The NBC News/ Marist poll was the first to come out on Sunday. The interesting thing here is the shift towards Cotton in this poll. Even among RV's, Pryor does not lead for the first time this year (among traditional pollsters. ie. Live interview). This leads me to believe that the race is tighter than this poll would seem to show. For the most part, the Demographics look in step with 2010 turnout and not too terribly far off from 2012. So, this might just be an odd sample.
-Unfortunately, we have no exit poll data from 2010 and 2012 for the state.
-As an overall observation, they seem to be polling Female and Democratic Voter turnout to be bellow 2010 levels. If the GOP is banking on having better than 2010 turnout levels (ie. even more women, minorities, and Democrats stay home): they are probably a little worse off than this set of data shows.
-One of the clearest problems is in Minnesota where they over weigh Republicans over Democrats by a few points.
-The biggest problem here is their weighting. Unfortunately, since there is little other data in these states: we don't have a whole lot of comparison to do. Even bigger of a problem is that many of these interviews were conducted in mid-late August. These are now a 2-3 weeks old and once the Gubernatorial portions come out, they will be 3-4 weeks old. So these polls should be looked at in the context that they are not the most recent polls we have for many of these races.
My projection now lands in the problematic zone where Kansas is where the balance of power turns. If Orman were to win (which I am projecting him to), the balance of power would depend on him. If he caucused with the GOP, they would have a 51-49 majority. If he caucused with the Democrats, they would have effectively a 51-50 majority (with the Senate President Joe Biden). With the large number of competitive races on the books right now: The GOP has the edge in the playing field as we get going through September.