Eleventy bazillion? Hey! What's your KND number?
;-)

Mitch whined,

"Or it may have been influenced by the fact that the Minnesota Poll is at best worthless"

I assume you have a copy of the questionaire so you can disect the methodology and demonstrate a bias right Mitch? Or you can somehow prove that the people conducting the interviews were intestionally falsifying the responses?

oops... Intentionally

Doug, you are as obtuse as ever.

When EVERY SINGLE MN POLL IN TWO DECADES (with one exception) shows a 5-12 point gap in the DFL's favor in the month before the election (which *just happens* to be the home stretch for voter turnout and final fundraising, whaddyasupposetheoddswere), which always vanishes on election day, then the burden of proof is on the Strib. Not on me, the news consumer.

I don't give a rat's ass about their methodology, or your obedient little rearguard action to support your obvious ideological masters. The pattern is clear and obvious, and the burden of proof lies squarely at the corner of Fourth and Portland.

"Lies" is an excellent word choice, in this case.

Mitch said,

"the burden of proof lies squarely at the corner of Fourth and Portland."

Dude, you're seriously losing it.

Republicans have boasted for the last, well since I can remember, about just how successful their last minute "get out the vote" efforts are. Would that explain your 5 - 12 point gap Mitch?

What else... In 2000 and in 2004, dozens of exit polls showed democrats leading in key races (including the Presidential) around the country and mysteriously, the official winners were, Hmmmm... Republican. You guys chalked that up to over-zealous democrats who took time to respond to the polsters. Maybe we're just over-zealous again Mitch and actually taking the time to do the survey. Is that good for a few points Mitch?

Of course there's always the possibility that the poll results are correct and you're just a bias pimp?

There is a simple solution to this. Bachmann can release one of her internal polls that shows the race to be closer. Her failure to release such a poll in the next few days will prove that her results track the Strib poll.

Just for the record. The final Strib poll in the 2005 St. Paul mayor's race was 61% Coleman / 26% Kelly / 13% Undecided. Final result 69% Coleman / 26% Kelly. Hardly a skewed result.

My fault:
that should be final result 69 coleman 31 kelly.

The error the MN poll has is oversampling Dems. In St. Paul, you can not oversample Dems.

"Dude, you're seriously losing it."

Doug, you can make the choice now - argue rationally, or sling ofay little insults.

Saying "you're losing it" implies you have some edge in "sanity", to say nothing of fact. Neither your facts nor your delivery evince this.

So - if you want to be treated as a name-calling jagoff, that's fine. If you want to be treated as someone whose input is worthy of a considered response, lose the condesencsion.

Since you're a lefty commenter, I'll assume the former until proven wrong.

"Republicans have boasted for the last, well since I can remember, about just how successful their last minute "get out the vote" efforts are. Would that explain your 5 - 12 point gap Mitch?"

Non-sequitur. GOTV efforts don't - are not designed to - swing the undecided. They are designed to make sure every GOP-leaning voter GETS TO THE POLLS. That implies that they are GOP-leaning in the first place. If they are GOP-leaning but *never* show up in the MN Poll, then that is neither the GOP nor the voters' fault; it's the fault of the Strib and its constant oversampling.

And the fact - and the timing - of the (consistently-one-sided) inaccuracies would make any rational person wonder after a while, no?

"What else... In 2000 and in 2004, dozens of exit polls...

Bzzzzzt. You really want to be considered a nutbar, don't you?

Exit Polls are scientifically utterly invalid. Their sampling controls are notoriously inconsistent. Even the population doing the actual sampling is, itself, skewed!

Discussion of exit polls as serious, empirical, dispositive fact will earn you nothing but ridicule on this site. Someday, when society is more enlightened, comparing "exit polls" to election results will be accorded the same disgust as foot-binding, clitorectomy and Klansmanship.

"Of course there's always the possibility that the poll results are correct and you're just a bias pimp?"

Well, your mind *was* made up before the first of your 20,000 comments, so that was never really in doubt, was it?

"In St. Paul, you can not oversample Dems."

As a matter of math you are wrong. It is no harder or easier to oversample Democrats in a population 65% DFL than it is to oversample Democrats in a population 35% DFL.

"It is no harder or easier to oversample Democrats in a population 65% DFL than it is to oversample Democrats in a population 35% DFL."

Really?

Are you saying that if I ran a poll with 1000 respondents in an area containing two to three million residents, I couldn't skew the respondent pool?

Paul:
"Are you saying that if I ran a poll with 1000 respondents in an area containing two to three million residents, I couldn't skew the respondent pool"

If I wanted to say that I would have said it. I said the 2005 poll didn't seem to have a skewed pool.

You can skew the pool of any poll. I pointed out that the 2005 Minnesota poll in the St. Paul mayor's race was accurate. Bill suggested that the poll could not have oversampled Republicans, presumably because St. Paul is heavily DFL. I pointed out that as a matter of math the DFL% in the pool has little imppact on a pollsters ability to skew the result.

You are right, the MN Poll could be skewed (hell Madonna could be a guy), the question is what evidence is there that it is skewed. Since many people like to point out the bad results in 2002 and 2000, I thought it usefull to mention the accurate 2005 results.

Interestingly the 2001 MN poll of the St. Paul mayor's race had Kelly up 6% when Kelly won 50.1% to 49.9%

Where to start...

"Non-sequitur. GOTV efforts don't - are not designed to - swing the undecided."

You're wrong Mitch. GOTV efforts have two components. To identify and to motivate. For example, anyone who identifies as undecided, lean Democratic or lean Republican go into the list to canvass.

Those that are undecided or lean lean left or right two weeks before an election, (the people who might appear in a minnesota poll), are usually decided the day of the election. If your network of volunteers are doing their job they will have turned a leaning voter into a strong.

Next...

"Discussion of exit polls as serious, empirical, dispositive fact will earn you nothing but ridicule on this site"

Hmmm... It appears that I never said exit polls were serious, empirical, dispositive facts.

I said that you fellows had a quick explanation for why the exit polls didn't match the results. Remeber Mitch? It was over-zealous voters.

Now, the Minnesota poll says something that you don't agree with and it must be the notoriously biased Star Tribune. Maybe it was over-zealous respondents...

Also, I offered a simple opportunity for you to validate your points. Here it is again...

"I assume you have a copy of the questionaire so you can disect the methodology and demonstrate a bias right Mitch? "

You respond with...

"I don't give a rat's ass about their methodology"

Ohhh... Them's powerful words Mitch. Ironically, in a new post, you cite King who does EXACTLY what I said and references the methodology used and low and behold, he actually proposes a rational explanation for the poll results.

Sorry Mitch but someone on you own side made you look like a reactionary fool.

And finally Mitch, regarding pre-election polls and exit polling. When they are within a couple of percentage points of agreement but the actual tallies are WAY off... you're going to have to start doing better than "Their sampling controls are notoriously inconsistent."

"Ohhh... Them's powerful words Mitch. Ironically, in a new post, you cite King who does EXACTLY what I said and references the methodology used and low and behold, he actually proposes a rational explanation for the poll results."

Right. And King shows that my impressions are *exactly correct*.

So - your complaint is *what*? That I'm right (again), but that you'd rather futz with numbers than deal with the overarching ethical issue?

"Since many people like to point out the bad results in 2002 and 2000"

And 2004, and 1998, and 1996, and 1994, and 1990...

Mitch said,

And King shows that my impressions are *exactly correct*.

Really. King showed that, in addition to the reasons cited by the Trib, there is an alternative demonstrable explanation for the results of the poll.

And, in addition to the Tribs explanation and Kings explanation, there is a half dozen other likely explanations INCLUDING the possibility that Bachmann is behind in the polls.

Of course for you that just couldn't possibly be.

The funny thing about idealogues like you Mitch is that you never see yourself objectively. What you can't see Mitch is that you're really just Al Sharpton without the bad jewelry.

"And, in addition to the Tribs explanation and Kings explanation, there is a half dozen other likely explanations INCLUDING the possibility that Bachmann is behind in the polls."

I never ruled that out. I DO say that it's unlikely, for a variety of reasons, and cite the MNPoll's lunatic past.

"Of course for you that just couldn't possibly be."

As usual - and by "usual" I mean "almost always" - you are wrong.

"The funny thing about idealogues like you Mitch is that you never see yourself objectively."

Ah. But if I just saw myself the way you do, all would be well.

Bullshit, Doug. Yet again.

Mitch:
"And 2004, and 1998, and 1996, and 1994, and 1990."

Do you have the results for 96, 94, and 90? Please post. I bet a Guiness the 90 result picked Boshwitz to win.

Just for the record.
2005. Accurate St. Paul poll.
2004. Pro-DFL poll.
2002. Pro-DFL poll in context of extra-ordinary. events surrounding Wellstone crash/funneral.
2001. Pro-GOP poll in St. Paul.
2000. Pro-DFL poll in State race.
1998. Pro-DFL poll in context of extra-ordinary Ventura surge.

So far, in ordinary circumstances, I count 1 accurate, 1 pro-GOP and 2 pro-DFL polls.

Let me finish this for you, Rick:

"2005. Accurate St. Paul poll." - there was no reason to be *innaccurate*, Coleman won in a landslide. There was nothing to game."

"2004. Pro-DFL poll.
2002. Pro-DFL poll in context of extra-ordinary. events surrounding Wellstone crash/funneral."

Extraordinary events, sure - but the last pre-election poll showed Mondale with a solid lead that didn't exist..

2002 - MNPoll right before the election shows Roger Moe beating Tim Pawlenty by a slim but convincing margin.

"2001. Pro-GOP poll in St. Paul." - Not sure what you mean, here - Jerry Blakey went out during the primaries.

"2000. Pro-DFL poll in State race."

Yep. The Strib Poll didn't NEED to do the dirty work - the News and Editorial departments took care of Rod Grams.

"1998. Pro-DFL poll in context of extra-ordinary Ventura surge."

Right - which showed Skip Humprey with a convincing lead in the final poll before the election.

1994 - MNPoll showed Ann Wynia with a solid lead over Grams.