I would remind everyone that the last Gravis poll in Iowa (conducted over a two day period from 1/11/2016-1/12/2016) showed Clinton at 57% to Sanders's 36% (O'Malley at 7%). This poll was off of the general trendline by 10.1% for Clinton and 7.6% for Sanders. It has a "C" rating at 538 and is in general awful. I would agree that Clinton is probably leading, but don't say that Gravis proves anything. It's a dumb outlier that shows almost nothing. In fact if it is off the margin it was last time (assuming the other polls are correct), Sanders would be at 49.6% beating Clinton's 42.9%.
(I took out the Gravis polls so they weren't affecting the trend.)
Never try to unskew Gravis, they're not even self-consistent. There was that crazy national poll with Webb at 12% and Chafee at 8%.
Sorry, my point wasn't to say that the polls were actually 49.6-42.9 Sanders, my point was to show how absurd their polls are to the people that were saying that this shows that New Hampshire will become competitive. If I was trying to unskew them, I would look at the breakdown of their polls for odd sample sizes, not just throw a flat rate from one candidate to another.
