RCP has the most complete collection of polls of any aggregator that I know of. Any other polling average (see: HuffPost, which excluded LA Times and other polls) would need to exclude polls that the aggregator didn't like. If Wang used an incomplete polling average, that's yet another huge mistake on his part. 1. You started out by claiming to explain why Wang's model was wrong. You just copied Wang's own explanation (meta-margin), showing again your poor critical thinking ability and over-reliance on "experts". 2. I showed your/Wang's meta-margin argument was wrong if you used a reasonable polling average. 2.5% or 2.2% state polling error both would have been sufficient to flip PA to Trump, assuming the RCP average or any other complete average. 3. Now you are no longer claiming you know the reason for Wang's mistake, saying you are just reporting what he reported. Advice: Try to read your "experts" (Nate Silver, Paul Krugman, Sam Wang, Harry Enten,...) with the same degree of skepticism that you read the loony right wingers. Reality will come into focus much more quickly.