Rasmussen Tracking Poll [Obama vs McCain]

[Sam Spade]

I expect Obama to be up about 2 tomorrow as a strong McCain night rolls off, and 3 Obama wins will be on the sample(I expect Obama to win tonight).

You have access to the internals, right?  What are they?

No, I don't personally. But, I know MAC is +5 with Indies today, after being +2 yesterday, and that the last three nights have been (most recent first):

Obama +2.69%
Obama +3.26%
McCain +5.33%

So the MAC night will fall off, giving Obama a lead of probably 2-3 points.

Well, thanks for the info.

I'm guessing that the Obama +2.69% of today replaced a similar number that dropped off today, right?

The one that fell off(Monday's daily result) was Obama + 1.55%.

Today's result in decimal form is Obama 48.00%- McCain 47.79%.

So, in other words, this table would be accurate for tomorrow:

McCain +4.46 or above = stays the same or better
McCain +1.46 to McCain +4.46 = Obama +1
Obama +1.54 to McCain +1.46 = Obama +2
Obama +4.54 to Obama +1.54 = Obama +3
Obama +4.54 or above = Obama +4 or more

I like all the info I can get.  Thanks. 

[Sam Spade]

Saturday- September 20,2008
Obama 48%
McCain 47%

Looks like McCain won last night by a few points, I'm waiting for the exact numbers and will post later.

Well, as my chart indicated yesterday, the result would have to be somewhere in-between M+1.46 and M+4.46 to get this number.

[Sam Spade]

A whole bunch of numbers could fit the pattern.

The difference is - as opposed to his Gallup estimations - that these are the actual numbers copied from Rasmussen internal crosstabs.

[Sam Spade]

Nice piece of news in todays Rasmussen release:

"Looking back, four years ago today, George W. Bush held a 49% to 47% lead over John Kerry."

The dynamics were different.

Very true. Conditions were fundamentally far more favorable for the incumbent party, thus there was no reason to really believe that Kerry was extremely likely to make a comeback, barring a change in those conditions.

Also, at this point in 2004, the Democratic numbers were depressed quite a bit through lack of enthusiasm (pre-first debate).  Of course, Rasmussen covers that up with weighting, but it would show up in the state polls.

[Sam Spade]

According to my math, this is my rough estimation as to what this week was.  The margin could be wider or narrower depending on what the sample for the first nine days of August was that dropped off, but I doubt it would be much wider or narrower...

DEM = (40.4%-41.0%)
GOP = (32.2%-32.7%)
IND = (26.6%-27.1%)

[Sam Spade]

That must have been a fairly strong "leaning" Obama sample, because the numbers w/o leaners moved from O47, M45 to O46, M46.  Or maybe one of those before it was...

Anyways, here are the new targets for tomorrow.

O+3.94 or above = Obama +2
O+0.93 to O+3.93 = Obama +1
M+2.06 to O+0.92 = Tie
M+2.07 to M+5.06 = McCain +1
M+5.07 or above = McCain +2

[Sam Spade]

All of these supposedly 'nightly' numbers don't jive with this posted on Rasmussen on their South Carolina Senate race.

"The South Carolina survey was conducted on Thursday night. Several state polls conducted that evening showed very positive results for Democrats and it was one of Barack Obama’s best recent nights in the daily Presidential Tracking Poll. It remains to be seen whether this was a temporary blip in the race or the beginning of a new trend leading to a significant advantage for the Democrats."

This sounds like bigger than a 2 point lead for Obama as stated in previous posts.  In any case, this night will drop off tomorrow so we will likely see a tied race tomorrow.

Rasmussen doesn't weigh his state polls to party ID, but weighs his national polls to it.  That explains the difference.

[Sam Spade]

Monday- September 22, 2008
Obama 48%(nc)
McCain 47%(nc)

So, where does this sample lie in the O+0.93 to O+3.93 range?

[Sam Spade]

BORING!

Why are Rasmussen's numbers so rigid?

It's called party ID weights.

[Sam Spade]

Obama won last night by 3.4%

So tomorrow...

O+4.98 or above = Obama +4
O+1.98 to O+4.97 = Obama +3
M+1.02 to O+1.97 = Obama +2
M+4.02 to M+1.03 = Obama +1
M+4.03 or above = Tie

[Sam Spade]

What makes this night different from all other nights?

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

Dunno, but remember always 1 in 20 and MOE issues.

Obviously, last night had to be M+4.03 to M+7.02 and when we have a feed to actual internals, we can solidify.

[Sam Spade]

Three night actual average:

Obama 48.07%-McCain 47.80%

Last night McCain won by 5.09%.

So the last three days(most recent first):

McCain +5.09%
Obama +3.40%
Obama +2.12%

Targets for tonight:

Obama +6.19% or above = Obama +2 or above
Obama +3.19% to 6.18% = Obama +1
Obama +0.20% to 3.18% = Tie
McCain +2.81% to Obama +0.19% = McCain +1
McCain +2.81% to McCain +5.80% = McCain +2
McCain +5.81% or above = McCain +3 or above

[Sam Spade]

In a just FYI mode, these are the state Rasmussen has not polled this month, and we can thus expect to see numbers out of before September 30 (maybe before the end of this week even), since he says all 50 states will be polled this month.  Kansas will definitely come out today.

Alabama
Arizona
Arkansas
California
Hawaii (haven't gotten any polls out of here so far from him)
Iowa
Kansas
Kentucky
Louisiana (hurricane-questionable)
Maryland
Massachusetts
Mississippi
Nebraska
New Hampshire
Tennessee
Texas (hurricane-questionable)
West Virginia

[Sam Spade]

Kansas will definitely come out today.

CA too.

And you know this how? 

[Sam Spade]

Kansas will definitely come out today.

CA too.

And you know this how? 

It's on their homepage.

Hmmm...  Must've missed that.  Guess I need new glasses (although I don't wear glasses). 

[Sam Spade]

Wednesday, September 24, 2008
Obama 49 (+1)
McCain 47 (-1)

Giant Obama sample last night, apparently.

[Sam Spade]

6, 7, 8, points?

According to my chart - Obama 6.19% to 9.18%

[Sam Spade]

6, 7, 8, points?

According to my chart - Obama 6.19% to 9.18%

so, it went from McCain +5 to Obama >+6 in the span of 24 hours?...time to start ignoring daily numbers, guys

I think this is why they have a three day average.

Way out of MoE.

It'll all correct itself in the long run.  But obviously, one or both of these samples screams outlier.

[Sam Spade]

Those one day totals are a total farce. The are not listed even on premium membership.  McCain did not win Sunday by 5 and Obama did not win yesterday by 6 or 7. 

By the way, McCain now leads in New Hampshire 49-47 and Palin is still more popular than Biden.

So, in other words you're saying that the one-day totals Rowan gives us are not on premium membership?  What say you, Rowan?  One-day totals can be outliers like this, but if I have to buy premium membership for myself to know for sure, I will.  

I see no reason to not believe the last two things you posted.

[Sam Spade]

By the way, McCain now leads in New Hampshire 49-47

That's one poll by UNH only.

The UNH poll was 47-45.  ARG's poll was 48-45  Either he's giving us Rasmussen's new numbers (which might be available to premium members), he's lying, or there is some other poll that will come out those numbers.

[Sam Spade]

Alright, I signed up for premium membership.  There are no daily numbers there.  If you're getting the numbers, Rowan, from somewhere else, their either guessing or attempting to use a complicated measure which I don't believe will work based on the data I'm getting.

Looking at the data, I'm guessing today's sample was pro-Obama, but not by that much.  The sample that drops off tomorrow I'm guessing should be fairly Obama, esp. among Indies.

I also have no state numbers, but pepper11 removed his remark, so he might have just mistaken that for another.

[Sam Spade]

If you goto the morning video it gives you a preview of the polls.

ah ha.  Thanks, I will remind myself to do that next time.

[Sam Spade]

Rowan, with leaners it's:

Obama 49% (nc)
McCain 46% (-1)

[Sam Spade]

Rasmussen with the Obama bounce this week.....finally.

Considering that there's a pro-McCain sample dropping off tomorrow, you could see the lead easily push up a bit tomorrow.  And then the day after that, the massive pro-Obama sample tracks off.  So, there will be some bumping

Still, today's sample was a bit more pro-Obama than the one that preceded it (i.e. more than 3.4% Obama, I'd guess somewhere around 5%), but it had more to do with movement from McCain to undecided than Obama gaining.

[Sam Spade]

The five state polls to be released this afternoon are consistent with a 2-3 point Obama lead.  More later.