Skip to comments.Election Polls -- Accuracy Record in Presidential Elections-Gallup Poll Accuracy Record
Posted on 10/21/2012 11:54:26 AM PDT by Signalman
Click the link to see the accuracy of the Gallup poll from 1936 to present. With a few exceptions, it's been quite accurate, usually not off more than a few points either way. And those that are off more than 3 pts on the plus side have usually gone to the Dem.
(Excerpt) Read more at gallup.com ...
Gallup’s top guy was just on with Chris Wallace and he refused to ‘predict’ the election outcome with two weeks to go.
On Oct 21 of 2008, Gallup had Obama +8. Obama won by 7.
Turn out the lights!
Is Zogby still around? Haven’t heard any polling from him. He used to be the most accurate, now it’s Rasmussen.
A lot of these polls will favor the leftist until the last one before the vote.
The last one they try harder to be accurate because that’s what they are judged on for future business.
Gallup has been polling US Presidential Elections since 1936. They’re October/November polls are excellent historically and their misses don’t favor either party. Their one big miss outside the MOE? You guessed, it “Dewey beats Truman!” :-}
In '48 there was a third party candidate that got 11 points and in '76 it was predicted at less than the margin of error.
And there is one pattern that is apparent if you look that the numbers...Gallup routinely underestimates Republican numbers and over estimates democrap numbers.
What on earth is wrong with Investor’s Business Daily???!
I guess we will know on Nov 7th. But I think Gallup is slightly over estimating Romney’s lead, and most of the rest are under estimating his lead. I believe he is in the +4~5 range right now nationally.I have thought it would end up being a 52-47 or 52-48 Romney win with 300+ EVs.
Regardless of poll or actual number, every poll is trending Romney big time the last 2 weeks. Virtually every poll shows him moving at least 4 pts.
And this movement seems sustained. We started to see slight movement right before debate #1, and it has accelerated since. And it doesn’t seem to be a debate bump. I think we have seen a fundamental shift in the election. Maybe Romney’s strong Debate#1 pulled the shift ahead.
Even the media and others are starting to acknowledge R&R are winning and have a good chance Nov 6th. If there is no major shift coming out of the final debate, this will begin to “wrap up” pretty quick. The last week will be more about down ballot help and of course GOTV.
Is Zogby still around?”
I have been wondering the same thing. Used to receive a request to participate in an on-line poll almost weekly and then nothing. Really liked their questions. Generally dealt with issues that were in the news and political questions.
You are most correct....From Eisenhowers re-election in 1956 (excluding Bush in 2004), the incumbent has also received about 2% less in the actual vote than their final poll...so count on Zero to slightly underperform his final poll number...
The last Zogby poll I remember seeing (maybe 2 weeks ago) had Obama up by 3. Very rarely look for Zogby results as it is a Muslim-run outfit not known for its honest polling methods.
***Gallup routinely underestimates Republican numbers and over estimates democrap numbers****
Right - Fox always mentions those two numbers but the Gallup
guy disputed the “Democrat weighted”- responders. Stated that they don’t ask for Party affiliation - only:
Registered to vote.
Likely to vote -or- enthusiasm.
Party registration is listed as a ‘foot note’.
Rasmussen may be showing a tight race but finally they are seeing a falling Obama approval rate, down to -16 today! I watch that closely and believe this is important to leveling out some of the volitility.
The Supreme Court will likely settle this thing.
Actually doesn’t look to be that accurate. Gallup was #17 overall in 2008.
Last 4 presidential elections since 1996 Gallup has been extremely accurate. Hard to argue with success, especially if your guy is winning. ;-)
“Is Zogby still around? Havent heard any polling from him. He used to be the most accurate.”
Step away from the bong.
Zogby has gone so far into the tank for his ‘Arab American’ causes (read that as CAIR because they seem to be in lock step) that his polling has become useless
And that miss was admittedly because they quit polling a month out of the election
their polster keeps using 2008 modeling for their sample split....good luck with that
Zogby is an Arab. After 9-11, Zogby became far more antagonistic to conservatives, far more prone to use polls to create opinion, instead of reflect it. However, polling was still his business, and he always seemed to give in to reality as the election neared. His final polls remained fairly accurate, as I recall. His ego demanded accuracy as to final results, when it was too late to push poll any more.
So, if he were still in the game, I would expect him to be showing polls that give momentum to Romney and a lead, as Gallup is doing. I haven't seen any Zogby polls, which is why I asked whether he is still out there. That's when you decided to be a jump in with a rude, uninformed comment.
Lighten up brother.
Zogby has never been as reliable or accurate as Rasmussen. Haven’t you ever heard of “Zogby sauce”?
In 2008 Zogby was 18th in accuracy out of 20 polls.
Interesting comments about Zogby. Certainly didn’t know about the Muslim connection or would never have responded. Last on-line poll they sent me included questions on BO versus Romney on various issues. I gave BO all zeros and Romney at least an 8 our of 10. Will just add them to my blocked list.
If you throw out 1936, yes...otherwise no.
Overestimate R by 0.33 pts...throw out 1936, it’s -0.03.
Underestimate D by 0.24 pts...throw out 1936, it’s +0.12
Overestimate liberal (Henry Wallace-McCarthy-Nader) by 0.21, or 1.3 if you consider they competed in only 3 elections.
Overestimate Dixiecrat-American (Thurmond-George Wallace)
by 0.05, or 0.5 if you consider they competed in only 2.
Underestimate Neutral (Anderson-Perot) by -0.26, or -1.67 if you consider they competed in only 3....almost entirely due to 1992.
Underestimate others by -0.09, or -0.24 for the 7 elections they appeared on the radar.
nothing you say contradicts anything I said. Does that mean you’ve been hitting the bong? Zogby was, however, the most accurate for quite a few election cycles, until after 9-11, after which he became progressively more pro-Arab and therefore, pro-Democrat and therefore less enamored with the truth. His brother became a radical spokesman for some terrorist front. And, it’s “special sauce”, not Zogby sauce.
You’re looking at a long outdated Fordham list. It was compiled just hours after the 2008 election, before the final numbers were available, and used an ESTIMATE of Obama’s win that was off by more than a percentage point.
The final Fordham report of pollster accuracy, based on the CORRECT election outcome, still has Zogby near the bottom of the list, but Rasmussen is no longer top of the heap. He was beat out by eight other pollsters.
He was 10th out of 16 in 2004. I don’t know how far back you want to go.
BTW I was kidding you, no need to get snotty, especially when you’re wrong.
I wasn’t being snotty. And what was I wrong about? You posted a url for the incorrect 2008 Fordham report and I provided the correct one. I wasn’t talking about 2004.
Sorry, no you were not. The post I directed at you was intenteded for Defiant.