Vorlon, very good analysis. I would like to make one point though.
There is a difference between a state like Arkansas and a state like North Dakota. Both states have 2 Democratic senators. Both states were won by Bush with ease.
However.
Arkansas used to be a pretty Democratic state. It was the MOST Democratic state in 1992, the only state in that electino to give any candidate an absolute majority of the votes. It has voted Democratic in most electinos throughout the twentieth century. But it has recently moved more and more towards the Republicans and now seems to be firmly in the GOP camp. It's reasonable to expect this to have an effect on the senators there. The same kind of analysis could be done on West Virginia.
But North Dakota is a different story. See, ND has NEVER been Democratic. It has voted Republican in almost every election since the FDR era. This means that the 2 senators there managed to get elected in an environment about as hostile to them as it is now. There is really nothing inherent saying they will go Republican.
Therefore, I think that the kind of analysis Vorlon does should only be applied to states with a significant recent trend. This excludes the following for the Democrats: South Dakota, North Dakota, Nebraska, Montana and Wisconsin.
For the Republicans: Rhode Island (though this is a rather special case) and Pennsylvania.
This leaves us with, as I see it, 5 vulnerable Democrats in Arkansas, West Virginia and Louisiana v 4 (5) vulnerable Republicans in Maine, Ohio, New Hampshire and (Oregon).
Basically, the trend should pretty much cancel out. Secindly, there seems to have been a historical fact that the senate has tended to reflect the national trend more or less, which suggests that these kind of analyses lead nowhere...
