Rasmussen: "core states" leaning to Obama

Because it's Rasmussen, the link to this poll might not work in the future. Scotty likes to make it difficult for people to check up on his previous predictions. Still, Rasmussen released a new poll (500 LV, ±4.5% MoE, 3/3 - 3/8) where he finds the "core States" (FL, NC, OH, and VA) leaning toward Obama 46/42. Hardly a surprise, that's exactly within the margin of error (MoE) for the survey. Also unsurprising is that this poll has major flaws:

First, a sample of 500 spread across four states is virtually meaningless. 500 would be a barely respectable sample size for each of the four States. I'm a little bewildered how Rasmussen arrives at a mere 4.5% MoE with that sample size. 

Second, it's a bit suspicious that Rasmussen was in the field for five whole days getting these 500 voters. 

Third, Rasmussen left PA off of his list of "core States", though he previously found PA to be a virtual tie.

Rasmussen avoids directly contradicting other polls of OH and VA that show Obama doing well by lumping his results into one poll. With other pollsters also showing FL and NC as a tie, Rasmussen also avoids shedding any light whatsoever on those States. This poll is not only meaningless and suspect; it's also worthless. 

These "core States" are absolutely crucial for the GOP to win. It's highly likely that losing any single one of them could produce a win for Obama in the general election. Any poll showing Obama leading in all of them would be death for the Republican chances. Is it any wonder Rasmussen combined his results into one meaningless composite, since that is the case? 

Here is the result that Rasmussen allows those unwilling to pay him to view:

President Obama now holds modest leads over Mitt Romney and Rick Santorum in combined polling of key swing states Florida, North Carolina, Ohio and Virginia. The numbers mark a shift from late February when Obama was tied with both candidates in the four states. 

Obama is now ahead of the former Massachusetts governor 46% to 42%. Six percent (6%) prefer some other candidate in this matchup, and six percent (6%) are undecided. (To see survey question wording, click here.)

The survey of 500 Likely Voters in Florida, Ohio, North Carolina and Virginia was conducted on March 3-8, 2012 by Rasmussen Reports. The margin of sampling error is +/- 4.5 percentage points with a 95% level of confidence.

Once again, we find the defenders of the Republican narrative to be settling for a tie. Since the poll results are within the MoE, it could be that Romney breaks even. Guess what partisan Republicans are going to believe? I'm not really sure how a tie supports their delusions that Obama is a hated figure who could be beaten by "anyone", but it seems to represent the limit of Rasmussen's willingness to manipulate his figures. Of course, the Right "knows" that the 6% of "undecided" voters is going to break for Romney, so the MoE is just icing on the cake of self-delusion. 

The important thing to note here is not the actual poll results. It is, instead, the delicate dance that Rasmussen is forced into doing to counter the growing evidence that Obama can reach 270 EV in 2012. Scotty concedes that Obama has improved by four points in a few weeks, while he also concedes that his poll results are so shoddy that they could indicate a tie. The mere fact that Rasmussen has to do this dance is further evidence that the Republicans are in trouble with their electoral votes. 

We are seeing the narrative of the Right shift from "anybody could beat Obama" to "Obama can't close the deal". That's huge. I'll take that. 

With the most Right-leaning national pollster in America slowly being backed into the "outlier" category, Rasmussen could find that Fox News is the only network willing to take his numbers seriously.