Princeton professor presents a meta-analysis of state polls. Compare and contrast with J. Scott Armstrong's site.

And see Henry W. "Chip" Chappell's "Presidential Election Forecasting Page" which, at least as of 10/26, was forecasting a Kerry victory, 282 to 256, even after awarding Bush both Ohio and Florida. That violates the conventional wisdom's Rule of Two--and I'd like to get a (small) bet down against it.

But also read Jim Webb on an underappreciated advantage of the Republicans: the Scots-Irish.

Finally, I recommend David Gelernter--no fool, he--on why he's voting for W.