What the heck is up with Rasmussen?!?

[Firebirds01]

I check RCP and every poll has Obama up in Ohio except Rasmussen.

Same with the following swing states:

Wisconsin
New Hampshire
Iowa
Colorado
Virginia


Are these guys party hacks? I thought they were reliable. They seem to be quoted often.

[WI_DEM]

1. Why yes, yes they are party hacks.

[mzteaze]

2. Somebody....

...has to work as the "hype man" for Robme. Guess Rasmussen is it.

[timlot]

3. Used to be the Fox News/Rasmussen Poll.

[Firebirds01]

7. Really.....

well that explains a lot. Thanks

[woolldog]

12. Fox dropped them b/c they were not accurate. nt

[Hawaii Hiker]

4. Rasmussen is a fraud, he tells the GOP what they want to hear

http://www.huffingtonpost.com/alan-abramowitz/the-rasmussen-difference_b_2030330.html?utm_hp_ref=@pollster

[fugop]

5. Where have you been?

Yes, they have a well-known Republican bias.

[John2]

6. Scott Rasmussen

has gone all in for Romney this cycle. The difference is, he has very unknown new Republican Pollsters like Gravis Marketing to support his numbers. It seems they all work in concert. I think Axelrod is right on target. We will know what is there on Election day. My prediction is the biggest fraud they are pretending about my own home state. The only Polls all cycle showing North Carolina is not a tossup throughout the cycle are mostly or all Republican Pollsters. Public Policy Polling even though they lean Democratic has got to be very incompetent to show all 28 of their Polls a tossup in North Carolina. They have done the most Polling in North Carolina and all of them are consistent. I think they have got it right. North Carolina is probably the closest swing state of all and no candidate has a significant lead in the state. It depends on turnout.

[pointsoflight]

8. Definitely hacks. But by showing a narrow Romney margin...

...they're just going to help us GOTV.

[SoFlaJet]

9. The ONLY One That IMO Matters

http://fivethirtyeight.blogs.nytimes.com/

[20895DEM]

10. All Robo-calls and only done on a Sunday

No live calls
Sunday (people are either at church or watching football)
No internal cross-tabs so I am unclear if these are all +50 year old

The only number that seems consistent is the early voting split. Thus far I have seen the following over multiple weeks and multiple pollsters

Percent voted 11% 18% 21% 23% 26% 26% 36% 32%
Obama 59% 57% 66% 54% 58% 58% 63% 62%
Romney 39% 38% 34% 39% 39% 32% 36% 36%

Obama E.V. Adv. 20% 19% 32% 15% 19% 26% 27% 26%
Election Adv. 2.20% 3.42% 6.72% 3.45% 4.94% 6.76% 9.72% 8.32%

[Jennicut]

11. Rasmussen is a robo poller, as is PPP and Survey USA. Rasmussen's issues are

with party ID in the OH poll. Showing it even when almost every other Ohio poll show it with Dems having a slight advantage to more like 2008 (D+8).
They also do a Party ID screen per month for the tracking poll. Most pollsters to it every time they poll if they weight by party ID. And Repubs talk about pollsters cooking the books for Dem-friendly polls. They take the cake on it.

[woolldog]

13. which pollsters weight by Party ID?

The only one I know of is Rasmussen.

It's not an acceptable practice, AFAIK

[Bleacher Creature]

14. Nothing wrong at all.

He is totally predictable.

[gravity]

15. It's part of their business model and marketing strategy.

If you are the only poll quoted by right wing media, then you get all that traffic to yourself.

Their main challenge is to make sure that they seem scientific enough for the mainstream media to take them seriously too.

[TroyD]

16. What's odd about Rasmussen today

Is that he showed a decline for Romney nationally for the second day in a row, and yet somehow Romney went up in Ohio.

Romney went from (+4) nationally on Saturday to (+3) on Sunday down to (+2) today.

Meanwhile Romney supposedly went up 2 points in Ohio.

[blazeKing]

17. He's starting the move towards the real results

Which is Obama leading. So he can still claim after the election that he was the most accurate.

[Azathoth]

18. He's not a pollster. He's an entertainer catering to a particular market. n/t