What is wild about today is how wide the potential range could be in the Senate. I touched on this on Saturday with the observation that it was not far-fetched that Republicans could keep their losses to only three in the Senate. This was before the Pew, ABC/WP and Gallup polls that showed a late shift toward the GOP. Just last night the Evans-Novak Political Report predicted that Democrats would only net two seats in the Senate. It is this late shift to the GOP that pushed us to tick down our Senate number that had been at five-six down to four. And when you go down the list of RCP's 5 toss up races the possibilities of only a two seat loss are certainly not impossible.

However, and this should give pause to Republicans who are feeling better on the back of the closing generic polls, even with this late move, the odds still clearly favor a loss in the House. And while we think the most likely outcome in the Senate has moved from five to four, there is still a very real possibility that Democrats could sweep all five of the Toss Up states of Maryland, Virginia, Rhode Island, Missouri and Montana. With four out of five of these races having Republican incumbents trailing, (albeit by small numbers) a very good argument can be made they will all lose for as simple of a reason as they are trailing incumbents polling below 50%. And if that were to happen and the Democrats just hold Maryland they would get the Senate.

Given each one of these races has its own peculiar dynamics at work and they are all well within the margin of error, coupled with the market odds giving Republicans any where from 25% - 40% chance of wining each race, we felt that a Democratic sweep of all 5 toss ups was less likely than Republicans winning one or two. A tabulation of the InTrade odds at the end of last night as well as first thing this morning pointed to the GOP winning two out of these five as the most likely outcome and given the closing generic numbers that is where we came down as well.

As far as listing the order of how these seats might rate for each side, hat is difficult as there are cross-currents working both ways in all five of these races. We settled on Montana and Missouri as the two most likely to stay Republican, but you could certainly argue it other ways. Robert Novak who is predicting only a 2 seat Dem pickup suggests Burns will lose and Chafee and Steele will win.

Ignoring the weird situation in Rhode Island, Talent and Allen are closest in the RCP Averages and also are bid the highest in the trading markets. However, they are well known incumbents who are stuck near the mid-forties and don't have any momentum. Burns, Chafee and Steele on the other hand all have varying degrees of momentum.

Of the momentum guys we think Burns is the most likely to win simply because Montana is a conservative sate and Maryland and Rhode Island are not. And then between Talent and Allen we give the edge to Talent because he has run a very strong campaign, has some wind from the stem cell amendment vote, and has perhaps the best GOTV operation in the country.

So while we think four seats is the most likely pickup for the Democrats, six or two are legitimate possibilities.