[Gabu]

I did look up the polls.  And honestly, from what I could see, I think people have a point. 

Only because you're looking for them to have a point.

What?  I looked them up to respond to your statement.  Isn't that how discussions normally go?  A person makes a statement, and then the evidence is evaluated to determine whether or not the statement has merit.

You must be logged in to read this quote.

In the case of Nevada all the silly caucus rules mean that the polls aren't much use (and both company's were pretty close anyway. See also earlier point about Zoggy and caucuses). In the case of SC, all companies got it "wrong" by rather a lot.

Oh sure, all companies got South Carolina wrong.  But Zogby was most certainly closer than Mason-Dixon.

You must be logged in to read this quote.

That Zoggy was the worst, by a reasonable amount as well, is significant though. Frankly I don't see why anyone would ever trust any of his polls (when there's a secret ballot anyhow). If he gets it right it might as well be a fluke.

Based on the evidence thus far for this primary season's results compared to polls, I quite simply see no reason to assert that Mason-Dixon is more trustworthy than Zogby.  Sure Zogby may have been a total joke in 2004.  But this isn't 2004 and his record has been far better than his record in 2004.  I see no reason why we should blindly trust the past over the present when the two are in conflict.

You must be logged in to read this quote.

I already mentioned that Zoggy tends to do well when there is no secret ballot. I'm not sure if "basically completely nailed it" is accurate, but hey, that hardly matters.

Well, the results were 31-27-24 or thereabouts, with the candidates in the same order as they actually came in.  The proportion between the two candidates is roughly identical to the actual results, although Clinton was a tad underestimated.

You must be logged in to read this quote.

Zoggy's last poll: Obama 41, Clinton 26, Edwards 19. Actual result: Obama 55, Clinton 26, Edwards 17. "Closer" than most, but still way out.

Granted.  But look at Mason-Dixon's last poll from there: Obama 38, Clinton 30, Edwards 19.  Even worse than Zogby's.  I'm not seeing a lot of reason to trust Mason-Dixon over Zogby here.  Now, if your assertion was that all polls ought not to be trusted that much, I would agree, but I would certainly not agree that this supports the idea that Mason-Dixon's polls are more trustworthy than Zogby.

You must be logged in to read this quote.

So did Zoggy.

To a lesser extent.

I should note that you should also look at the polls and results for the Republican primaries. Some important differences between the two sets o/c, but that shouldn't matter as far as polling goes.

I'm not sure what you mean there.

EDIT: I should *also* add that as my main concern is for an interesting set of primaries an Obama "lead" of 2pts in a CA poll would be a good thing (as, taking early voters into considerations, it would mean that the state would be extremely close). I'm just sick of the Obamahack Hive Mind which has become so dominant here. Any poll that's good for their candidate is drooled over as if it were political pornography, any poll that's not good for their candidate is dismissed out of hand. In no case does it matter who the pollster is (though at least no one trusts ARG. That's something) and a total lack of consistency is easy to see.

I'm not asserting any of this because I want it to be true; I'm asserting it because I think it is true based on the evidence that I have seen.  Come on, you should know better than to lump me in with the "hive mind" (which do admittedly exist to some degree).

[Filuwaúrdjan]

Like I said I exaggerated a few things to demonstrate something. I think your main argument is wrong (there's not enough evidence to make many conclusions about pollster accuracy this year yet, meaning that it is legitimate to look at polling from other years to make assumptions about accuracy) though.

I'm not asserting any of this because I want it to be true; I'm asserting it because I think it is true based on the evidence that I have seen.  Come on, you should know better than to lump me in with the "hive mind" (which do admittedly exist to some degree).

I wasn't lumping you in with the hive mind and I'm sorry if it looks like I might be. I actually added that last section to try to make that clear.