Polling Post-Mortem

Mark Blumenthal has a nice piece on the final polling results compared with outcomes and his conclusions are basically along the lines of what we thought:

1) The Bradley Effect simply did not appear anywhere. The question of whether it ever existed is something that can be reasonably debated. But there is simply no question that the Bradley Effect did not exist here and we know that because ...

2) The polling was very, very accurate at both the national and the state level. If anything, the "undecideds" broke a little more towards Obama than McCain but they were about evenly divided.