INTERNAL POLL DUMP (PUT ALL INTERNAL HOUSE POLLS HERE)

[Meeker]

MORE INTERNAL POLLING GARBAGE PLZ

[Torie]

Campaign Spot has put up an internal GOP poll for CT-5 showing Murphy (D) ahead by 1%, 40%-39%.

I'm curious what goes on in your brain when you read a poll like that (other than a burst of happy surprise.) What adjustments do you make to the numbers for the final expected outcome, if any? Do you see the pendulum swinging further Republican in the next two months? What adjustments should Democrats make to Democratic internal polls, ig any?

I fit them into a matrix in my mind, compare them to other polls in other races, and am fully prepared to assume that they have a party bias, particularly if they seem out of line. The context of the internal matters too. Having a Dem poll dump like the above with numbers generally way out of line, suggests to me that it is quite likely that they are fund raising, or Pelosi triage insulating, spam or on the order thereof.

I don't take the CT-5 result too seriously. The pubbie is probably down by at least 5% in reality, and probably closer to 10%.  I discount it in part because it was released by the GOP candidate. It is also out of line with my vision of the current size of the wave; the elevation of the CD is a bit too high to be in that much danger of getting wet. Heck, it is not even one of the 113 districts that RCP has put up as being in some play, so maybe my use of the word "bit" should be changed to "considerably."  I am unaware of anything to adjust for individual candidate issues, in any event.

Oh, and if an internal polling outfit puts something out, which is then approximated by a respectable poll, and I remember the outfit, I then take their numbers a bit more seriously.

Another problem is data mining. It may be some campaign took 10 internal polls, and maybe stopped half of them in midstream when the numbers looked bad, and then finished one up when the numbers were looking much better (or finished 4 up, and then picked the best one to release). In that regard, it helps to keep the sample size down. Addendum: Notice that the Dems sat on this batch of polls for a week, and then dropped them against the generic bad news, and the triage story. I wonder how many internals the Dems did?  Did they just pick the best few out of 50?  That is another way to data mine.  Someone should ask them this question.

So, as usual, context is everything.

I think the odds are about 3-2 that the pendulum will swing back to the Dems somewhat, as voters near the end regress some towards prior voting patterns, and all that money the Dems spends, gets traction in some races. But it won't be enough for the Dems to avoid losing the House and probably losing around 50 seats. (I have the no swing number in my mind at 65 seats at the moment.)  And there is some chance the pendulum will swing a bit more against the Dems, or the turnouts on election day bite them in the ass more than anticipated. As a wild guess, I put 90 seats as the maximum "still on this planet" GOP gain.

[Mr.Phips]

Today is really a good day for internals.

Action Solutions for William Lawson (R-NC 04):

46.5% - William Lawson (R)
46.1% - David Price (D-Incumbent)

http://news.yahoo.com/s/usnw/20100907/pl_usnw/DC60126

wtf?

Well, remember that Price was elected in the 80s, then got defeated in the 94 GOP wave and got the seat back in 1996.

It´s not that impossible that he loses again in a similar wave this year. In 2008, he got 63% to 37% against Lawson and a 13-point swing from 2008 to 2010 is not impossible.

We'll see ...

The NC-04 that Price lost in in 1994 was a signficantly more Republican district than the current NC-04.  Heavily Democratic Durham county is now wholly in the district, when none of Durham was in the district in 1994(it was split between NC-02 and NC-12).  Also, marginal to Republican leaning Wake county(where Heiniman got all of his margin) was wholly in NC-04 in 1994, where now, only a small portion is in the district and it will be completely drowned out by Democratic margins in Orange and Durham.

[ajc0918]

And here's yet another:

FL-24

Adams(R): 49%
Kosmas(D-Inc): 37%

http://blogs.tampabay.com/buzz/2010/09/gop-poll-shows-sandy-adams-beating-suzanne-kosmas.html

Good.
Kosmas could have been saved by not voting for healthcare. But she didn't.
She also COULD HAVE at least tried to stop NASA from cutting 14000 jobs after the shuttle program.

Oh well. bye bye

[ajc0918]

Since today seems to be internal day, here's another:

KS-04

Pompeo(R): 50%
Goyle(D): 47%

http://www.scribd.com/doc/37051149/Raj-Goyle-August-2010-Poll

This poll is almost 1 month old and was conducted at the same time as the SUSA poll, which had Pompeo up by only 7 points - before the racist postings of Pompeo.

The DCCC should probably put some money into this district.

Do you really think this district is the least of their concerns?

[Brittain33]

Kosmas could have been saved by not voting for healthcare. But she didn't.
She also COULD HAVE at least tried to stop NASA from cutting 14000 jobs after the shuttle program.
Oh well. bye bye

1. Not in a Republican wave this size, in a district full of olds, which was drawn by a Republican for a Republican who lost because of scandal.
2. Yes, a freshman representative could have single-handedly saved the shuttle program beginning in 2009. (14,000 jobs in that district? Really?)

I'd like to think that various Republicans who went down in 2006 and 2008 could have been saved if only they'd supported my legislative priorities and gone all-out for gay marriage, but I don't think that's reasonable. (Except maybe Musgrave...)

[ajc0918]

Kosmas could have been saved by not voting for healthcare. But she didn't.
She also COULD HAVE at least tried to stop NASA from cutting 14000 jobs after the shuttle program.
Oh well. bye bye

1. Not in a Republican wave this size, in a district full of olds, which was drawn by a Republican for a Republican who lost because of scandal.
2. Yes, a freshman representative could have single-handedly saved the shuttle program beginning in 2009. (14,000 jobs in that district? Really?)

I'd like to think that various Republicans who went down in 2006 and 2008 could have been saved if only they'd supported my legislative priorities and gone all-out for gay marriage, but I don't think that's reasonable. (Except maybe Musgrave...)

Did I say 14000, oh I meant 23000
http://www.floridatoday.com/article/20100226/NEWS0204/2260321/23-000-now-expected-to-lose-jobs-after-shuttle-retirement

and maybe she couldn't alone save the program, but she could have tried. I didn't hear anything on that.

[Eraserhead]

Somebody poll NY-19 and 20 already. Sheesh.

[Brittain33]

Did I say 14000, oh I meant 23000
http://www.floridatoday.com/article/20100226/NEWS0204/2260321/23-000-now-expected-to-lose-jobs-after-shuttle-retirement

and maybe she couldn't alone save the program, but she could have tried. I didn't hear anything on that.

http://www.kosmas.house.gov/index.php?option=com_content&task=view&id=269&Itemid=1

[ajc0918]

Did I say 14000, oh I meant 23000
http://www.floridatoday.com/article/20100226/NEWS0204/2260321/23-000-now-expected-to-lose-jobs-after-shuttle-retirement

and maybe she couldn't alone save the program, but she could have tried. I didn't hear anything on that.

http://www.kosmas.house.gov/index.php?option=com_content&task=view&id=269&Itemid=1

I stand corrected, is this still in legislation?

[Smash255]

Today is really a good day for internals.

Action Solutions for William Lawson (R-NC 04):

46.5% - William Lawson (R)
46.1% - David Price (D-Incumbent)

http://news.yahoo.com/s/usnw/20100907/pl_usnw/DC60126

wtf?

Under 40,  9.7%..........   Why bother releasing that??

[TheDeadFlagBlues]

Guess who won't be paying attention to this thread.

[JohnnyLongtorso]

I don't think Price is facing another 1994, somehow. I think the district was more Republican in the 90s, although I'm not sure (it had more of Chatham and Person and some different bits of Wake).

[Niemeyerite]

Oliverio is safe?? oh, so WV will have a dmeocrat delegation after 2010, it's becoming republican this year, like arkansas =)

[Mr.Phips]

I don't think Price is facing another 1994, somehow. I think the district was more Republican in the 90s, although I'm not sure (it had more of Chatham and Person and some different bits of Wake).

And the fact that Elaine Marshall will almost certain win the district solidly will help him big time.  In 1994, there was nothing at the top of the ballot to draw out Democratic voters.   

[xavier110]

NY-24
Mike Arcuri (D-inc): 50
Richard Hanna (R): 37

LOL

[Associate Justice PiT]

Guess who won't be paying attention to this thread.

     Internal polls are obviously inferior to regular polls, but they aren't completely worthless. Releasing internals showing yourself tied or trailing is tantamount to admitting that you will lose, though there are uncommon cases where they do it on purpose to get money from the national party for races that are not yet on the radar.

     Point is, internals do contribute to developing one's picture of a particular race, though not in the same sense that regular polls do.

[Tender Branson]

Rep. Kissell (D-NC 08) is probably safe. Take a look:

The polling war in North Carolina's 8th district continued Wednesday with a new survey from Republican Harold Johnson's campaign showing the former sportscaster down just 5 points to freshman Rep. Larry Kissell (D).

Johnson's poll, which was in the field Aug. 29-30, showed Kissell with 39 percent to Johnson's 34 percent, with the remaining 27 percent either undecided or favoring Libertarian Thomas Hill. The survey had a 4.9 point margin of error.

The poll of 400 likely 8th district voters was conducted for Johnson by the GOP firm Public Opinion Strategies.

The results paint a very different picture of the race than a Democratic Congressional Campaign Committee poll from Aug. 25-29 that showed Kissell ahead 48 percent to 36 percent for Johnson. And just two weeks ago, Kissell's campaign released an internal poll that showed the Congressman up 49 percent to 32 percent. That survey was in the field Aug. 19-24.

http://blogs.cqpolitics.com/eyeon2010/2010/09/nc-johnson-poll-shows-race-tig.html

[Associate Justice PiT]

     Challenger down by 5 in his own internal poll with huge undecideds? Yeah, Kissell is in great shape.

[Vepres]

These internal polls are so contradictory when compared, ugh.

[JohnnyLongtorso]

And here I was thinking Kissell was in serious trouble. Now he's just in some trouble.

[TheDeadFlagBlues]

Guess who won't be paying attention to this thread.

     Internal polls are obviously inferior to regular polls, but they aren't completely worthless. Releasing internals showing yourself tied or trailing is tantamount to admitting that you will lose, though there are uncommon cases where they do it on purpose to get money from the national party for races that are not yet on the radar.

     Point is, internals do contribute to developing one's picture of a particular race, though not in the same sense that regular polls do.

I know, it's just that it is really annoying to see people overreact to internals. It's obvious when a polls reflects what a certain candidate wants for fundraising. A good example is the internal showing Driehaus two points behind Chabot. All credible polls show Driehaus behind by at least 10 points. Chabot needed for the race to be thought of as tossup in order to get funding from the party and from individual donors. There is nothing credible about the poll, yet red avys got all excited. This applies to Republicans too.

[Filuwaúrdjan]

'polls'

[Mr.Phips]

Kissell better thank god that there is a Senate race at the top of the ballot this year, unlike 1994 to draw out Democrats in Charlotte, which will likely be his entire margin of victory. 

[Tender Branson]

A poll for Betsy Markey (D-CO 04):

38% - Markey (D)
38% - Gardner (R)
  5% - Waszkiewicz (I)
  2% - Aden (C)
17% - Undecided

http://www.coloradoan.com/article/20100909/UPDATES01/100909005/1001/NEWS/Betsy+Markey+campaign+poll+shows+her+tied+with+Cory+Gardner