Internal polls

[Grad Students are the Worst]

I would not have a huge problem with Options 2 or 3.  However, under the current system -- where the polls get functioned into the average and are only indicated in the description -- I don't want them.

If anyone isn't familiar with the black hole that is internal polling, please ask.

Opinions?

(I'm really more interested in written opinions than poll votes, BTW)

[Aizen]

No, why would any sane person actually want to add them? It's truly regrettable that some people feel the need to add anything resembling a poll regardless of its validity. I guess it gives them their jollies.

[Vasall des Midas]

Maybe in races where there aren't any recent decent polls. But we don't have a database of house races anyways, so there aren't many such.

[Tender Branson]

No, I won´t add them.

[Senator Ben]

We should add as many polls as possible.  Just because it's an internal doesn't mean it's wrong.

[DownWithTheLeft]

Polls all together at this point mean nothing, so who cares?

[Grad Students are the Worst]

We should add as many polls as possible.  Just because it's an internal doesn't mean it's wrong.

Investigative journalism:Press release::Poll:Internal poll

That's all I can really say to that.  It would be like posting press releases on the AP wire.s

[Wiz from Wis in Mass]

No. Polls need to come from third parties, the more objective, the better. Otherwise, we could be adding polls with ridiculous question wordings, orderings, etc.

[Inks.LWC Supports Chuck Hagel]

We should add as many polls as possible.  Just because it's an internal doesn't mean it's wrong.

I agree.  Dave has never said anything about entering internals, nor ARG polls.

And where would we draw the line?  Research 2000 does polls for DailyKos, and they clearly have a bias, so should we ban Research 2000 polls too?

[Senator Ben]

No. Polls need to come from third parties, the more objective, the better. Otherwise, we could be adding polls with ridiculous question wordings, orderings, etc.

If it's an absurd poll, we just won't add it.  I trust the judgement of the posters here.

[Sam Spade]

No.

[All life is a blur of Republicans and meat.]

No. Polls need to come from third parties, the more objective, the better. Otherwise, we could be adding polls with ridiculous question wordings, orderings, etc.

If it's an absurd poll, we just won't add it.  I trust the judgement of the posters here.

This is goldmine material.

Option 3 (normal), by the way.

[Prez Duke]

If we add ridiculous Uni polls, polls sponsored by the DailyKos, and ARG, then we should add internals.

[Grad Students are the Worst]

We should add as many polls as possible.  Just because it's an internal doesn't mean it's wrong.

I agree.  Dave has never said anything about entering internals, nor ARG polls.

And where would we draw the line?  Research 2000 does polls for DailyKos, and they clearly have a bias, so should we ban Research 2000 polls too?

We have clear indications that DailyKos does not futz with methodology.  That's a clear place to draw the line.

[Sam Spade]

We should add as many polls as possible.  Just because it's an internal doesn't mean it's wrong.

I agree.  Dave has never said anything about entering internals, nor ARG polls.

And where would we draw the line?  Research 2000 does polls for DailyKos, and they clearly have a bias, so should we ban Research 2000 polls too?

We have clear indications that DailyKos does not futz with methodology.  That's a clear place to draw the line.

time will tell on that one...

[Grad Students are the Worst]

We should add as many polls as possible.  Just because it's an internal doesn't mean it's wrong.

I agree.  Dave has never said anything about entering internals, nor ARG polls.

And where would we draw the line?  Research 2000 does polls for DailyKos, and they clearly have a bias, so should we ban Research 2000 polls too?

We have clear indications that DailyKos does not futz with methodology.  That's a clear place to draw the line.

time will tell on that one...

Actually, my asking Research2000 will tell on that one

[Sam Spade]

We should add as many polls as possible.  Just because it's an internal doesn't mean it's wrong.

I agree.  Dave has never said anything about entering internals, nor ARG polls.

And where would we draw the line?  Research 2000 does polls for DailyKos, and they clearly have a bias, so should we ban Research 2000 polls too?

We have clear indications that DailyKos does not futz with methodology.  That's a clear place to draw the line.

time will tell on that one...

Actually, my asking Research2000 will tell on that one

R2000 won't lie to you will they?

[Grad Students are the Worst]

We should add as many polls as possible.  Just because it's an internal doesn't mean it's wrong.

I agree.  Dave has never said anything about entering internals, nor ARG polls.

And where would we draw the line?  Research 2000 does polls for DailyKos, and they clearly have a bias, so should we ban Research 2000 polls too?

We have clear indications that DailyKos does not futz with methodology.  That's a clear place to draw the line.

time will tell on that one...

Actually, my asking Research2000 will tell on that one

R2000 won't lie to you will they?

Insofar as I'm still not convinced Scotty Rasmussen isn't pumping them out for Michael Badnarik

[Inks.LWC Supports Chuck Hagel]

Dave, were you going to weigh in on this?

[Dave Leip]

Hi,
please no internal polls.  this database is for scientific polls done by private companies, not conducted by campaigns.  If i could get some time, I want to create a weighting system based on past accuracy that would help with the "quality" issues of some of these companies.
Thanks,
Dave

[Inks.LWC Supports Chuck Hagel]

Thank you Dave!

[Gustaf]

Hi,
please no internal polls.  this database is for scientific polls done by private companies, not conducted by campaigns.  If i could get some time, I want to create a weighting system based on past accuracy that would help with the "quality" issues of some of these companies.
Thanks,
Dave

That would be awesome. And some of the rest of us should be able to help out. The poll system here now works ok but there is room for improvement. One way could be to use average error over the last 3 elections or something along those lines. So for this year the average results for 2002, 2004 and 2006 could be used. But that's only an idea.

[Lunar]

The problem with internal polls it that the fact that they are released is self-selecting.  Anytime one in twenty senators, governors, or whatever, hit an abnormal result that could generate excitement and contributions they are encouraged to release the poll.  But if the poll is in line with other polls (or expectations) of the state, they are encouraged to not do so.

So while the poll could be perfectly conducted, it is still flawed.  That'd be like if Gallup only released polls that were +5 McCain or better and Rasmussen only released ones +5 Obama.  If we knew this to be true, we wouldn't be liike "oh, lets just accept them both because they are solid firms and the more polls the better."

The main reason why this is even debated is inherent in the logic above: we want internal polls to be true because wheen we see them they are a hell of a lot more exciting, on average, than Scott Rasmussen or even Mr. Zogby.