[Kerrigan]

Zogby, which thus far has been most favorable to McCain in its polling data, suggests an Obama landslide is in the making. One has to wonder if this is not a continuation of their pro-McCain sentiment in that they are attempting to scare conservative voters to the booth. The basis of this speculation is that polls with greater accuracy, like AP, show that it is once again tied.

[besson3c]

Whatever floats your boat, guys. All I'm saying is that there simply is no logical reason to prepare for a McCain victory, while there is some reason to be cautiously optimistic about Obama's chances.

[Kerrigan]

A little LSAT fun.. Which one of the following arguments most closely resembles Besson's pattern of reasoning?

Correct answer:

D) 44% of meteorologists assert that the hurricane will not hit my house. 43% of meteorologists assert that the hurricane will hit my house. Therefore, there is no logical reason to prepare for the hurricane hitting my house.

[besson3c]

Nicely played Kerrigan, refute an argument with an insult!

[Kerrigan]

More LSAT fun. Besson's attempt to counter Kerrigan's argument is most vulnerable to criticism on the grounds that it:

Correct answer:

C) Attempts to discredit Kerrigan by citing a fallacious component that is not present in Kerrigan's argument.

[stupendousman]

Originally Posted by Kerrigan 

Zogby, which thus far has been most favorable to McCain in its polling data, suggests an Obama landslide is in the making.

Again, Zogby has been UP and DOWN. Right now they have McCain down. Tomorrow you might see him 12 point ups.

One has to wonder if this is not a continuation of their pro-McCain sentiment in that they are attempting to scare conservative voters to the booth. The basis of this speculation is that polls with greater accuracy, like AP, show that it is once again tied.

I don't think any are "accurate". The most accurate poll last election (IBD) has it at a statistical tie as well, but I don't figure that to be "accurate" either. There's really no way to tell where the election truly is unless you're dealing with maybe more than a 8 point aggregate advantage, IMO.

[Kerrigan]

So if most polls show at least an 8 point aggregate advantage, then they are "accurate," and if most polls indicate a statistical tie, then they are not "accurate?" Huh? I'm not trying to be contradictory, I just don't get your point regarding accuracy.

[besson3c]

At any rate, reality still stands: I've yet to see a poll aggregator that doesn't show Obama significantly ahead of McCain, and it has been this way consistently for at least a couple of weeks.

I keep pushing this point because I'm just of curious as to how far this strange bizarro world of not acknowledging facts in the name of truthiness will go.

[Kerrigan]

Everyone knows Obama is favored to win. The only "bizarro world" of unacknowledged facts which triggered my initial response, was saying that there was "no logical reason" to be prepared for something that has a realistic chance of occurring.

[kobi]

After seeing the AP results I knew they were a little to good to be true for McCain.

Here's the proof:

AP poll flawed

If you only poll a majority of Evangelical Christians then of course McCain is going to win.

So much for the Liberal Media, it's more like the Right Wing media.

[besson3c]

Originally Posted by Kerrigan 

Everyone knows Obama is favored to win. The only "bizarro world" of unacknowledged facts which triggered my initial response, was saying that there was "no logical reason" to be prepared for something that has a realistic chance of occurring.

So you are saying now based on what data we have that McCain has a realistic chance of winning?

I guess this depends on whether you define realistic to mean "likely", or "possible"...

[besson3c]

Five Thirty Eight has the probability of a McCain victory at 6.5%. That is, if you play out all of the possibilities based on the poll data we have (i.e. McCain sweeping those 7 or 8 contested states), he can win, but the probability of that happening is as noted.

[stupendousman]

Originally Posted by Kerrigan 

So if most polls show at least an 8 point aggregate advantage, then they are "accurate," and if most polls indicate a statistical tie, then they are not "accurate?" Huh? I'm not trying to be contradictory, I just don't get your point regarding accuracy.

No. I think that if you can show as an aggregate, more than an 8 point lead (most polls allow for about 7 points for the margin of error) then you can assume a trend one way or another as far as a lead. Especially when it's the Democrat in the lead since they've been way oversampling them.

Originally Posted by kobi 

After seeing the AP results I knew they were a little to good to be true for McCain.

Here's the proof:

AP poll flawed

If you only poll a majority of Evangelical Christians then of course McCain is going to win.

So much for the Liberal Media, it's more like the Right Wing media.

There are apparently more minorities and especially African Americans voting this time. That's the excuse pollsters use for giving astronomical amounts of plus points for Democrat turn-out. African American voters tend to consider themselves born-again Christians at a greater rate than normal. I just googled and came up with this in seconds:

Religion and Politics in America ... - Google Book Search

Originally Posted by besson3c 

Five Thirty Eight has the probability of a McCain victory at 6.5%.

They seem to slant in favor of their Chicago brethren. For instance, they had a story on the candidates operations in certain states like West Virginia where they claimed Obama had something like 15 satellite offices in the state and McCain had one, despite every county Republican election office also serving double duty as a McCain election headquarters. I've been to two major cities in West VIrginia of late and after asking around they both had "Obama Headquarters" and both shared the space with the local party headquarters same as McCain.

[zerostar]

I can tell you from our last polls FL will be tight, but I think Obama has huge momentum and even without FL will win handily from everything I am hearing from just around the SE.

[kobi]

Looks like were seeing a huge Powell bounce since the Colin Powell endorsement for Obama. The Powell endorsement has helped with Military families and undecided voters. Obama is up 11 pts in the ABC-News/Washington Post poll from this Powell bounce.

Powell Bounce

Obama is ahead by double digits in the latest Big Ten Poll in all the battleground states in the Midwest.

Big Ten Battleground poll

The new Zogby/Reuters/C-SPAN poll has Obama up by 12 pts.

Zogby poll Thurs 10/22/08

[Eug]

Yeah, at this point I'd be very surprised if Obama does not win.

Things can change quickly, but barring any real stupidities on the Obama camp side, it seems he's pretty much got it now.

[stupendousman]

Originally Posted by kobi 

Looks like were seeing a huge Powell bounce since the Colin Powell endorsement for Obama. The Powell endorsement has helped with Military families and undecided voters. Obama is up 11 pts in the ABC-News/Washington Post poll from this Powell bounce.

The last ABC News poll showed him at 10 pts. before Powell. HUGE BOUNCE!!!

...at the same time he sank in some polls, was up in some polls and was tied in others. If you're looking for guidance as far as performance goes in the polls....keep looking. Tomorrow there will be something else you'll like, and probably hate.

[Big Mac]

I love this overconfidence. Keep it up, lemmings.

[zerostar]

The fact is without OH McCain has no way to win right now. Forget PA, forget here in FL, its all moot without OH.

Obama leading in last 3 polls:

University of Wisconsin: 53-41
Quinnipiac: 52-38
TIME/CNN: 50-46

The big thing here, (and even with the TIME/CNN it is the closest poll) is that Obama has crossed 50% in all 3 polls. That is huge for him, on top of this McCain is running out of money, while Obama is still very flush.

Because some are confident and getting excited they are lemmings? What a jerk. This was the exact opposite last time and I don't remember anyone being called a lemming.

[zerostar]

Originally Posted by stupendousman 

The last ABC News poll showed him at 10 pts. before Powell. HUGE BOUNCE!!

Actually with undecideds, independents, and soft party IDers that has been a huge contributor, you got to realize 2-3% gain in these groups is HUGE and can clinch an election big time.

[Eug]

Overconfidence? I'm not even voting for the guy. Cuz, I don't live in the US.
It's just a reasonable assessment of the situation.

The reality is that it's going to be extremely difficult for the McCain camp to overcome this difference. It's possible, but very unlikely. Palin sure isn't making it easy for him either.

[kobi]

Originally Posted by stupendousman 

...at the same time he sank in some polls, was up in some polls and was tied in others. If you're looking for guidance as far as performance goes in the polls....keep looking. Tomorrow there will be something else you'll like, and probably hate.

I love how you ignore the Zogby and Big Ten poll result. That's were the bounce is, not in the ABC poll.

I do agree with you, tomorrow it will be different numbers. There's a ebb and flow to these polls, plus it's how the media keeps us interested in the election. There has to be some sort of drama, otherwise they will start loosing viewers.

[kobi]

At this point, McCain has to keep every state that Bush won in 2004, plus turn a Blue State Red in order to win. McCain is putting all his eggs into Pennsylvania and is ignoring all the other states, that's not a good move. Florida is his best shot to keep Red, but without Ohio there's no McCain win.

The math is against him.

With the McCain and GOP pulling money out of traditional Red States, and McCain is having to campaign to DEFEND traditional Red States such as Virginia, Indiana and Florida this late in the game means that your campaign has screwed up. With little to no campaign money left and Palin/McCain on the defensive, it's not looking good for ol' Johnny boy's campaign.

[zerostar]

You're wrong, McCain has Obama "right where he wants him."

[Big Mac]

Originally Posted by kobi 

At this point, McCain has to keep every state that Bush won in 2004, plus turn a Blue State Red in order to win.

Uh, no, if he could do that he would win with just those states.

[besson3c]

Originally Posted by zerostar 

The fact is without OH McCain has no way to win right now. Forget PA, forget here in FL, its all moot without OH.

Obama leading in last 3 polls:

University of Wisconsin: 53-41
Quinnipiac: 52-38
TIME/CNN: 50-46

The big thing here, (and even with the TIME/CNN it is the closest poll) is that Obama has crossed 50% in all 3 polls. That is huge for him, on top of this McCain is running out of money, while Obama is still very flush.

Because some are confident and getting excited they are lemmings? What a jerk. This was the exact opposite last time and I don't remember anyone being called a lemming.

Why Ohio? That's kind of a weird way of looking at things, don't you agree?

McCain has to sweep all of the contested states that are not comfortable Obama leads including, off the top of my head, FL, CO, VA, OH, NM, NV, IN, MO, NC. As of right now, Obama has leads of some level of comfort in most, if not all of these states.

[stupendousman]

Originally Posted by kobi 

The math is against him.

EXACTLY.

It's the "math". The "math" that shows despite having even turnout last Presidential election and at VERY BEST when Republicans stay home during congressional races getting 3, maybe 4 points better turn-out than Republicans, many of these polls are weighing the results +9 for the Democrats. In a state like Virginia, I think they are giving Obama +7 points in most of the state polls when almost every election shows a +4/5 advantage by the Republicans.

Yes, it will be hard to win with those odds. It's a good thing for McCain that those odds are totally unrealistic and reflect nothing that can be shown to be past precedent.

As I've said before, some pollsters are really going to end up with egg on their face. I'd suggest that maybe they'll end up learning something, but EVERY YEAR we get the same exaggerated claims about how Democrat turn-out will trump Republicans and it seldom occurs. People who aren't inclined to vote....aren't inclined to vote. No Obama "magic spell" is going to cure that.

[Eug]

Originally Posted by stupendousman 

as i've said before, some pollsters are really going to end up with egg on their face. I'd suggest that maybe they'll end up learning something, but every year we get the same exaggerated claims about how democrat turn-out will trump republicans and it seldom occurs. People who aren't inclined to vote....aren't inclined to vote. No obama "magic spell" is going to cure that.

Originally Posted by big mac 

i love this overconfidence. Keep it up, lemmings.

.

[Big Mac]

This country will definitely get the president it deserves in 12 days. If the people are gullible enough to be conned by Barack Obama, then so be it.

[zerostar]

Originally Posted by stupendousman 

many of these polls are weighing the results +9 for the Democrats.

What polls are those? Are you STRICTLY talking about STATE polls?

lets look at the major ones RCP tracks:
http://www.realclearpolitics.com/epo...obama-225.html

Here is how the lean in 2008. POLL [LEAN #] (Todays Poll) is how I have it setup


Zogby Lean [REP +1 or +2] (TODAY: Obama +12)

Reasearch 2000 / Daily KOS [DEM +2] (TODAY: Obama +10)

Rasmussen Reports [REP +.5] (TODAY: Obama +7)

The Diageo Hotline Poll [NONE detected] (TODAY: Obama +12)

Battleground / Tarrance / George Washington U. [REP +3 Early now Neutral] (TODAY: Obama +4)

Gallup [2008 Neutral] (TODAY: Obama +5 Trad +8 Expanded)

Investors Business' Daily / TIPP [not enough data] (TODAY: NONE- Obama +6 Yesterday)

ABC News / Washington Post [DEM +3] (TODAY: Obama +11)

Factor in all the leans and it shows exactly what the RCP average shows, about a +8 point lead nation-wide.

[zerostar]

I thought this fit with the discussion:

Pollster John Zogby: "Obama now has a huge lead among young voters, independents, and Hispanic voters. It's obviously not over. Frankly, this could tighten up and then loosen up again before Election Day. We saw movement on Election Day in New Hampshire, but at least for now, Obama has a very big lead. In the absence of news, McCain is not connecting. He seemed to be connecting during and immediately after the last debate, but got lost in issues that are not on people's minds. At some point, there are some issues that just overwhelm, and McCain has been particularly weak on the economy. He misstated the problem, confused his position, acted in a frantic way, and then looked like he wanted to run away from it. Meanwhile, Obama has been cool and confident, which worked for FDR in 1932 and worked for Ronald Reagan in 1980."

"I am very comfortable with our sample, especially given our track record in the last three presidential elections. Look at other polls and ask - Do they have enough college educated respondents? Enough Hispanics? Enough young voters? We do. And we have more Republicans in our sample than anyone else."

http://www.zogby.com/news/ReadNews.dbm?ID=1605

Today Zogby has Obama up 52.2%-40.3%. 11.9 points

[stupendousman]

Originally Posted by zerostar 

Here is how the lean in 2008. POLL [LEAN #] (Todays Poll) is how I have it setup

I'm not sure we are even talking about the same thing. I'm not referring to when Pollsters force leaners to chose. I'm talking about the percentage of republicans to democrats the pollsters use in their sample of likely voters. Last Presidential cycle, the turn out had equal republicans and democrats.

For instance, you list ABC News. The last internal numbers I can find for their weighting is that they give Democrats 9 extra points in their poll. The +5 and +8 seems about right for Gallup, but I also know that Pew gives them 14, and Fox gives them 9 as well. I even think I got the numbers from fivethirtyeight.com. They complained that those numbers were too low and that they should be weighing Democrats even higher!

If you believe that the Democrats are going to double or triple their best turnout in the last 30 years, you can trust the polls. If you have a brain and you do not, then you have a reason to doubt that the current polls are very accurate for much else than a tool for propaganda.

[zerostar]

Originally Posted by stupendousman 

For instance, you list ABC News. The last internal numbers I can find for their weighting is that they give Democrats 9 extra points in their poll. The +5 and +8 seems about right for Gallup, but I also know that Pew gives them 14, and Fox gives them 9 as well. I even think I got the numbers from fivethirtyeight.com. They complained that those numbers were too low and that they should be weighing Democrats even higher.

I believe we are talking about the same thing :-) [House Effect/Lean], I got the ABC # from 538 as well, the ones I listed are the methodology used for 2008. PEW and FOX aside (as I can't find those) the polls are all showing the same thing, and our poll here (now Dem +2) in FL shows that as well.

[MindFad]

Originally Posted by Big Mac 

This country will definitely get the president it deserves in 12 days. If the people are gullible enough to be conned by Barack Obama, then so be it.

Here you go, chief. You know, no matter the outcome.

[besson3c]

Originally Posted by stupendousman 

EXACTLY.

It's the "math". The "math" that shows despite having even turnout last Presidential election and at VERY BEST when Republicans stay home during congressional races getting 3, maybe 4 points better turn-out than Republicans, many of these polls are weighing the results +9 for the Democrats. In a state like Virginia, I think they are giving Obama +7 points in most of the state polls when almost every election shows a +4/5 advantage by the Republicans.

Yes, it will be hard to win with those odds. It's a good thing for McCain that those odds are totally unrealistic and reflect nothing that can be shown to be past precedent.

As I've said before, some pollsters are really going to end up with egg on their face. I'd suggest that maybe they'll end up learning something, but EVERY YEAR we get the same exaggerated claims about how Democrat turn-out will trump Republicans and it seldom occurs. People who aren't inclined to vote....aren't inclined to vote. No Obama "magic spell" is going to cure that.

Except for 2006, of course, which you keep glazing over.

[stupendousman]

Originally Posted by zerostar 

I believe we are talking about the same thing :-) [House Effect/Lean], I got the ABC # from 538 as well

The number I got from them is +9, not +3. It would be odd if the poll suddenly changed weighting in favor of McCain and the numbers still stayed essentially the same.

the ones I listed are the methodology used for 2008. PEW and FOX aside (as I can't find those) the polls are all showing the same thing, and our poll here (now Dem +2) in FL shows that as well.

Again, I don't think we are talking about the same thing. The "house effect" is different than what I'm referring to. The "house effect" is how a pollster trends based on their method of operation. How they ask questions, where they get samples from, etc. I'm referring to the specific weighting a poll gives Democrats and Republicans based on predicted poll turn-out. That's something totally different and is what pollsters are using to give up to a 14 point advantage to Obama before even looking at the numbers.

[zerostar]

Originally Posted by stupendousman 

Again, I don't think we are talking about the same thing. The "house effect" is different than what I'm referring to. The "house effect" is how a pollster trends based on their method of operation. How they ask questions, where they get samples from, etc. I'm referring to the specific weighting a poll gives Democrats and Republicans based on predicted poll turn-out. That's something totally different and is what pollsters are using to give up to a 14 point advantage to Obama before even looking at the numbers.

Interesting, not sure what is different in saying the Zogby favors asking Republicans by 2points and what you are saying.... Why would no one else list this? Where can I find this info for all the polls (the relevant polls, not these outliers)

I don't think any I watch are that +14 you talk about, got any links to this? Will be interested to see afterwards what the outcome is.

[stupendousman]

Originally Posted by besson3c 

Except for 2006, of course, which you keep glazing over.

What was the ratio in 2006? I thought it was +4 Democrats. I'd read that somewhere.

[stupendousman]

Originally Posted by zerostar 

Interesting, not sure what is different in saying the Zogby favors asking Republicans by 2points and what you are saying....

What I'm saying is that polls will decide what they think actual turn-out will be, then use that as their sample percentages. Turn-out the last presidential election was pretty much equal Dem. and Rep. As I've explained, some are saying that the Democrats will get 14% more people to the polls. They've NEVER been anywhere even close in the past 30 years. They are either purposely skewing the numbers or living in a fantasy world.

This is different then just showing a trend that a certain pollster is X% more likely to have polls that show one party or the other more favorably than others.

Why would no one else list this? Where can I find this info for all the polls (the relevant polls, not these outliers)

They don't always release the info. This is known as the "internals" of the poll. Some pollsters do explicitly tell you (Gallup for instance). Others will tell if asked and others treat it as a trade secret. There are ways to decipher weighting by looking at the break-downs and comparing them week to week, but it's a pain in the *ss to do with every poll. Like I said, I believe that 538 had them listed a week or so ago. Those are the numbers I've been using.

[besson3c]

Originally Posted by stupendousman 

What was the ratio in 2006? I thought it was +4 Democrats. I'd read that somewhere.

I don't have the numbers in front of me, but the race went to Democrats in greater number than the pollsters were predicting on average.

I think that the best way one can gauge the sort of x variables on election night in advance is looking at the number of independent voters and switchers going one way or another in a particular demographic. These people can go either way depending on the political climate, there is no inherit Democratic or Republican bias in there that is set for eternity.

Really, I think you need to open your eyes to the state of disarray that the Republican party is in right now. Look at exit polls, the number of Republicans that turned out to vote early, the size of McCain rallies, operations on the ground, number and quality of endorsements, donations, etc. If you were to do this objectively, I really have a hard time understanding how anybody could rightfully say that the momentum isn't in the favor of the Democrats this year. I'm pounding you with this, stupendousman, because I really don't want you to feel disappointed like many of us did in 2004.

[besson3c]

McCain also sucks at dancing

[xi_hyperon]

ahahahahaha. That's awesome.

[stupendousman]

Originally Posted by besson3c 

I don't have the numbers in front of me, but the race went to Democrats in greater number than the pollsters were predicting on average.

True. They were probably going based primarily on the numbers from 2004. They didn't take in to account that 2004 wasn't a Presidential election year (where turnout is greater) and the GOP wasn't crazy about it's own people in office.

To me, it looks like they are overcompensating for 2004. Even if you assume it's going to be another 2004 (and I don't think that you've really got a depressed GOP base like in 2004), that means maybe a 4 point Dem advantage. Even if you are generous, and say they'll do even better, that might mean 5 points. 14 points? Someone needs to pass me some of whatever they are smoking.

Really, I think you need to open your eyes to the state of disarray that the Republican party is in right now. Look at exit polls, the number of Republicans that turned out to vote early...

...is meaningless. Republicans aren't typically the ones that have troubles getting their people to the polls on election day. It's the Democrats who have traditionally had difficulty despite their advantage in numbers.

....number and quality of endorsements, donations, etc.

Again, mostly irrelevant. Republicans always get fewer endorsements, and McCain didn't lie like Obama and decide to start accepting unlimited $200 donations from people like Mike Hunt and I.P. Freely. Some objectivity you have there.

Again, I'm not going to be disappointed. Unless something drastic happens, my only prediction this year will come true.

IBDeditorials.com: IBD/TIPP Economic, Presidential Election, and Political Polls -- IBD/TIPP Tracking Poll: Day Eleven

With last election's most accurate poll putting it at a tie, and Pew Research putting it at 14 points, someone is going to end up with egg on their face.

[besson3c]

Perhaps I'm misunderstanding your position then? Are you saying that McCain will win, or that Obama's win won't be as great as 14%?

[olePigeon]

Originally Posted by besson3c 

Perhaps I'm misunderstanding your position then? Are you saying that McCain will win, or that Obama's win won't be as great as 14%?

Good luck with that.

[Wiskedjak]

Originally Posted by stupendousman 

To me, it looks like they are overcompensating for 2004. Even if you assume it's going to be another 2004 (and I don't think that you've really got a depressed GOP base like in 2004), that means maybe a 4 point Dem advantage. Even if you are generous, and say they'll do even better, that might mean 5 points. 14 points? Someone needs to pass me some of whatever they are smoking.

So, with all of your statistical calculating, are you taking into account Republican turnout for the 2008 primaries?

[Wiskedjak]

Originally Posted by stupendousman 

With last election's most accurate poll putting it at a tie ...

You mean last election's luckiest guess, right? Because, based on everything I've read from you, "polling" is just "guessing". Just because one poll was close to right in the last election doesn't mean it will be accurate this time, unless it's got a trend of being accurate.

[ort888]

The polling data is showing me that Obama is up by around 35 points. I'm getting these numbers by looking at the actual numbers and replacing them with number I like better. It's a very scientific process. Very technical.

[stupendousman]

Originally Posted by besson3c 

Perhaps I'm misunderstanding your position then? Are you saying that McCain will win, or that Obama's win won't be as great as 14%?

Neither.

I'm predicting that there are going to be major polling firms this time out who are going to be very embarrassed. Next election, they aren't going to have much credibility.

[stupendousman]

Originally Posted by Wiskedjak 

You mean last election's luckiest guess, right?

Correct.

Originally Posted by ort888 

The polling data is showing me that Obama is up by around 35 points. I'm getting these numbers by looking at the actual numbers and replacing them with number I like better. It's a very scientific process. Very technical.

Ort, I'm sending your resume to Pew Research. They could use more like minded thinkers! Based on this year's results, I think you'd fit in great as a pollster!