If you don’t like the projections of poll aggregators like Sam Wang or Nate Silver, blame the polls, not the aggregators.
In Wang’s case, for example, he has *completely* documented his inputs and methods - and his results mirror the polls, that’s for example why Romney’s chances improved dramatically after the first debate and them declined again afterwards.
Nate Silver’s methods area bit less transparent as he is also including factors such as general economic performance.
All the mainstream aggregators acknowledge that it’s *possible* that there is systematic bias in the state level programming, for example Nate Silver states that it’s exactly this possibility that creates the remaining probability of a Romney EC victory.
The Pundoids across the political spectrum hate the aggregators, because the former are in the business of selling their subjective opinions, and if the aggregators continue to be as accurate as they have been in the last few electrons (which is much more accurate that the average pundit), the pundits will be out of a job (or at least potentially less well paid to do it).
Not going to agree with you there. Aggregation does nothing to make polling results more accurate, all it does is allow the "aggregator" to stake out a middle-of-the-road claim that will of course place them closer to the end result than roughly half of the individual pollsters, whom the aggregator will use highlight his "success" (along with a wide enough MOE to account for reality to be different than anyone's guesses).
It is quite true that it's GIGO (garbage in, garbage out) as well -- and the input polls are garbage, even more so than usual.
San Wang’s final prediction:
And... before you shoot the piano player... remember that the aggregators are the *messengers*, the state-level polling writes the message.