Battleground Tracking Poll

[Sam Spade]

Battleground had kinda acted like they were going to do a tracking poll last week, and I believe someone had posted one of the results here, but I was leery that they were actually going to start.

With them actually posting results today, I see no reason why they're not going to start, except that they don't poll on Friday/Saturday and last week didn't poll on Monday/Tuesday.  Nonetheless, we have this language:

The firms are currently conducting a rolling national tracking poll with N=200 interviews per night on Sunday night through Thursday night.

Usual caveats apply - 200 per day means strong swings and this poll does have a slight GOP lean, imho.  But it does have a pretty good track record.

Today:  McCain 48, Obama 47, Undecided 6
Last Friday:  McCain 47, Obama 47, Undecided 6
Last Thursday:  McCain 48, Obama 46, Undecided 6

http://www.tarrance.com/files/2-way-ballot-trender-9-22.pdf

[Rowan]

It has a slight GOP lean because it uses a hard weight of roughly D+3. They have a hard time believing the electorate changes all that much in four years, and frankly so do I.

[minionofmidas]

Unsticky. Or sticky DailyDemocraticHackery.

[Nym90]

The rule is that every tracking poll being included in the RealClearPolitics averages gets stickied.

It's also worth mentioning that the sticky is primarily to assist those who will be posting daily updates of the numbers, not to confer any degree of legitimacy to the results posted therein.

[minionofmidas]

The rule is that every tracking poll being included in the RealClearPolitics averages gets stickied.

Since when is RCP a legitimate source?

[Nym90]

The rule is that every tracking poll being included in the RealClearPolitics averages gets stickied.

Since when is RCP a legitimate source?

Touche.

Still, I think most would agree this is a more legitimate poll than DailyKos.

[Sam Spade]

Unsticky. Or sticky DailyDemocraticHackery.

Careful now, because if history is our basis, it deserves more to be stickied than Diageo for example (what is its record?).  

Though someone will always be happy to bump Daily Kos up, whether it's me or one of their members.

Truthfully, I have no problem with the Daily Kos thing being stickied.  We're smart enough, hopefully, to be able to make the distinction.  My problem is putting internal polls in the database.

[minionofmidas]

The rule is that every tracking poll being included in the RealClearPolitics averages gets stickied.

Since when is RCP a legitimate source?

Touche.

Still, I think most would agree this is a more legitimate poll than DailyKos.

Hardweighting to just a 3% Dem advantage? Yeah, right. If R2000's partisan weighting is partisanly motivated, then so is this.

[Nym90]

The rule is that every tracking poll being included in the RealClearPolitics averages gets stickied.

Since when is RCP a legitimate source?

Touche.

Still, I think most would agree this is a more legitimate poll than DailyKos.

Hardweighting to just a 3% Dem advantage? Yeah, right. If R2000's partisan weighting is partisanly motivated, then so is this.

True. Ok, I stickied DKos now too. The rabbits told me to do it.......

[Eraserhead]

I can't wait for the ARG tracking poll!

[minionofmidas]

I can't wait for the ARG tracking poll!

By 2012 at the latest... possibly by next week.

Yeah, if these tracking polls keep sprouting all over the place, it might make sense to unsticky all but Gallupsen again... not that I absolutely trust these, but at least they have a much longer track record...

[Aizen]

I only really care about Rasmussen and Gallup. I find this poll and the DailyKos one useless.

[The Vorlon]

The rule is that every tracking poll being included in the RealClearPolitics averages gets stickied.

Since when is RCP a legitimate source?

Compare the final Battleground poll in each of the last 4 presidential elections to the actual results.

That will answer rather clearly why this poll deserves to be included in the averages.

We can argue a point or two here or there regarding the weighting, (I would call it a point or two GOP friendly, but I have been wrong before....) but this is a very solid poll run by two very well respected pollsters (One a Democrat, one a Republican)

As with all polls, you need to be able to trust the trendlines, even if the absolute number may bounce around a point or two...

The Terrance Group (Eg Goas) and Celinda Lake are well known and respected pollsters

'nuff said

[Rowan]

This is a BIPARTISAN and reliable poll.

[CARLHAYDEN]

A very good polling organization.  I'd trust their numbers more than Gallup, and perhaps a little more than Rasmussen.

[mypalfish]

What was the Rep/Dem breakdown in the 2006 elections?   Wouldn't that be the worst case scenario for the GOP?  That was the most depressed and unmotivated I can remember our local headquarters being.

[Sam Spade]

What was the Rep/Dem breakdown in the 2006 elections?   Wouldn't that be the worst case scenario for the GOP?  That was the most depressed and unmotivated I can remember our local headquarters being.

The CNN national exit poll for the HoR was Dem 38, GOP 36, Ind 26.

http://www.cnn.com/ELECTION/2006/pages/results/states/US/H/00/epolls.0.html

Rasmussen's party weights at that point were around Dem +6.

[big bad fab]

I only really care about Rasmussen and Gallup. I find this poll and the DailyKos one useless.

I agree.

But just have a look each day on "other national trackings" only for their trend, not for their absolute level: this is why it can be interesting, just to confirm or not the trend of Gallup and Rasmussen.

[Sam Spade]

Tuesday, September 23, 2008
McCain 48% (nc)
Obama 46% (-1)

[Rowan]

Very good news. Shows that when you use a realistic party ID, you get realistic results.

[Brittain33]

What was the Rep/Dem breakdown in the 2006 elections?   Wouldn't that be the worst case scenario for the GOP?  That was the most depressed and unmotivated I can remember our local headquarters being.

I don't believe that's a worst-case scenario, no, because my understanding is that Democrats, Independents, and Republicans all turned out solidly for their candidates, but the second group voted heavily with the first. This is why Republicans held their own in the southeast, a few exceptions aside, and also held the ordinarily safe Republican open seats.

That said, I can't imagine a worse scenario transpiring this year, although the Republicans do have the potential to lose even more Senators than they did in '06 if two of Oregon, Minnesota, and North Carolina tip to the Democrats.

[mypalfish]

Tuesday, September 23, 2008
McCain 48% (nc)
Obama 46% (-1)

I'd love for this result to be right, but we know it's off.

A question for any experts out there...what happens when a polling firm like this get a result that is out of whack with every bit of other polling out there.  Do they know they screwed things up and try to make corrections in the next sample?

[Sam Spade]

What was the Rep/Dem breakdown in the 2006 elections?   Wouldn't that be the worst case scenario for the GOP?  That was the most depressed and unmotivated I can remember our local headquarters being.

I don't believe that's a worst-case scenario, no, because my understanding is that Democrats, Independents, and Republicans all turned out solidly for their candidates, but the second group voted heavily with the first. This is why Republicans held their own in the southeast, a few exceptions aside, and also held the ordinarily safe Republican open seats.

That said, I can't imagine a worse scenario transpiring this year, although the Republicans do have the potential to lose even more Senators than they did in '06 if two of Oregon, Minnesota, and North Carolina tip to the Democrats.

Unless you think Uncle Ted's going to be acquitted, you only need one of the three.

Of course, I suspect that if the Dems get more than 5, they'll kick Lieberman out, so maybe your numbers are right.

[Brittain33]

Unless you think Uncle Ted's going to be acquitted, you only need one of the three.

Republicans lost 6 seats in 2006: RI, PA, OH, MO, MT, VA. In 2008, VA, NM, NH, CO, and AK make 5. To lose more seats than they lost in 2006, the Republicans would need to lose seven, i.e. two more from OR, MN, and NC.

[ChrisFromNJ]

Very good news. Shows that when you use a realistic party ID, you get realistic results.

LOL. You're a far right hack. Change your avatar to blue.