2004 Elections: Democratic Party Contenders

They provide a clue about the total. And in this case the clue is that your math doesn't add up. You are taking pockets of data and saying it adds up to showing that Kerry is the worst candidate but when they are really added up it is showing that he's the best by far.

<et cetera et cetera>

This is getting to be very convoluted. Your math doesn't add up now but it'll add up later and the polling data you like is hard data while the rest is snapshots.

Frankly I think that when you extrapolate based on this limited data set and ignore and dismiss larger datasets you do so in error.

Quote:

In all, a quarter of the WI voting age population (VAP) went to the primaries. The REP primary turnout was bizarrely low, suggesting most solid Republicans didnt go - which means that on the part of the VAP that might actually vote Dem, turnout was significantly higher, still. I'd say at least half of the people who might vote Dem in the general elections, came out now.

And I'd still say that you are pulling those calculations out of the air. And stringing them together to contradict what similar data that refutes your position.

1) It was the primary for ONE PARTY nimh. There will, of course be TWO PARTIES in the running for president.

2) Edwards has not been the front runner. The data you use to "confirm" that Kerry is the "worst" candidate could just as easily be data saying that the "independents" tend to side with underdogs.

Quote:

These match-up polls are useful in measuring how the mood turns from or away Bush resp. the main Democratic candidate. But they are more problematic (not useless, just more problematic) in comparing how different Dem candidates perform.

Sez you, again basing this on nothing but your own opinion.

In the primaries polling Edwards and Kerry were against each other. Kerry won almost every single time. So you select even smaller subsets of data to seize upon.

What you don't seem to even be aware of is that it's entirely possible for Person A to come out on top vs Person B while Person B is a better person to run against Person C.

Quote:

So Kerry is shown more "electable" than Dean. But Edwards, we just dont know, cause he hasnt had the chance yet to be projected on in this way.

Then how can you assert that Kerry is the "worst"? LOL

Right. Analysing the specific results concerning key voter groups is moving away from the main truth - and the main truth is that Kerry is "the best by far" - because the national snapshot has him in place one.

According to that same logic, Dean was the most electable Democratic candidate on January 2, and Lieberman was the most electable candidate last August.

I'm just saying that your line of argument here - analysing subsets of polling results is just "resorting" to "smaller slices of negative data" when the overriding truth of the main national snapshot is so blatantly clear - is dilettantist.

Craven de Kere wrote:

This is getting to be very convoluted. Your math doesn't add up now but it'll add up later and the polling data you like is hard data while the rest is snapshots.

Nonsense. That is not at all what I said.

The latter is interesting, but not hard data. The former usually approaches the actual election results within very narrow margins.

Compare the last opinion polls before the WI primaries (Kerry had, what? a nineteen point lead or something?) and the exit polls, which differed only a percent or two from the official results.

That was all of that point. Exit polls are not "just another poll" - they are much more useful and credible as analysis material.

Craven de Kere wrote:

Frankly I think that when you extrapolate based on this limited data set and ignore and dismiss larger datasets you do so in error.

You yourself brought in the S.V.S.S. criterium - swing voters in swing states. It's the EC, remember? Well, here they are. Of course its still a sample with limited use. But its the closest to your criterium we have thus far.

Pulling them out of the air? I've provided all the links, dammit. But OK, sure, I'll list them again for you:

Hey! Guess what. The number of voters in the Dem primaries this week was exactly half of a very royally calculated utter maximum likely number of general election Dem voters. That's pretty much exactly what I said, isn't it? Shoddy work and "stringing together calculations that I'm pulling out of the air", my a**. <mutters>

1) It was the primary for ONE PARTY nimh. There will, of course be TWO PARTIES in the running for president.

See calculation above.

What about white men, suburbians, last-minute deciders, people who vote on the basis of whom they "agree with on the issues", and voters who are "satisfied, but not enthusiastic" about the incumbent president?

I based my case - that, 1) Kerry has a problem in appealing to the exact segments of voters a Dem victory will hinge on,

2) it will become worse if and when the bubble of being the most electable, which his turnout now hugely depends on, collapses, and he's judged on the issues again - on a range of indicators, not just the Independents.

You have to be pretty controversial (e.g. Dean) to largely miss out on this 'leader of the pack bonus'. I submit that we see the same now with Kerry vs the still-more-unknown Edwards.

Craven de Kere wrote:

In the primaries polling Edwards and Kerry were against each other. Kerry won almost every single time. So you select even smaller subsets of data to seize upon.

Platitude: primaries are no general elections. You can win every single Dem primary by appealing to your core party support, and lose the general elections big time, exactly because your appeal is to your core party support.

Its the extra people who will make the difference - the floating voters or, OK, the non-voters you manage to pull in (though thats depressingly hard). So it makes sense, when analysing primary results, to focus on whatever indications you can get on how they will act - on those "smaller subsets", yes. <shrugs>

The higher the turnout in a primary, the more it indicates how a candidate might play out in the general elections. The more non-party faithful take part, the more it indicates how etc. The more the state is a "bellweather state", the more it indicates how etc. Swing voters in swing states. (That was a brilliant line, btw). All of those criteria suggest WI is not just any random "subset". <shrugs again>

Craven de Kere wrote:

What you don't seem to even be aware of is that it's entirely possible for Person A to come out on top vs Person B while Person B is a better person to run against Person C.

Ehm ... <big grin> ... thats the exact case I've been making, Craven.

Kerry has, until now, been resoundingly coming out on top of Edwards et al. Yet I submit that Edwards (et al.) would be a better person to run against Bush. <shrugs>

'K, gotta stop for now, gotta go.

Quote:

Craven de Kere wrote:

the polling data you like is hard data while the rest is snapshots.

Nonsense. That is not at all what I said.

I disagree. I think that summarizes what you said.

Quote:

The latter is interesting, but not hard data. The former usually approaches the actual election results within very narrow margins.

No it doesn't. Actual election results are usually very different from a subset of one party's primary in a subset of the nation.

What do you think will be the main issue these elections? Who does better on it, Kerry or Edwards?
Which comes closest to your feelings about the Bush administration?
What do you think will be the main issue these elections? Who does better on it, Kerry or Edwards?
Did you vote for your candidate today more because you think...

Quote:

Hey! Guess what. The number of voters in the Dem primaries this week was exactly half of a very royally calculated utter maximum likely number of general election Dem voters. That's pretty much exactly what I said, isn't it?

Incorrect, you do not have any safe way to base your assumption that your calculations on "people who might vote Dem" is correct [..] So citing the statistics that you use on your way to assuming how many people "might vote Dem" does absolutely nothing to support your assumption about how many will vote Dem.

What that means, though, is that the participants in this week's primaries represented an even greater share of those who will vote Democratic in the presidentials - not just half of them, but considerably more.

Lookit, I don't think that I have "magically isolated the very exact people that this whole election hinges on" - all I have said is that, in terms of indications we have to go on, the WI exit polls (and before that, the MO exit polls) are the best we have.

"Isolating" the people an election hinges on is a question of speculating on the basis of prior election results. Soccer moms. NASCAR dads. You know the game.

You are not going to get those specifications out of a general match-up number (Kerry 52, Bush 43).

Those "day ratings" will still fluctuate wildly up and down. They show whether the trend is up or down - which is important. But this far in advance, they have little predictory value. But you can say that, if a candidate has significant problems in appealing to electorate groups that are commonly referred to as the ones that tilt the balance (soccer mums, NASCAR dads, Reagan Democrats, Latinos), he's going to get into trouble in the general elections if the race does get close.

One, opinion polls that go into that much detail. Problem: they're opinion polls, and merely poll nebulous voting intentions. Two: exit polls on elections that actually took place - state primaries. Problem: though these are actual voting decisions and not just nebulous intentions, the sample represents just the one state - and only those who turn out for the primaries.

But it is nonsense to say that all these punters are "wildly" stringing together assumptions when they point out that the numbers do suggest these problems are there.

Again, you do not know whether they will vote Dem or not. This is assumption. You assume that all said independents will vote Dem for e.g.

Is it a reasonable assumption to say that the Democrats will not likely score over 10% better than their best performance in the past 20 years, and that turnout will not - or at the very least not simultaneously - be higher than it's been in the past 40 years?

I've never said those "punters" are "wildly" stringing together assumptions.

I said you are. Examine the difference between what you say and what they are saying.

You say Kerry is "the worst". They, as far as I know, do not.

Sure there are "problems". There are always problems. You took it a step further to say that Kerry is the "worst"

-#1-
My case that Kerry is the worst the Dems could plausibly have picked was based on estimations of how his personality and style [etc etc] would play out against Bush.

All personal opinion / estimation.

I've since been observing how this estimation is reflected (and sometimes disproven, as in the case of his unexpected blue-collar support) in the specifics of polls and election results.

Much of those suggest confirmation for my case for Edwards over Kerry. None of them suggested confirmation for my case for Clark over Kerry - so I'm not so sure about that one anymore. Gephardt was already out, so - we'll never know. My rather hesitant case that even Dean might possibly have been better was never based on polls (which clearly argued against him)

-#2-
Finally, let me point out that I have made no arguments on the basis of polls alone, here. In this thread and the "lies and foibles of John Kerry" thread, you'll find that I'm consistently using polls to doublecheck, illustrate and confirm - and at times, [..] disprove - my independently argued theories about why Kerry is the worse candidate.

Why is Kerry more electable than Edwards (or Clark, or Gephardt, for that matter), 'ccording to y'all? [..] though the polls themselves should be relativated, they provide plenty of leads that confirm otherwise argued concerns about Kerry, and those at least, deserve to be addressed.

nimh wrote:

Is it a reasonable assumption to say that the Democrats will not likely score over 10% better than their best performance in the past 20 years, and that turnout will not - or at the very least not simultaneously - be higher than it's been in the past 40 years?

Perhaps.

I've said that individually your assumtions may be reasonable, but that stringing them together is not (if there are multiple points of failure).

the point was that it was yet another assumption. As reasonable as it may be it's a possible point of failure to the conclusion if it's considered substatiation for it.

Hmm. So the punters are not wildly stringing together assumptions, but I am, when I call Kerry the "worst".

I.e., I'm suggesting that the polls sometimes "illustrate or confirm" (and sometimes disprove) my otherwise argued "theories" on "why Kerry is the worse candidate". I'm saying they "confirm otherwise argued concerns about Kerry".

So how and where are they OK, but am I "wildly stringing together assumptions"?

<shrugs> Any kind of non-beta-science theory involves series of formally unprovable, but generally considered assumable parameters.

Political theories start out from assumptions based on historical track records etc. They are not geared for sudden flukes - they sketch scenarios on the basis of what can be reasonably expected.

I'm suggesting that you have to move on at some point from raw numbers to the chess game of individual states or regions.

Correct. As far as I know, the "punters" you reference are not stringing assumptions together to support the previously stated notion that Kerry is the worst candidate for the Dems.

I do not see any relevance whatsoever in regard to what the "punters" say. <shrugs>

I don't have time to list all the assumptions in the series

1) The election will hinge on "swing voters".

Yep

2) The election will hinge on "swing states"

Yep

3) The popularity of the candidates is not the main reason for the expressed opinions in the polls.

In matchup polls, frontrunner status is as much a function of name recognition/"winner" image as of anything else

4) The exit polls samples up to half of the "people who will vote Dem".

The Wisconsin exit poll accurately reflected the results of the Wisconsin primary - in which over half of the "people who will vote Dem" there took part.

5) The exit polls thus contained a statistically significant data sampling.

The exit polls thus says significant enough things about a sizable body of swing votes in Wisconsin. Which is relevant to the extent that Wisconsin is a swing state and, in terms of past presidential elections, a MOR American state.

6) The exit polls contain data that is sound to extrapolate.
a) The fact that the data is on Democrats is not important.

The data is not on Democrats - it was an open primary. Only 62% of those who voted in the Dem primaries were Democrats

b) The fact that people who vote in primaries are the most politically motivated and thus atypical is not important.

Turnout was huge enough to make this primary's electorate less atypical than any other's. See 4).

7) That the dynamics of a Dem vs. Dem campaign is not sufficiently different from a Dem vs. Bush campaign to render most of the data you cite irrelevant to a Dem vs. Bush matchup.

Yep. If a candidate does badly outside the Dem hardcore in primaries, he can be reasonably expected to have problems against the Rep opponent, too. Example: Dean.

8) That the unknowns about Edwards carry no surprises (otherwise you'd not be able to assert that Kerry7 is the worst, you'd have to wait for said revalations).

Huh?

I explicitly and repeatedly agreed on what an Edwards' frontrunnerdom would bring in terms of increased scrutiny. So dont even go baiting me there.

I have repeatedly made the case that Kerry is a worse risk, cause of his long exploitable track record. Open door, that.

9) That "independents" who vote at party primaries represent the bulk of the "swing voters".

Nope. Never said that either. Not "the bulk of". But swing voters, yes, many or most of them would be.

I have to leave the office so I need to cut this short. But if you know of any other "punters" who are making the same series of assumptions

Strip away your colorful additions and changes to the submissions I actually made, and you get a smorgasbord of pretty commonplace points ...

That surely was of an irresponsibility worthy of the dozen-post indignant Craven wrath.

<puzzled & pissed>

Enough poll talk......

Let's find some polls from those states voting next week and see what chance he's got...