Going and checking the
Real Clear Politics website for election results and I noticed something odd.
For every race where a D won, the polls were pretty close to the actual percentage of votes the D got. However, in races where an R won, the polls were off by anything from 4%-10% or even more.
Virgina Senate, for example, showed Gillespie at 38% in the polls, but he got 48.2%, pending count of (military) absentee ballots.
Illinois Governor: Rauner at 44.8% (and losing) in the polls, but garnering 50.8% of the vote and winning.
NC Senate: Tillis polled 43.4% (and losing), garners 49% and wins.
I've heard that polls generally "undercount" R votes by 3-4% but this election it was off by much more then that, in almost every race.