After posting about the risks of partisan bias unduly influencing election ratings and forecasts, I thought it would be fair for me to put my money where my mouth is. I claim no expertise in prognostication, but I will try as hard as possible to avoid any wishful thinking in my ratings. You'll note that I rate some races Safe that others would rate as Likely, because I don't believe in fooling around when I really don't see a path to victory. Have a look at the ratings after the fold, and please discuss and argue with them as you see fit. All states are listed in alphabetical order, with a summary below, grouping states by rating and discussing different scenarios for greater or lesser Republican and Democratic gains.
Ratings key:
Safe = No path to victory for the other candidate, absent some unforeseen disaster to the leading candidate.
Likely = Perhaps a 10-20% chance of victory for the trailing candidate.
Lean = Slight edge to one candidate, with the other having about a 25-35% chance of winning.
Tossup/Tilt = A very tenuous edge to one candidate
Arizona: Tossup/Tilt R - Carmona has been coming on, but I'm still unconvinced that he will beat Flake and would like to see more polling. It would also help if Obama could get closer and possibly win the state, which I do think is conceivable, though I still doubt he has more than a 35-40% chance of carrying the state, even if OFA saturates the state with advertizing and mailers about Romney and Ryan in relation to Medicare.
California: Safe D.
Connecticut: Likely D. It's hard to completely ignore Linda McMahon's money, but if she couldn't win in 2010, it's difficult to envision her winning this time. My prediction is Murphy by 15.
Delaware: Safe D.
Florida: Lean D - Some polls are too close for comfort, but I still doubt this is a real tossup, and I also think that Obama has a better than 50/50 shot of taking the state, especially now that Ryan has been selected as Romney's running mate.
Hawaii: Safe D. Others rate this as Likely, but I see no path to victory for Lingle and expect Hirono to win by at least 20 points.
Indiana: Tossup/Tilt D. Polls have shown this race just about dead even. My feeling is that Mourdock is an extremist and Donnelly is a moderate, and Hoosiers tend to prefer moderates of both parties over extremists in statewide races. So I think that even if Obama loses the state, as currently seems quite likely, if the polls are tied going into election day, Donnelly has a slightly better than even chance of winning.
Maine: Safe I (caucuses with the Democrats).
Maryland: Safe D.
Massachusetts: Tossup/Tilt D - In the end, I expect about a 5-point win for Elizabeth Warren, but for now, polling can't justify more than Tossup/Tilt D.
Michigan: Safe D. Hoekstra utterly destroyed his chances with a grossly racist ad campaign that Michiganders hated. He was an underdog to begin with, but now, I expect him to be completely flattened in the general election. Stabenow by at least 15.
Minnesota: Safe D.
Mississippi: Safe R.
Missouri: Lean R - McCaskill's strategy is similar to Donnelly's: To run as a moderate against an extremist. However, she has a more liberal record than Donnelly, and Missouri has shown itself less reluctant than Indiana to vote for extreme or stridently partisan Republican candidates in statewide races. I also think McCaskill is more unpopular than Donnelly. So while she definitely has a chance with the strategy she's pursuing, I have to give the extremist, Akin, an edge for now.
Update: After Akin's bizarre comments about rape, I cannot justify a Lean R rating for this race. I am changing my rating to Tossup/Tilt D, because I think McCaskill will eventually win this race, even though it's probably more of a Tossup/Tilt R race right now.
Montana: Tossup/Tilt D - This might turn out to be the closest Senatorial race in the country. Montana tilts Republican but does have a tradition of ticket-splitting in statewide races. Tester has run terrific ads and is very authentic and credible. Denny Rehberg showed some independence from his Party Leadership recently, but his record is damaged by his involvement in a boating while intoxicated incident. By no means is this close to a safe race for either side, but I give the slightest of edges to the incumbent.
Nebraska: Safe R. Kerrey wins this only if some unforeseen scandal breaks out, and maybe not even then, considering his unpopularity, the unpopularity of the Democratic Party in Nebraska as a whole, and his own sordid history in Vietnam.
Nevada: Tossup/Tilt D. My rating is a bit daring because polling is showing rather the reverse (Tossup/Tilt R), but Nevada polling seems to suck and specifically seems to greatly underestimate Latino turnout. I think Latinos will turn out in droves and Shelley Berkley will win by at least 3 points.
New Jersey: Likely D. Except in Gubernatorial elections, this state just about never comes through for Republicans anymore. I vividly remember when things were different, but some of you surely don't! In any case, Menendez is not very popular, so I don't think he's completely safe, but anything less than an 8-point win for him would surprise me.
New Mexico: Lean D. I don't expect anything closer than a 5-point win for Heinrich, and my prediction is 7 points. At the same time, Obama will probably win this state by more than Nevada.
New York: Safe D.
North Dakota: Tossup/Tilt D. This race is hard to rate because of the paucity of publicly-released polling. It's also hard to know how to interpret the release of Democratic internal polling and the lack of an answer from the Berg campaign. And of course the state's lean is Republican, but that's more true of Presidential voting and voting for some state offices, not so much in elections for either House of Congress. North Dakotans frequently vote for Democrats for both the House and Senate. So this rating is based, with some caution, on a combination of the polling that has been released, the heavy expenditures on ads from both sides, polling showing that Berg is fairly unpopular, and a feeling that Heitkamp's ads have been mostly excellent. I think this is a slightly better shot for the Democrats than Tester's defense in Montana, because I believe Berg is less popular than Rehberg. However, if some polling shows otherwise, please correct me.
Ohio: Likely D. This is Ohio, so it's hard to consider someone as liberal as Brown a total shoe-in, but as things stand, I think he'll win by at least 8 points, and probably more than 10. The Republicans have thrown everything at him, including the kitchen sink, and he has continued standing tall, so it's hard to imagine what they could do that would fell him. Even in the relatively unlikely event that Obama loses the state and even if Romney is elected President, I still don't see a likelihood of Brown losing.
Pennsylvania: Safe D. Casey is not going to be defeated, period. He's a moderately liberal pro-life senator in a Democratic-leaning state where Obama will probably win by about 10 points again. Casey will run ahead of the President. My prediction is Casey by 15.
Rhode Island: Safe D.
Tennessee: Safe R. This one is a bit sad, considering how close it was 6 years ago, but if I were living in Tennessee, I wouldn't vote for the Democratic nominee. I guess we have to be glad that Corker is relatively reasonable for a Southern Republican senator.
Texas: Safe R. Texas may become a tossup state in a couple of decades or sooner, but it sure isn't one now.
Utah: Safe R.
Vermont: Safe I (caucusing with the Democrats).
Virginia: Tossup/Tilt D. This election has been maddeningly close, but in the end, I think that the state will go for Obama by at least 5 points and the heavy turnout from Northern Virginia and areas with heavy black population will put former Governor Kaine over the hump. And I don't think there will be many Obama/Allen voters, and somewhat more Romney/Kaine voters, because Kaine is not as liberal as Obama, whereas Allen lacks the moderate record Romney had as Governor of Massachusetts (which Romney is either completely running away from or touting, depending on the day or context).
Washington: Likely D. I think this is close to Safe, and I'm kind of undecided on where to rate it. Suffice it to say, I would be very surprised if Cantwell doesn't win, and win by at least 10 points.
West Virginia: Safe D. If Raese couldn't win in 2010, he won't win this time, either. Manchin by 15.
Wisconsin: Lean R - I think this is between tilt and lean, and as of now, I have to give Thompson an edge, but Baldwin definitely has an opportunity. Some of the factors influencing my rating are Thompson's record of broad-based statewide victories in the past and a perception that he is a moderate Republican, whereas Baldwin is a very liberal Democrat - and particularly, a liberal Democrat from Madison, site of one of the nation's most highly-respected universities and target of many Wisconsinites' ire. It also can't be overlooked that a few voters may vote against Baldwin just because she is a lesbian. All that said, Baldwin comes across in ads as a nice, soft-spoken person, which may reassure many voters that whatever her ideology, she is moderate in personality and of good temperament to do the job. As I see it, her difficult challenge in this campaign is to convince voters that former Governor Thompson is either no longer a responsible moderate who cares about good government (whether that's a true image or not, it's what many, many voters have thought) or/and no longer has the proper temperament or competence to do an important job for the people of Wisconsin, and convince the voters of this while retaining an image of being pleasant, soft-spoken - Midwestern nice. She's getting some help from independent expenditures on ads, but her own most recent ad slammed Thompson for his lobbying work and said that "he's not for you anymore." I'm not sure how far beyond that she should go in her own campaign's ads. She can't let Thompson walk over her, but if she is seen as too unladylike as well as very liberal, she really risks having her attack ads backfire. This will be one of the most interesting Senate races to watch this year.
Wyoming: Safe R.
To summarize:
Safe D: CA, DE, HI, ME (I), MD, MN, MI, NY, PA, RI, VT (I), WV
Likely D: CT, NJ, OH, WA
Lean D: FL, NM
Tossup/Tilt D: IN, MA, MT, ND, NV, VA
Tossup/Tilt R: AZ
Lean R: MO, WI
Safe R: MS, NE, TN, TX, UT, WY
If the Republicans end up winning all the races I rate Tossup/Tilt D (the most unlikely of which, I believe, is MA), they would turn over 6 Senate seats while losing 1 (Maine), with a possibility of 8 turnovers and 1 loss if they also win the Lean D races (though I think beating Heinrich is more of a stretch than beating Nelson). On the other hand, if the Democratic candidates in the Tossup races sweep, the Democratic Party would turn over 5 Republican seats while losing 3 (update: 2 - I now predict that McCaskill will narrowly win), with a possibility of losing only 1 seat if they also sweep the Lean R races. So it's anybody's ballgame, but my ratings at this point call - albeit tenuously - for a Democratic gain of 4 seats and loss of 3 (update: 2), amounting to a net Democratic gain of 1 Senate seat (update: 2 Senate seats). And given that that's where these ratings end up, I feel somewhat reassured that I probably have not made any wildly biased predictions. But feel free to disagree. I'll see you in the comments.