Rasmussen: Romney expands lead to +2

[candycorn]

The odd thing about Rasmussen is this:

In 2008, Rasmussen got the election pretty close; they had Obama by 6 in their final poll, Obama won by 7.3.

Many people have cited that as the reason Rasmussen is reliable.

BUT:

In 2008, if you go back through the fall and summer, Rasmussen was in line the whole time with the other polls. Rasmussen was close to the average all that time.

NOW, however, in 2012, Rasmussen has been consistently well off the average, to the Romney side, for months. In sharp contrast to 2008.

Why????

It's very simple. Rasmussen samples using ONLY a likely voter screen, and it bases its sample on its own (very accurate) poll of party identification- in other words- they use the most up to date numbers. The other pollsters are using the 2008 party identification figures- numbers that skew the polls by D+7 or higher. Marquette University conducted a poll recently that used a D+11!! Reality will be a bitch come Nov 6th...

Yes it was.

 

[AceRothstein]

The odd thing about Rasmussen is this:

In 2008, Rasmussen got the election pretty close; they had Obama by 6 in their final poll, Obama won by 7.3.

Many people have cited that as the reason Rasmussen is reliable.

BUT:

In 2008, if you go back through the fall and summer, Rasmussen was in line the whole time with the other polls. Rasmussen was close to the average all that time.

NOW, however, in 2012, Rasmussen has been consistently well off the average, to the Romney side, for months. In sharp contrast to 2008.

Why????

It's very simple. Rasmussen samples using ONLY a likely voter screen, and it bases its sample on its own (very accurate) poll of party identification- in other words- they use the most up to date numbers. The other pollsters are using the 2008 party identification figures- numbers that skew the polls by D+7 or higher. Marquette University conducted a poll recently that used a D+11!! Reality will be a bitch come Nov 6th...

Sounds like Rasmussen needs to update the way that they calculate party ID.

 

[C_Clayton_Jones]

Rasmussen is pure evil. Everyone knows that Obama can't lose. The polls prove it, and everyone knows it is true. All that has to happen is Democrats turn out in the same numbers as they did in 2008 and we ignore 2010.......

Actually this is a very interesting and telling post.

It’s one of the contributing factors not often mentioned as to why republicans and conservatives were so convinced Romney would win.

Most on the right incorrectly inferred that the 2010 midterms were somehow a ‘portent’ of the General Election.

Needless to say they were wrong.

 

[candycorn]

The odd thing about Rasmussen is this:

In 2008, Rasmussen got the election pretty close; they had Obama by 6 in their final poll, Obama won by 7.3.

Many people have cited that as the reason Rasmussen is reliable.

BUT:

In 2008, if you go back through the fall and summer, Rasmussen was in line the whole time with the other polls. Rasmussen was close to the average all that time.

NOW, however, in 2012, Rasmussen has been consistently well off the average, to the Romney side, for months. In sharp contrast to 2008.

Why????

It's very simple. Rasmussen samples using ONLY a likely voter screen, and it bases its sample on its own (very accurate) poll of party identification- in other words- they use the most up to date numbers. The other pollsters are using the 2008 party identification figures- numbers that skew the polls by D+7 or higher. Marquette University conducted a poll recently that used a D+11!! Reality will be a bitch come Nov 6th...

Sounds like Rasmussen needs to update the way that they calculate party ID.

Wonder how that University of Colorado computer programmer is doing about right now? Remember...it had never lost an election.

 

[AceRothstein]

It's very simple. Rasmussen samples using ONLY a likely voter screen, and it bases its sample on its own (very accurate) poll of party identification- in other words- they use the most up to date numbers. The other pollsters are using the 2008 party identification figures- numbers that skew the polls by D+7 or higher. Marquette University conducted a poll recently that used a D+11!! Reality will be a bitch come Nov 6th...

Sounds like Rasmussen needs to update the way that they calculate party ID.

Wonder how that University of Colorado computer programmer is doing about right now? Remember...it had never lost an election.

How many times was that posted here? 1,000 times? 1,500 times? There was one poster in particular that had to have dreams about it, he posted a link to it so many times. Maybe he was one of those professors?

 

[Black_Label]

The odd thing about Rasmussen is this:

In 2008, Rasmussen got the election pretty close; they had Obama by 6 in their final poll, Obama won by 7.3.

Many people have cited that as the reason Rasmussen is reliable.

BUT:

In 2008, if you go back through the fall and summer, Rasmussen was in line the whole time with the other polls. Rasmussen was close to the average all that time.

NOW, however, in 2012, Rasmussen has been consistently well off the average, to the Romney side, for months. In sharp contrast to 2008.

Why????

It's very simple. Rasmussen samples using ONLY a likely voter screen, and it bases its sample on its own (very accurate) poll of party identification- in other words- they use the most up to date numbers. The other pollsters are using the 2008 party identification figures- numbers that skew the polls by D+7 or higher. Marquette University conducted a poll recently that used a D+11!! Reality will be a bitch come Nov 6th...

Yes it was.

Great bump, look at all the over the top stupidity from the right wing shitheads in this post

 

[mamooth]

If they had simply used the raw registered voter data, Gallup and Rasmussen would have been very accurate. However, Gallup and Ras both insisted on adjusting the data with a likely voter screen which, as history shows, just plain sucked ass. In general, the polls which did not apply likely voter screens did much better than those which did.

Gallup's likely voter screen was especially mysterious. Non-hispanic whites were 74% of voters in 2008, yet Gallup's screen assumed whites would be 78% in 2012. Huh? Whites ended up being 72%.