Pollster.com

Why would anyone do a poll of RV when they know that LV is much more accurate.

I was wondering the same thing. The only thing I can think is that it's a little cheaper and faster to do RV versus LV. Not a good enough excuse, but that's all I've got.

This is CNN/Time. Maybe close races make for better news stories, which are what these outfits need to sell?

Bukama when you said this is CNN, all I could think of was James Earl Jones saying that, what ever happened to him??? Anyways using an RV screen right now has a bit of a dem tilt to them cause dems have a good registration advantage in all three states (yes even Kentucky). I'd trust these RV polls more if it was a presidential year where everyone shows up to vote.

I guess it depends on what likely voter screen you are using. While a registered voter poll my not be preciuse, a bad likely voter poll could be wildly off of the mark.

Likely voter screens are inaccurate this far out. You shouldn't take those seriously until October.

For what it's worth.

CNN/Time polled Obama/McCain in Florida in 2008 (they didn't poll California or Kentucky).

Their 9/14-16/08 poll, also of Registered Voters, had the race tied at 48/48.

They switched to likely voters in their next poll that month (9/28-30/08) and got a +4 Obama race (51/47). Their next two polls showed similar results: 51/46 (10/11-14/08) and 51/47 (10/23-28/08)

Obama ended up carrying Florida by 3 points (51/48). Looks to me like CNN/Time have the right idea.

Now, it's certainly possible that Likely Voters this year will be more favorable to Republicans. But it's too early to predict. And that's exactly what pollsters are doing when they use Likely Voter screens this far out from election night.

"....a bad likely voter poll could be wildly off of the mark."

I agree - there is nothing that fundamentally says that an LV screen makes for a more accurate forecast. Remember that a sizable number of registered voters who are disinterested will refuse to take part in the survey to begin with so there is always a small interest screen in every poll. Since response rates have been steadily declining for decades, that effect keeps getting stronger.

It is hard for me to judge the LV models used by different pollsters since they rarely make them public. Past experience makes the likely voter models used by Rasmussen and Gallup close to the election look pretty good. However - perhaps this election is "special" in ways that make these models biased. We won't know that definitively until November 3.

On the issue of RV vs LV polling, I don't think there will be a big differential in either CA or FL. In CA, there are two very high profile races, plus a bunch of initiatives that will draw voters to the polls in large numbers, and it is a very "easy voting" and "early voting" state. So turnout will be quite high.

FL also has two very high profile races, and with the huge number of retirees there, voting percentages are always very high. So I wouldn't see a lot of differential on the so-called enthusiasm gap.

KY is another issue and I don't know enough about that state to have an opinion. It may be that there would be a large differential in RV vs. LV, but I'm not one to be comfortable in reaching any conclusion there.

After the election, Senate will be:

Dems 54
Reps 44
Indies 2

Dems will hold the House as well.

The problem with LV screens is that they are meaningless. They are completely based on a guess by the pollster of who are likely voters. Two months or more out from election day, and not knowing what might happen in the intervening time, the screens are a joke.

The fact is, that with every registered vs likely voter model, the likely voter models have given better advantages to the GOP candidate. When both Boxer and Brown are under 50% in RV polling, and Rubio is ahead of Crist in RV polling, it's obvious that if the model was likely voters, the gop candidate would be in pretty good shape.

Now this poll I think is more accurate. If this had been KY or NC, and the Democrat had the lead, I would think they are wrong, but based on the state of CA, I know, this sounds about right. Survey USA, had made the Sacremento area seem like Alabama with about 70-30 for Fiorina; I don't buy it. There are a lot of progressives near Sacremento, and conservatives too, but I doubt even in that part of CA, it is a lot closer than the stupid Survey USA poll.

Just for the record, I remember after the 2008 Republican convention, NC had Mccain up by 18 over Obama and the next day I saw another poll that had Mccain up by 3. Obama wins the state by about a 1/2 percentage point. If they can be that off on a state like NC, am not buying their polls in Washington and CA.

I am still concerned about Meek being the spoiler in FL. Based on the average American polling, and FL would probably have fewer progressives than some other states you have about 19 percent Liberal
41 percent moderate
40 percent Conservative.

That makeup might move most moderates to Crist, and maybe 6 percent of progressives. This will be a close one, that is for sure.

I think that Crist may have to roll the dice, and hint he might possibly caucus with the Democrats. On the other hand, if he runs more like a Scott Brown or Olympia Snowe Republican, he may get a few conservatives, who like him as governor, and will stay loyal. I would have a tough decision to make if I were him, and which direction he goes, it will be a risk.

I think Crist will pull it out, and it might be the older Reagan Democrats who might pull through for him.

I think Crist will pull it out, and it might be the older Reagan Democrats who might pull through for him. I'm almost positive, that if Reagan were alive, and well he would endorse Crist too.

I wonder if Sink is helping Meek at all?

I think the people who are supporting Meek are the ones who think Crist is going to win anyhow, and have probably not looked at recent polls.

I think polls in primaries are helpful to me, and obviously many fierce partisans don't look at hypothetical matchups. I am likely to usually support the candidate in the Democratic primary who polls the best against the Republican. Obviously the GOP in Alaska were not thinking of that, and same with Sharron Angle. I was delighted when Angle won, because both Tarkanian, and Lowden who were more mainstream had huge leads against Reid in the general election. Wouldn't I love to see the radical candidate win the NH GOP primary. It would almost assure that Paul Hodes is going to be Senator Hodes.

I don't believe the KY race is that close.

Hey lat, both Palin and Romney came out against the Koran burning today. (As if there was any doubt). Happy now?

As a republican, I have to say I couldnt care less about Paul. His opposition to the liberation of Iraq is a disgrace. It would be nice to see more republicans in office but other than that I couldnt care less if he loses.

I think Conway may still have a shot. Now the biggest thing Conway is against, is federal funding for drug enforcement or treatment, and awareness programs for high school kids. You could bet Nancy Reagan wouldn't endorse Paul, and not sure if Ron would either. What happened to the party that founded the DEA and "Just say no".

Except for Kentucky, these numbers make sense.

"I wonder if Sink is helping Meek at all?"

I think it's the other way around.

If I lived in Florida, I'd have a hard time voting for Crist. I'd probably wait till the last polls, or even election day, before deciding to do so.

I agree with Paleo. Meek is going to pull the black vote. Had Meek not been on the ballot they wouldn't come out in decent numbers. They will come out to vote for Meek but it will pull them out for Sink as well.

Christ won't get any of the black vote which is why he won't have enough steam to pass Rubio.

I've got to agree with the wingers here, sad to say, on LV vs RV. Gallup makes their transition in the coming week or two. There's been far too much evidence for months now that Dem enthusiasm lags far behind repubs' enthusiasm and it's foolish to think that this won't depress the Dem vote to some degree.

Additionally, as a partisan I've started to notice left wing sites like Daily Kos put aside their foolish purity-based diaries and posts criticizing Obama, Reid, & co. from the left and begin to focus on the battle ahead. If all polling firms were like CNN and we were getting a constant stream of results like these I seriously doubt this would happen.

Dems need a real wake-up call; Ras and, of late, PPP (and SUSA, this year) have provided the jolt.

Now we just need a highly energized leadership delivering a much more populist message, and going on the attack.

Obama's public rejection of Boehner's offer and specifically his refusal to compromise at all on ending tax cuts for 250k+ earners has begun to bring the left back, a left he was losing badly just a few weeks ago.

We also saw that from Obama today in Cleveland. If it took all this to finally bring Obama back into campaign mode it's almost worth the suffering we Dems have endured.

Obama's speech skillfully intertwined populist rhetoric with scathing contempt for both the strategy and tactics of the other side. It's a very tricky path but if Dems can make a strong case that Republicans are directly harming the middle class through their actions and if they can show that there's nothing at all that's new in Republican policies (other than what was previously attempted and which failed miserably); in other words if they can make the case that Republicans want to implement the same policies which caused real income for the middle class to go down while their costs skyrocketed they won't necessarily look like shining knights but they will certainly knock their opponents down a few pegs.

If the public not only hates both parties (which it does in every opinion poll) but see no hope in pursuing Republican policies which have previously failed that will bring the likely voter makeup back closer to the RV one. Moving it just 3-4 points will make all the difference in the world.

The other effect of turning populist is that it revs up the base...that's what I'm now seeing at left wing blog sites; much less "Obama didn't give me my pony" posts and far more calls to action, money bombs, GOTV-oriented diaries and rah rah ones, without the usual sniping by the above-mentioned purity trolls. We need these self-absorbed folks to phone bank and hit the streets, and it has to start now. We need more people to give money, etc. So, I think this was the right time for Obama and, I assume, national leadership to get tough and aggressive; to hit hard, to make no apologies, to offer no compromise.

MikeyA:

"I agree with Paleo. Meek is going to pull the black vote. Had Meek not been on the ballot they wouldn't come out in decent numbers. They will come out to vote for Meek but it will pull them out for Sink as well.

Christ won't get any of the black vote which is why he won't have enough steam to pass Rubio."

He's polling in the 30s with AAs right now. Why do you think that will drop off in the future. I'd say AAs, just like whites, have this same dilemna facing them; do I abandon a real Dem, a liberal, someone aligned with my own ideology, to vote for a guy who, a few months ago was touting his conservative credentials?

The answer for about a third (and for me at this point too, unless Meek goes on a real tear and Crist's money dries up, etc.) is to abandon Meek, who they think has no chance, and vote for the lesser of the two evils. If Meek continues to poll in the mid twenties we may just see Crist get desperate enough to announce that he'll caucus with Dems. There's just no path for him if people think he's stick with the Republicans. He needs too many moderate and liberal Dems.

Can Paul really win by simply keeing his mouth shut? Maybe this poll is showing that isn't working. Problem for Paul is that if he does open his mouth and say what he really thinks, he is too far out there even for conservative KY voters, who have a very appealing center right Democrat in Conway as an alternative.

The difference with the Raz poll is striking. Raz was conducted on one day (9/7) and involved 500 respondents. CNN was over 5 days (9/2-9/7) and involved 869 respondents. LV v RV also an already noted difference. Future polling of this race is going to be fascinating.

Just checked out the details of the CNN polling on KY. There were no "non white" respondents? No respondents 18-34? Am I misreading this???? Help!

If this is true it means that the pool of respondents was defintely tilted towards Paul yet the race is still tied. This is hard to believe even for me.

"There were no "non white" respondents? No respondents 18-34? Am I misreading this???? Help!"

VW: You are misreading the crosstabs. For example the white respondents were 43% for Conway and 48% for Paul (as opposed to the 46% to 46% tie for all respondents). Thus the "non-white" respondents clearly went solidly for Conway. The pollster simply didn't list those percentages separately.