I read the headline, and read the raw figures and thought, it just keeps getting worse for Obama.

Then I looked inside, and realized whoever wrote the title might have missed something significant----the 33 point lead from 9-30 is now 23.

I also noticed that it is possible for the number of people who say they support a candidate strongly---to actually go down.

Since July, Obama's strong support dropped from 56 to 50. Hilaary's dropped from 68 to 57.

Hillary may still have a lead, but she dropped 4 points while Obama gained 6. Maybe your headline should be Clinton's lead narrows by 10 points. That would be a more accurate reflection of the numbers.

Oh, and to further underline the point, Eric even noted McCain was up 7 points since the last Post/ABC poll. But Obama's 6 point rise by the same measure? No mention of it by Eric.

The debate was not a disaster or even a major blunder for the Clinton campaign as far as national polls are concerned. Some people are really overstating the importance of what happened at the last debate. She's going to be the nominee.

Looks to me like Hillary dropped a bit and Obama went up. Tells me that the debate questions had a slight effect after all. Perhaps if this fact of flip/flopping gets out there more....Democrats just might end up walking away from Hillary. All they need is another (what seems like) flipper.

kjoe wrote on November 4, 2007 9:27 PM:

I read the headline, and read the raw figures and thought, it just keeps getting worse for Obama.

Then I looked inside, and realized whoever wrote the title might have missed something significant----the 33 point lead from 9-30 is now 23.

I also noticed that it is possible for the number of people who say they support a candidate strongly---to actually go down.

Since July, Obama's strong support dropped from 56 to 50. Hilaary's dropped from 68 to 57.

I am sorry to rain on this parade but now that you finally believe in polls this "early", I see little for Obama and his supporters to rejoice about in this poll, which is really just a correction for its last numbers that few other polls could reproduce (Mark Blumenthal @ pollster.com has maintained all along that that poll was a clear outlier). The numbers just published are more inline with the trends that have emerged over the past couple of weeks, in fact, the past month. I had written about it last night elsewhere: Clinton is at mid- to high 40s and Obama at low to mid 20s, which is consistent with the latest WaPo/ABC.

The overall picture however, is one in which Clinton bas been gaining ground at the expense of Obama, with Edwards just a spectator 12%.

Virtually everyone seems to be showing about the same lead for HRC, so, therefore, there is nothing for her supporters to worry about unless you think a 23% lead is worrisome (a strategy would be for her detractors to raise the expectation that unless she is leading by 30%, she cannot win, which seems to be what many are trying to do...Good luck).

dcshungu,

This is now the third time that you have uttered that talking point about Blumenthal viewing that 9/30 poll as an outlier, and I have already pointed out to you that he reassessed his outlier claim in a 10/16 post.

Frankly, you should be ashamed of yourself. It is one thing to spin on your own behalf. It is another to misrepresent the views of another person to aid your spin.

Two points need to be made. First, Dschungu is only partially right. He/she/it is right that Hillary's 23 point lead (down from 33) is hardly a cause to hoot and holler if you're an Obama supporter. However, there were two polls last month (including this very same Post poll) that had Hillary 30 plus points ahead, and the coverage of those was breathless, including on this esteemed site. Had the initial coverage of those been downplayed and caveated, then Eric would be right to glide by the ten point change, but alas, the coverage of those earlier polls was hyped to the max. So the subtle bias of Election Central in the way it covers polls sadly continues. (Another example: immediate play on EC to a two day partial Rasmussen poll after the debate that Rasmussen himself said was not statistically signifcant or terribly relevant, but no play to a Rasmussen poll that showed that Hillary has trouble getting more than 48% against any Republican, including Romney and even Ron Paul, whereas Edwards and Obama cream all Repubs other than Giuliani (against whom they poll similar to Hillary).

Second. The poll that is the immediate subject of discussion on this thread was not taken entirely after the debate, and the poll did not ask whether respondents watched or saw coverage of the debates. The more seasoned pollsters will tell you that these primary debates are watched by very few, and that they move public opinion only with a substantial time lag as the press digests what happened and opinion leaders render their verdict over a period of days.

I'm cautiously optimistic that this lag effect will put Obama at least a little closer by the middle to end of this week, because the aftermath of this debate has been a bit odd. The morning after the debate, there was sort of a split verdict: some pundits thought Hillary did quite fine albeit falling short of her previously high standard, and others thought Hillary clearly lost. Then Hillary's strategists, instead of spinning it as "there have been seven debates; Hillary has won six and tied one, let's move on," could not stand the idea of losing even one news cycle so they came up with their ridiculous excuses. The excuses extended the story of Hillary's marginal performance for days and well into the Sunday talk shows. And Obama's people handled the post-debate analysis far better than Hillary and better than Edwards as well, so though even many Obama supporters thought Edwards outperformed Obama on the night of the debate, the meme started to emerge that Obama won the debate and the week. So let's see what happens next.

Hillary is down in every poll that has late September/early October history. Penn's internals are clearly showing the same thing. It tends to explain both, why she was a little rattled at the last debate why they started the "all those mean men are picking on Hillary" victim card even before the debate began. One might speculate that they're now desparately afraid that she in the first week of October.

But that's okay. you Hillaroids just keep telling yourself how it doesn't mean anything and she's still invincible and inevitable. She's still running a perfect campaign and shouldn't change a thing, not a thing. (Although little brie to go with all that whine she's uncorked recently might be nice.)

Last month's Post/ABC poll was a statistical outlier, as HRC's jump in a couple of polls came out of the blue and did not jive with the rest of the trends, which caused Mark Blumenthal of pollster.com to make a comment that now appears to be eerily prescient in the light of today's Post/ABC poll that is more inline with the trend:

Finally, about my speculation twelve days ago:

Of course, we do not yet know whether this poll is really a statistical outlier. Other polls have obviously been showing a more gradual increase in her support recently, and it is still theoretically possible that Clinton's support suddenly lurched up ten points last week...

So we will wait and see. But I'll wager that a month from now the real trend will not look nearly as dramatic as the one suggested by yesterday's news.

A month has not yet passed and the trend line for September might still change, but I stand by my wager.

Emphasis mine. So, let stop the hyperventilation. If you'd like to have a sense for the campaign dynamics, study the aggregate polls: For about a month now HRC has been leading by ~20% and gaining, while Obama had either stalled or was losing ground. Edwards' numbers have not moved at all, despite his increased aggression. He is done unless he wins Iowa...but even then...

LOL.

DTM wrote on November 4, 2007 11:51 PM:

dcshungu,

I have already corrected your misrepresentations of Blumenthal elsewhere.

Dear DTM: People can read English, I think. What Blumenthal had written in Standard American English (see my preceding post and follow the link) speaks for itself. You accused me of lacking "intellectual honesty", please show some yourself...right about now.

Gotta go.

Keith, agreed. Was Ferraro always this pathetic? I'm retrospectively embarrassed.

Whatever happens, the Clinton campaign can't fault Kleefeld for effort.

But independents and undecideds would have to be fools to trust Kleefeld or TPM Election Central for complete polling/campaign coverage.

POLLING NEWS YOU WON'T FIND POSTED AT TPM ELECTION CLINTON

Quinnipiac poll results from October 23-29:

October 31, 2007 - Treat For Giuliani As He Inches Ahead Of Clinton, Quinnipiac University National Poll Finds; Trick For Clinton As She Drops Overall

Former New York City Mayor Rudolph Giuliani, the Republican front-runner, leads New York Sen. Hillary Clinton by a razor-thin 45 - 43 percent, [even though] voters say 58 - 37 percent that President George W. Bush's low approval ratings will make it difficult for any Republican to be elected President, according to a Quinnipiac University national poll released today.

Giuliani's lead reverses a 46 - 43 percent Sen. Clinton lead in an August 15 poll by the independent Quinnipiac (KWIN-uh-pe-ack) University, but Giuliani was up 49 - 40 percent May 3.

Illinois Sen. Barack Obama gets 43 percent to Giuliani's 42 percent [while] Arizona Sen. John McCain ties Clinton 44 - 44 percent, wiping out a 47 - 41 percent Clinton lead August 15.

Clinton also has seen her favorability among American voters drop from 48 - 43 percent August 15 to 46 - 46 percent today.

In other head-to-head matchups in the current survey:

* Clinton leads former Tennessee Sen. Fred Thompson 46 - 41 percent and former Massachusetts Gov. Mitt Romney 48 - 38 percent, with both leads down from August 15;

* Obama ties McCain 43 - 43 percent and beats Thompson 45 - 37 percent and Romney 46 - 36 percent;

* Former North Carolina Sen. John Edwards trails Giuliani 44 - 41 percent; ties McCain 42 - 42 percent, but beats Thompson 46 - 36 percent and Romney 47 - 34 percent.

TPM and Election Central are losing credibility because they SPIN what they are reporting, always in favor of Hillary. IF we want the Hillary spin on all the election news, we can just go to her website. Nothing in EC item on the WashPO/ABC poll about Hillary losing ground or her growing "negatives." Yet when Hillary gains a smidgen in a poll of dogcatchers in Crossroads, Kansas, we get a banner.

dcshungu,

Again, I have already documented how you misrepresented Blumenthal, and the fsct that you continue to do so does indeed demonstrate your lack of intellectual honesty. And yes, Blumenthal was quite clear--quite clear in stating that after more polls came out, he was no longer asserting the 9/30 poll was an outlier.

But I know better than to expect you to admit you made a mistake (I suspect that originally you were simply unaware of Blumenthal's 10/16 post, but for whatever reason are unable to admit as much). My only goal was to get you to stop misrepresenting Blumenthal, and since you are fleeing this conversation I take it my goal has been accomplished.

By the way, I also know that you know that I know exactly what that graph you linked shows and why it is different than this graph:

http://www.realclearpolitics.com/epolls/2008/president/us/democratic_presidential_nomination-191.html

As we discussed before, these graphs are not actually inconsistent (although as an aside, pollster has not yet incorporated the latest polls) because of the difference in methodology.

What I was objecting to is that you are going beyond noting that pollster's methodology has not yet confirmed a legitimate trend to claiming with certitude that the trend is still upwards. That is false: the whole point of the pollster methodology is to try to eliminate trends which are the product of random fluctuation. But that doesn't mean there are no trends at all to eliminate on that basis, and in this case we know there is such a trend--the one reflected in the realclearpolitics graph.

In fact, if you really understood Blumenthal's 10/16 post, you would understand that given their methodology, the reported trend in the past can change based on new polls. So at some point in the future, pollster may report a downward trend for Clinton in October, even though the new information would be from November or later. Or maybe not--we just don't know yet what pollster's final analysis for October will be.

Frankly, I don't see why you need to spin all this in such a dishonest way. Do you really think Clinton will never experience a downward trend before Iowa? Aren't you just setting her up for a fall by working so hard to deny such a thing may have happened? Doesn't that just suggest if it does happen, it is somehow devastating news for Clinton?

Anyway, how you want to spin things for Clinton is your business. I just don't see how you think you are helping her with these tactics, but that is true of much of what you do here.

party-of-one,

It is also worth looking at the overall favorable/unfavorable scorecard, which is reported under Question 7 here:

http://www.washingtonpost.com/wp-srv/politics/polls/postpoll_110407.html

One of the things to note is the number of "no opinions". For example, while it is true that no one has over 51% favorable, that is not really an inherent problem for people with a relatively large number of "no opinions" and a good favorable/unfavorable balance. For example, Edwards is at 49% favorable, but he also only has 35% unfavorable, with 16% no opinion. This is a bit crude, but if we imputed those no opinions at the same ratio as Edwards' current fav/unfav, he would end up somewhere around 58% favorable and 42% unfavorable, which is an acceptable balance for a presidential candidate at this point in the process.

And for people with really high no opinions, such as Thompson, Romney, and particularly Huckabee, even that approach is misleading, because pollsters know that there are a decent number of people who tend to report an unfavorable opinion instead of admitting they do not know enough about the person to form an opinion. So, it is often the case that people with such numbers will improve their favorability balance as they get better known.

Accordingly, really there are only a couple people on that chart who appear to be heading toward (or already at) dangerously low favorables and high unfavorables: McCain and Clinton. All three of Giuliani, Edwards, and Obama currently have pretty good numbers. And with Thompson, Romney, and Huckabee, in light of their high no opinion counts it is simply too early to tell.

One final note on McCain and Clinton: I think it is interesting to compare their numbers in depth. Note that Clinton has both relatively high strongly favorables and relatively high strongly unfavorables, as well as the lowest no opinion count on the chart. In comparison, McCain's favorables and unfavorables are both more concentrated in the "somewhat" category, and he has more no opinions. I think that reflects the fact that people have pretty hard opinions about Clinton, but with McCain they are still on the fence. So, I wouldn't rule out McCain becoming more popular (or unpopular, of course) quite yet. But he certainly has some work to do to get to an acceptable balance.

DTM wrote on November 5, 2007 7:26 AM:

dcshungu,

Again, I have already documented how you misrepresented Blumenthal, and the fsct that you continue to do so does indeed demonstrate your lack of intellectual honesty. And yes, Blumenthal was quite clear--quite clear in stating that after more polls came out, he was no longer asserting the 9/30 poll was an outlier.

This is silly high-school debating tactic that you use on a constant basis. Just because you challenge something by providing false or irrelevant "evidence" and keep repeating it does not constitute a "documentation". What is it about Blumenthal's language that you did not understand that you must keep repeating this moronic line?

Let me take this one very, very slowly so that you might finally understand or be shamed into admitting the obvious.

After last month's Post/ABC poll initially came out and it showed HRC with a 33% lead, Blumenthal was skeptical of this sudden surge in her support, prompting him to say that he suspected that it was an outlier. Then there was discussion when two other polls seemed to show HRC with a similar lead. After much discussion that even involved ABC sending him their raw number, Blumenthal stuck to his guns that the jump in HRC's support was an anomaly. If you do not understand the following annotated comment by Blumenthal, please ask someone else to explain it to you, because I am through trying:

Finally, about my speculation twelve days ago:

Of course, we do not yet know whether this poll is really a statistical outlier. Other polls have obviously been showing a more gradual increase in her support recently, and it is still theoretically possible that Clinton's support suddenly lurched up ten points last week . [note: statisticians are not dogmatic fellows, and usually recognize that polling is not an exact "science", therefore there is always caveat in what they say... call it their MOE. The preceding statement is equivalent to Blumenthal providing a MOE to what he really believes, which I have emphasized below]

So we will wait and see [note: caveat or MOE]. But I'll wager that a month from now the real trend will not look nearly as dramatic as the one suggested by yesterday's news. [This is his view: He is saying that he does not believe that the trend is as "dramatic" as those few polls had suggested]

A month has not yet passed and the trend line for September might still change, but I stand by my wager.

Note how he was eerily prescient as the Post/ABC poll has just done exactly as he had suspected. So: What does "I stand by my wager" mean in English and what does the wager refer to in this context? Remember that this is a public forum and that your credibility is at stake here.

What you have "documented" is either that you do not understand simple Standard American English or that you are intellectually dishonest. I wager that it is the latter but it is more likely that it is both.

Good day.

dschungu,

I will document this one more time, since I have not yet in this thread.

First, note this line from what you posted:

"we do not yet know whether this poll is really a statistical outlier."

Second, note this sentence from the same 10/16 post by Blumenthal that you conveniently omitted:

"The 53% result for Clinton on the ABC/Post poll remains on the high end of support for Clinton compared to the trend line, but is no longer 'more of an outlier' than various other polls conducted earlier in the year."

And finally, here is what you originally claimed:

"Mark Blumenthal@pollster.com has maintained all along that the poll was a clear outlier"

That statement of yours is impossible to reconcile with the two statements from Blumenthal that I just quoted.

So there you go: your error documented in plain English, as you requested.

By the way, it did not escape my notice that in your most recent posts, you changed your claim. Now you say: "Blumenthal stuck to his guns that the jump in HRC's support was an anomaly." So, you have replaced your original term ("clear outlier") with a new term ("anomaly").

Of course, Blumenthal did not use the term "anomaly" either, but I guess you think it helps your defense to stop using the phrase "outlier" when Blumenthal specifically addressed that term. But I will just take your attempt to change your claim as an implied concession of your original error.