Oh PLEEEEEAAAASSSSSEEEEEE....
Let's start attacking Gallup now
Facts:
2000 Race
Final Gallup Poll
Bush 48
Kerry 46
Nader 4
Actual Result:
Bush 47.87%
Kerry 48.38
Nader 2.73%
Error:
Bush 0.13%
Gore 2.38%
Nader 1.27%
Wow - Average Candidate error was 1.28% - you're right a really sloppy outfit - I would throw this poll out...
So how did they do in 2002 you may ask...?
Well, according to an anaylsis done by the Natioinal Council on Public Polling, of all the firms surveyed, the firm with the lowest average error was, you guessed it... Gallup...
http://www.ncpp.org/2002SenGovPoll/2002ElectionPolls.htmlOk...so Gallup go lucky in 2000 and 2002...
Well... going back to 1936...
The record of accuracy achieved by Gallup in these "horse-race" or "head-to head" measurements is unsurpassed. For example, in the sixteen presidential elections since 1936, the deviation between Gallup's final pre-election survey figures and the actual election results is 2.2% and, since 1960, only 1.5%.
http://www.gallup.com/poll/content/?ci=1210Gallup has been polling for 70 years - yes you can pick on one poll here and there - in 70 years you get a few bad polls.
But year in year out, poll after poll after poll, Gallup can go toe to toe and come out looking just fine
That being said, I will post AGAIN for those of you who missed it the 1st, 2nd, 9th, and 43rd times I posted it...
Point #1The Gallup poll is designed to do one thing, and one thing only.... Predict the outcome of an election when you are
very close to the actual election.
The Gallup likely voter screen is very sensitive to changes in voter motivation and interest and hence when you are many weeks out produces
large and artificial swings in the race.Point #2Any poll taken right now is, by definition "wrong" in the Gallup likely voter screen is based on if the election were held tomorrow - the election is NOT tomorrow
Point #3Gallup has Bush +8 among RVs
Gallup has Bush +13 among LVs
This 5% gap is historically a bit large, but simply represents that GOP supportes are really cranked right now. Their candidate is up and doing well. DanRatherBiasedMemoGate has the GOP faithfull all worked up, while Dem supporters are a bit depressed for all the same reasons. - Of course more GOP supportes are interested and thus "likely" right now.
Will this huge intensity gap between the GOP and Dems last until election day..?
Probably not.
But this poll is a snapshot of Today, NOT a prediction of Tomorrow.
and TODAY the GOP is more energized
What does Gallup predict for Nov 2nd..?
Lets wait for their Nov 1st Poll to find out
Point #4Gallup uses the "purest" of all methodologoes in the sense that they weight very little and place vey few "boundry conditions" on their sample.
The benefit of this is that when the electorate truly actually does change, Gallup will catch it when many other pollsters will throw the proverbial baby out with the methodological bathwater by weighting changes away.
In 2002 when most pundits were "stunned" (To use Zogby's words) that the GOP gained in both the House and the SEnate, Gallup predicted it perfectly - The final Gallup "Generic" congressional ballot had the GOP up 4% (actual was 5%) while Zogby, ABC, etc all had the Dems up 4 or 5
In 1994 when the GOP wave took control of the House and Senate, everybody was stunned - except Gallup who also had it right. (Mason Dixon also had an amazing year in 1994 as well to be fair)
The price Gallup pays for this is they get a bit more pure random noise in their results than other pollsters.
It's not a "flaw" it's a choice.
Bottom lineIs Bush up 13%..?
I personally don't think so.
I think he has a "real" lead of 4% or so, plus some "froth" due to the GOP being rather cranked up right now.
For the record, I also expect that we will see a lot of national polls in the next few days where Bush is up high single digits.
It likely will not last, but that's where the race is now
