McCain landslide shaping up?

It sounds way too good to be true, but some reports are showing McCain doing far better than mainstream polls are indicating.

Zogby reports that on Friday (only, no 3 day average) McCain surged into the lead, 48-47%, with undecided's beginning to break heavily for McCain - I have seen reports that the margin may be as high as 4 to 1.

There is also a poll out, supposedly from GOP internal polls, that indicates McCain having a larger than 20% lead in PA, an 8 point lead in New Jersey, a 10+ point lead in Michigan, and within a point in California. (link)

While this sounds the most far-fetched of anything I have seen yet, and I was ready to dismiss it mostly when I read it, I found some other links that seems to verify the trends it represents, if not the actual number themselves.

As well, Paul Marston (hats off to Casey's Critical Thinking blog for the link) predicts McCain will win by at least 42 electoral votes, due primarily to PUMA voters.

Finally, there was a post at redstate.com that supposedly was from a former Hillary Clinton staffer turned Barack Obama staffer.

In it she explains the 4 "true weaknesses" of the Obama campaign.

Here is a link for the full thing, but there are some interesting parts I would like to point out.

First, the four weaknesses themselves:

1. Hillary voters - apparently the Obama campaign has internal data suggesting they are only getting about 75% of them...
2. Sarah Palin - really threw them a curve ball when she was picked, they have fumbled it since the outset and are now seeing some negative reactions from it.
3. Obama's radical connections - apparently it doesn't play well with voters to see the full line-up of Wright, Ayers, Rezko, Khalidi, etc.
4. The "Bradley Effect" - they seem to think it is going to be as much as 10 points they have to accept as lost to it.

Now, there really isn't much new here from first glance, regardless of whether this is a real report, or a fake.

The post goes on to point out how many of the mainstream polls are receiving as high as 80% "refusal to respond" results - and that the majority of the RTR's are McCain supporters. Combine that with the already known issue of some polls taking a much higher sampling of Democrats as it is, and you can get some pretty skewed results.

Now, if that above statement is true, that could indicate the numbers being reported in the GOP internal poll are perhaps more accurate than not.

Again, I think the above results are a bit too good to be true, but they are interesting to think about.

Let's see what happens if we try to quantify and apply all these "biases" in the polls:

1. Assuming for the moment that the Bradley effect is real, we'll put it at 4 points (considering the peak to be about 10 points 20 years ago, and an apparent weakening of the effect since then).
   
2. Now assuming it is correct that up to 80% of people contacted for polls refuse to respond in some polls, and there is a bit in all, we will put it at about 50% average at least refuse to respond.
   
3. Next, assuming it is accurate that a majority of the refusals to respond are McCain supporters (ok, now how can they really know that?) let's put it at 60% for McCain - which means a total disregarding of about 10% of McCain support (20% more RTR for McCain for 50% total refusals...).
   
4. Finally, assuming Democrats are indeed sampled 15 points more than Republicans, we will subtract away 7 points for the approximate weighting of Democrats to Republicans as reported by rasmussen, that leaves us with an 8 point bias to Democrats.
   

Taking the math that way - rasmussen currently reporting a 51-46% lead for Obama - it would not be hard to get it to McCain having a 52-45% lead after factoring in just the Bradley effect and the sampling bias. If we also include the refusal to respond bias for McCain, it could be as much as 57-40%.

Now, all those numbers are pure conjecture on my part, but it is liberals who keep going on and on about the Bradley effect, not me. If we take my number for that effect alone, then the race is basically tied, with Obama having an insignificant 1 point lead.

So I think, for the purposes of trying to understand that data, that bit of math is at least good for showing how skewed the polls could/might be - whether they are really or not we will have to wait for Tuesday to truly see, but this just confirms for me that this is indeed very much McCain's race to win or lose now.

Will it be a landslide? I doubt it.

Will McCain pick up the states he needs to win? I think it is plausible, if not likely.

Go McCain!