Last year's most accurate Poll: Obama by 1.1pts

[Prince Valiant]

From Investor's Business Daily...these guys got last year's race closest, which may or may not equal similar results this year:

Overall: Obama by only 1.1 pts according to this poll

http://brewcitymuscle.com/forum/attachment.php?attachmentid=5700&stc=1&d=1224855443

It's interesting to see the breakdowns of who is more likely, or less likely to be and Obama or McCain supporter.

Wish that it had a state by state poll to work with...to see how the electorate would shape up :(

[GTSLOW]

^ Obama supporter. :fire

[07ROUSHSTG3]

it will be closer than most think. there are a lot of people that are riding the obama wave right now, but when the time comes to pull the lever, they won;t be able to do it. kerry/edwards was a lock before the 04 election too, remember how that played out. the news stations were all but popping the corks on the champange (sp) when the exit polls were coming in, LOL!!! lots of people, especially middle-upper class whites, are almost being pressured into being obama supporters. they will come around come election day though.

[Prince Valiant]

^ Obama supporter. :fire
umm...yeah. :goof

[animal]

Polls = bench racing = worthless :devil

[Prince Valiant]

Polls = bench racing = worthless :devilTo a large extent I agree...even just releasing polls has an effect on poll numbers. And the only poll that matters of course, is the actual voting tally on election day.

To me, I put only a modicum of weight into polls (to me, it suggest trends rather than outcomes)...and I certainly give some polls more weight than others since the design and sampling techniques of the poll will have a HUGE difference in the outcome. Even then, national polls like above only tell a small part of the story...since it's state by state polls that'll give you the best idea of who'll win.

[animal]

Eh. Polls deserve zero weight. Polls are as manipulatable as statistics are. Select your sample the right way and you can get a poll to bend just like statistics. They're more an indication of who the poller wants to win than they are an indication of anything real.

The mere presense of a statement like "among likely voters" that is mentioned whenever poll results are mentioned is a cover your ass statement. They might as well replace that with "but this doesnt mean anything".

[VroomPshhTsi]

^ True

Just look at the sampling size of these polls, very very small.

The poll above only counts just over 1,000 people.

[Prince Valiant]

Never-the-less, the pollsters methodology that called last year's correctly could well be the methodology that calls this years correctly.

As far as 'likely voter' though, that's the only poll worth even beginning to consider...those among simply 'registered voters' or of those of 'voting age' (which you'll never see in this day and age) are rahter utterly worthlless since it measure many's sentiment who won't vote...whereas ''likely voters" measures a group that votes ~90-95% of the time.

But to your point Animal...polls GET zero weight. They don't determine the election at all.

As far as sampling size, they've given a good enough size imo to give a reasonable margin of error for the purposes of tracking a trend...

[Crawlin]

Look at that 18-24 age group !!!!

now if you could just GET the majority of those voters to the actual POLLS it would make a difference.

[animal]

Maybe that pollster just got lucky last year ;)

Also actual sampling size doesn't matter if you're hand picking samples or portions of your samples to arrive at a predetermined conclusion. Obivously they have no weight to throw around when it comes to pulling levers. By deserving zero weight I meant media exposure. Half the time near elections, polls get more weight than any other newsworthy story, and they're nothing more than bullshit. Not sure why during a news update on the radio I hear one sentence on a big news story and a paragraph on results of a couple polls. It does get kindof tiring though. Sigh. I hate when elections come around... just one big ass scam.

[ND4SPD]

Look at that 18-24 age group !!!!

now if you could just GET the majority of those voters to the actual POLLS it would make a difference.

That tells me something is screwed up in the poll... the 18-24 age group should show heavily towards Obama, not the other way around.

[Prince Valiant]

Look at that 18-24 age group !!!!The poll states that it's not valid since the sample size is too small...the margin of error has to be in the +/-50% range, lol.

Two possible reasons for this imo...1) Most 18-24 haven't registered to vote yet or have only registered recently, as this is most likely their first election...so as a whole they aren't firmly entrenched on the voter rolls as other demographics and 2) The pollsters call households (including apts) only...not cell phones...which younger people are less likely to have land lines than older people even in their house/apt.

Based on the percentages, I'd guess that they got either 25 or 50 in that age group. Which would STILL indicate a rather high pro-mcCain response among that age group (my guess it's 60-40 in that group, pro-Obama)

[Russ Jerome]

Polls = bench racing = worthless :devil

Oddly enough Obama is as useless as bench racing
and polls...coincidence or could he be Satan?

Welcome to North American candidates can I help you?
Yes, I would like a token adult male for my puppet.
OK fine, and his compexion?
Well not to dark and not to light.
Thats fine he just needs to be stupid and do everything
his lobbiest suck off for.
Oh we have you candidate already to go!

[Goat Roper]

For the past year (as opposed to the last presidential cycle), this has been the most accurate; http://www.realclearpolitics.com/epolls/2008/president/us/general_election_mccain_vs_obama-225.html

During the entire caucus/primary season it was spot on.

[Goat Roper]

And using the orginal poll, which the RCP poll takes into consideration, here is the "Day 12" poll

http://www.ibdeditorials.com/IMAGES/smltrackpoll102508.gif

[Prince Valiant]

I'm actually a big fan of RCP's...I check them daily. I read many of their linked articles...I love their electoral map.

BUT, the thing is, I do think there is a difference b/w the last congressional election cycle and a presidential cycle...it brings a different dynamic since many ONLY vote in presidential cycles, etc. So the RCP formula might be skewed because of it's use of other, habitually wrong polls (re: Zogby)

That's not saying I think McCain will win...quite the opposite; I personally AM confident that he'll lose. But, I'm also confident that the outcome of the election will track tighter than RCP polls, and more importantly, many pundits are currently predicting...I believe that the difference will probably be a Obama win by ~3 or so percentage points.

[subliminal1284]

Obamas the next president, better get used to it.

[GHOSST]

Obamas the next president, better get used to it.

Couldn't agree with you more, people just need to get over it, especially on here. I mean, initially nobody wanted Bush to win... but he farking won 2 twice.. I dunno...

[Rocket Power]

Couldn't agree with you more, people just need to get over it, especially on here. I mean, initially nobody wanted Bush to win... but he farking won 2 twice.. I dunno...
I disagree I voted for him twice.
I think polls can influence an election. People see one candidate way ahead of another and they say "forget it my guy can't win anyway" thereby losing a vote for the other side also some people just want to be on the winning side when so and so is winning I'll vote for him.
Seems like you hear more about polls that Obama is father ahead than ones that are closer in general on the news. But I am sure that's just coincidence .