RE: "Why not Predict" and your conitnued use of this imaginary and sadly CW "Bradley Effect." When you talk about polls undercounting Hillary's numbers, you are selectively choosing evidence. In five minutes on pollster.com I was able to check some of the bigger states that Obama has won (VA, WI, GA, SC) and the polls undercount HIS numbers by substantial margins (much bigger than they undercount hers). Most of these are in the south and look a lot like NC, but more importantly, they point to momentum at the end of the race, and not any other possibility, as the explanatory factor in the undercount of the polls. In states she wins (and where she had last minute momentum) her poll numbers are lower than the outcome, and vice versa in states he wins. This means that there seems to be no consistent or general evidence of a Bradley effect, and there is no reason to generalize that just because the polls have unercounted her numbers recently (in PA and OH) that this will continue to happen today.

NC - he wins by +10

IN - she wins, but it is a lot closer than people expected (within 5 points). He has the momentum for the undecideds in the last few days.

Thanks! I was finally right about something. Wait. Two things! I was finally right about TWO things! My wife will be surprised.