The problem is not how the polls are weighted, as that remains for the most part static within a given poll. The problem is the MOVEMENT within that weighting. It is movement away from McCain, which suggests that undecideds are breaking differently than many here may have thought.
Of course, this is just one poll, but the movements in the other polls are similar. In a little over an hour IBD/TIPP will release their final poll, and perhaps that one will have it back down to a race within the margin (under 2%.) Let’s hope.
Actually weighting is an issue. They are assigning a higher +D based on assumptions.
I’m no pro but I love reading the variety of breakdowns from folks who know how to do such a thing.
After reweighting back to closer to historic norms, the numbers run a lot tighter.