At the end of the day, Silver and Wong’s predictions were *extremely* accurate:
We may not have liked the results of their techniques this time around.
But after producing these sorts of results three elections in a row with impressive and increasing accuracy and precision, we ignore them at out peril in terms (for example) of judging the ongoing effectiveness of campaign strategies.
Most impressive was not just calling all 50 states, but the way that Florida turned the faintest whisp of blue on his model two days before the election. Final projection was 49%-49% with Obama ahead by a few tenths of a percent or so.
Florida right now, Obama 49.8%,Romney 49.1%.
Pretty dang good.
Also, his model projected Obama to win the popular vote by 2.5%, which is also EXACTLY dead on as of now.
Fact is, people who said he was a hack and who quoted Unskewed Polls (a true moron) were just looking what they were hoping was true.