Silver just got lucky when the results matched his biases a few times in a row. If there was a random or structural problem with polls, the errors wouldn’t all seem to benefit one side of the political spectrum.
Right, polls are utterly inaccurate. You may remember the 12 elections. Many conservatives came up with all kinds of inside baseball reasons the polls were completely wrong, understating the vote for Romney by 3 to 7%.
I fully expected a GOP victory. Then we held the election and dang if the polls weren’t in aggregate dead on. Most notably Silver, who pulled off an amazing level of prediction.
So don’t just tell me the polls are wrong, with the assumption they’re wrong in the direction you’d prefer. Not when actual elections keep proving them right.