This past year, I've been a frequent critic of how Charles Franklin weights his polls. I was proven wrong in the recall when the seven point margin exactly mirrored Franklin's final poll margin.
This time around, the Republicans recycled my weighting arguments in reverse, arguing that Franklin was weighting the poll too much in favor of Democrats.
Now another election has passed and once again Franklin nailed it. The final poll had Obama winning by an eight point margin and last night Obama won by seven points.
While all polls, with the exception of Rasmussen, had Obama winning, most had Obama winning by a smaller margin than the eventual result. Below is a ranking of the polls based on how close their final margin prediction compared to the actual result:
We Ask America - Predicted seven point Obama win / exactly right
Marquette Law School Poll - Predicted eight point Obama win / (+1) from actual result
St. Norbert (+2 from actual result)
Project New America (-2 from actual result)
YouGov (-3 from actual result)
PPP (-4 from actual result)
NBC/Marist (-4 from actual result)
Pulse Opinion Research (Rassmussen affiliate) (-6 from actual result)