An advice on Senate Polls Average.

[Senator Cris]

An advice, please
I'm doing an average of the polls for every senate race but my doubt is: it is right including internal polls?
An example: for Arkansas, with the internal poll (Pryor 45% Cotton 39%) the average is Cotton 48,7% and Pryor 48,7% (to be precise, it is Cotton 48,69% and Pryor 48,66%). Without the internal poll, the average is Cotton 49,1% and Pryor 48,5%.
What do you think?

[Free Bird]

I would maybe do both. It also depends on what you call an internal. If RCP is to be judged, everyone's precious PPP would have to be gutted out of the equation.

[Smash255]

Most internal polls I think should be excluded.  However, I would include PPP simply due to their track record.  PPP from time to time will do polls for liberal groups, those polls I would not include.

[Free Bird]

Most internal polls I think should be excluded.  However, I would include PPP simply due to their track record.  PPP from time to time will do polls for liberal groups, those polls I would not include.

Other internal polls have good track records at times. I would exclude it completely, even if it is a sacred cow to this forum.

[7,052,770]

Most internal polls I think should be excluded.  However, I would include PPP simply due to their track record.  PPP from time to time will do polls for liberal groups, those polls I would not include.

Other internal polls have good track records at times. I would exclude it completely, even if it is a sacred cow to this forum.

The only reason people here like PPP is because it has been very accurate in the past. If that weren't the case, everyone would hate it.

[IceSpear]

The problem with internals is that they run the gamut from "highly accurate and right all along when nobody else was" (i.e. Harry Reid's in 2010), to complete and utter bullcrap (i.e. Eric Cantor's showing him up 40 points). You're better off just excluding them entirely.

I wouldn't include quasi-internals either. For example, PPP and Magellan are okay, but PPP being commissioned by the John Walsh campaign or Magellan being commissioned by the NHGOP are not.

[Free Bird]

The problem with internals is that they run the gamut from "highly accurate and right all along when nobody else was" (i.e. Harry Reid's in 2010), to complete and utter bullcrap (i.e. Eric Cantor's showing him up 40 points). You're better off just excluding them entirely.

I wouldn't include quasi-internals either. For example, PPP and Magellan are okay, but PPP being commissioned by the John Walsh campaign or Magellan being commissioned by the NHGOP are not.

That's why I would exclude them, period. If they have been commissioned even once, ignore them.

[Free Bird]

Is this really an honest pollster?

http://washingtonexaminer.com/nate-silver-calls-out-ppp-for-liberal-bias-on-gun-control-poll/article/2535603

http://www.newrepublic.com/article/114682/ppp-polling-methodology-opaque-flawed

[free my dawg]

>citing conservative sources crying liberal bias

Math is not a vast left-wing conspiracy.

[Free Bird]

>citing conservative sources crying liberal bias

Math is not a vast left-wing conspiracy.

The lower article employed actual information. Not skewed to my knowledge. You skew the reality of it to show it skewing reality. I love circular logic.

And how has Nate Silver ever been a conservative darling for calling something genuinely off-putting out. If he called out Rassy or that article defamed Rassy, you would love it. Rassy sucks, too, by the way.

[free my dawg]

>citing conservative sources crying liberal bias

Math is not a vast left-wing conspiracy.

The lower article employed actual information. Not skewed to my knowledge. You skew the reality of it to show it skewing reality. I love circular logic.

And how has Nate Silver ever been a conservative darling for calling something genuinely off-putting out. If he called out Rassy or that article defamed Rassy, you would love it. Rassy sucks, too, by the way.

Ah, gotcha. You're partially right - New Republic is actually liberal - sometimes I get that confused with freep, so that might have some problems. The Washington Examiner is actually a conservative source - I've read the firsthand source from Nate Silver and it was more concerned with them using "feelings" on whether to release the poll or not. However, the fact that he still includes them in his average, and even Erick Erickson of all people defended PPP should show that PPP's generally more well-respected than Rasmussen.