I'm not ambitious enough (at least at this moment) to work up specific predictions, but I think I've got a general rule that I want to announce and thereby test:
I predict that on average the vote for Republicans in tight races will be around 5 percent greater than the polls predict.
From this prediction, it follows that the (R) side will keep both houses.
Sunday palate cleanser: The Proms, 2004
38 minutes ago