Is this a normal election year? Part II

A couple of weeks ago, I wrote a blog post criticizing the conventional wisdom as being flawed. I thought that analyzing traditional voter models would give us a false sense of who is likely to vote, and that, therefore, the polls predicting the outcomes of this year's elections were faulty.

I also criticized Nate Silver, from 538.com, whom I admired for his success in predicting past races, because, although, he has in the past been one to ignore the conventional wisdom, I felt that, this year, he is only slightly modifying the analysis of the conventional models, rather than relying on better models like he had in other elections.

Yesterday, though, Mr. Silver wrote a great piece, explaining that not only are conventional models are flawed, but how they are flawed, and how that could actually mean a BIGGER Republican landslide than the conventional models have been predicting.

Here's the piece: http://fivethirtyeight.blogs.nytimes.com/2010/10/31/5-reasons-republicans-could-do-even-better-than-expected/

He also promises a follow-up piece explaining how this could mean that the Democrats could hold on and stave off a devastating tidal wave of Republican resurgence. I hope that comes today, as I'm very curious to read the other side of this analysis (no, not because I hope the Democrats are successful; I just really am curious). I find Mr. Silver's honesty refreshing, and I wish he and others had been as honest about this faulty polling and data analysis long before this weekend.