Ranking the Senate Races....

[Talleyrand]

Upgrades:

Nebraska - Fischer is up by 25 according to WAA. I think we can call this.
Wisconsin - Thompson up 11, although it is Rasmussen. Probably a Lean R at least though.

Downgrades:

Missouri - you know why, although there is withdrawal potential.

With these changes, I think your ratings are much more accurate than Cook, Rothenberg, and Sabato. I completely agree with them.

[Talleyrand]

Sabato has said on Twitter he's moving MO from Lean R to tossup. Jennifer Duffy of Cook said she's moving MO from Tossup to Likely D. Needless to say, Sabato's rating is more accurate.

Likely D!? Cook's ratings are fairly ridiculous anyway. He has virtually every competitive race in the Tossup category.

You must be logged in to read this quote.

...That isn't what the polls say

[Talleyrand]

He's usually very good, just an overreaction. After all we're still talking about an incumbent who's at 50/40 disapprove and stalled in the low 40s. Which won't matter if Akin is on the ballot come Nov. 6- just ask Harry Reid.

In the past he has been pretty good, but his senate ratings this cycle haven't been too great. I think he started off with West Virginia as a tossup and for some reason Ohio, Hawaii, and Wisconsin are still tossups. I wonder if he'll update any of them soon, especially the last one. It's pretty clear Baldwin is a genuine underdog against Thompson.

[Talleyrand]

BTW, it appears Cook no longer has their ratings available for free. Which major rating systems, do?

Both Larry Sabato and Stuart Rothenberg do.

Sabato- http://www.centerforpolitics.org/crystalball/articles/category/2012-senate/

Rothenberg- http://rothenbergpoliticalreport.com/ratings/senate

Oops, looks like RogueBeaver already posted.

Also, Nate Silver is releasing his model for this year's senate cycle soon.

http://fivethirtyeight.blogs.nytimes.com/2012/08/15/in-senate-races-politics-are-local-again/

You should check back for its launch.

[Talleyrand]

I think Montana is even more likely to switch parties than Massachusetts. Tester has lost his personal popularity, and the only poll which has shown him ahead is PPP, and even they showed undecideds favor Romney 71 to 8. He's pretty much finished.

Donnelly is only close in the polls right now because Mourdock is underperforming Romney heavily. The Democrat hasn't been able to break out of the low 40s, even in his best polls. Therefore, I think the undecideds will move decisively to Mourdock in the end, closing the deal.

In North Dakota, Heitkamp, while the best candidate Democrats could have gotten for this seat besides Dorgan or Conrad, is not much more popular than Berg, according to the most recent Mason-Dixon poll. Despite outmaneuvering Berg, running a much more charismatic campaign, and managing to avoid scandal for the most part, she's only tied with him, and she has seen a decline in the polls since the spring. While I think she is more likely to win than Tester or Mourdock, it's probably Lean R.

Connecticut looked like a potential McMahon upset, and she could still pull it off, but the only recent poll showing her up was Quinnipac, and that was a virtual dead heat and a decline from their last poll. Plus, Rasmussen and PPP have her down six each, and she was routed in the most recent debate. Probably Lean D, but she's likelier to win than any of the Democrats I mentioned earlier.

[Talleyrand]

Idaho: No change. Safe D.

I was unaware of that situation, but everything else about your analysis is excellent.