I'm not any sort of expert on the Senate, but I've pulled together a bunch of metrics and researched many of the candidates as I consider adding Senate polling to our tracking. Some early conclusions? Incumbency is only an advantage of the incumbent is reasonably popular, and ideology matters a lot in Senate races. Although there are exceptions here and there (such as Rick Noriega in Texas), you can usually tell whether the opposition party is serious or not based on whether or not they nominate a candidate who is a good fit for that state's political environment.
Without further ado, here are the first and possibly last FiveThirtyEight.com Senate Race Rankings.
1. New Mexico (Open, was R-Domenici). I list this race above Virginia for a couple of reasons. Firstly, Tom Udall's lead is actually larger in the polling than Mark Warner's. Secondly, the GOP picked a candidate in Steve Pearce who is too conservative for his state, particularly on the Iraq war. And thirdly, there is no chance (at least I don't think) that Udall will get tabbed for the Vice Presidential nomination, whereas Warner still could. Pearce may pick up a handful of points as his name recognition improves but it should be a double-digit victory for Udall.
2. Virginia (Open, was R-Warner). Simply put, people like Mark Warner and don't like Jim Gilmore. Warner also has about a 5:1 edge in fundraising so far. But we're still talking about flipping a long-held GOP seat in a red (albeit purpling) state. A 90 percent+ plus chance of a Democratic pickup so long as Warner remains on the ballot, but I think it will tighten a bit.
3. Colorado (Open, was R-Allard). I'm ranking this ahead of New Hampshire based on gut-feel. Mark Udall's politics are very much in line with Colorado's: quite liberal on issues like the environment, the war, and civil liberties, but relatively conservative on immigration, taxation, and entitlement programs. Also, if we work from sort of the transitive property, Bob Schaeffer lost the Republican primary in 2004 to Pete Coors, who lost his Senate bid to Ken Salazar. Udall ought to be roughly as appealing a candidate as Salazar, but Schaeffer is not as appealing as Coors, and Colorado is turning blue pretty rapidly.
4. New Hampshire (R-Sununu). Leans pickup, but Sununu can make a race of it, particularly as he has about a 2:1 advantage in fundraising so far. John McCain's also liable to spend a lot of time on the ground in New Hampshire as he has relatively few places he can play offense.
5. Alaska (R-Stevens). Ted Stevens has the worst favorability/approval scores of any incumbent running for re-election -- he's polling at about a -5. In a change election, incumbency is only an advantage if the incumbent is reasonably well-liked; it can easily turn into a detriment if there is a 'throw the bums out' mentality. Alaska has a very young population that should turn out in large numbers for Barack Obama; if most of them vote for Mark Begich too, the state leans his way. Stevens has a huge, 11:1 advantage in fundraising so far, but how much money does it really take to compete in Alaska?
6. Mississippi-B (R-Wicker). This really almost ought to be considered an open seat, since Roger Wicker, though reasonably well-liked, is pretty far from having achieved 100 percent name recognition. Ronnie Musgrove is running as a very conservative Democrat -- if elected, he would probably be the most conservative Democrat in the Senate -- and is actually closer to the median point of Mississippi's electorate than the predictably right-wing Wicker. Musgrove doesn't even have to advertise the fact that he's a Democrat since there are no party names on the ballot in a special election.
7. Minnesota (R-Coleman). Although Norm Coleman's approval ratings aren't terrible, at the end of the day he is a little too conservative, and Minnesota a little too liberal, for this race not to tighten some. But we've passed the point where the chances of a takeover are better than 50:50.
8. Oregon (R-Smith). Gordon Smith is in the odd position of being a moderate in a state in which the electorate tends either to be very liberal or very conservative. Still, his approval ratings are passable, and Jeff Merkley is enough of a true liberal/progressive that Smith shouldn't have to worry about turning out the Republican base.
9. North Carolina (R-Dole). Elizabeth Dole is still reasonably well liked, but her voting record is conservative enough that Kay Hagan ought to be able to seize the political center if she runs a reasonably effective campaign. Dole has raised more funds than any other Senate candidate, however, so whether Hagan can win may largely depend on whether the Democrats decide to go for the jugular here.
10. Kentucky (R-McConnell). Mitch McConnell approval ratings are hovering around the 50:50 mark, which makes him quite vulnerable in a state where Democrats still have a substantial advantage in voter registration. Bruce Lunsford, his opponent, might prove to be slightly too far left for a state like Kentucky, and isn't likely to get many coattails from Obama, but he's run a well organized campaign and stands to pick up a couple points as his name recognition increases to 100 percent.
11. Louisiana (D-Landrieu). Louisiana is routinely listed as a potential pick-up opportunity, perhaps out of politeness to the Republicans who have few other targets. But Mary Landrieu is actually quite popular -- she had a 70:25 favorable/unfavorable score in an April poll from Southern Media & Opinion Research -- and the Republican John N. Kennedy is extremely poorly organized, with almost no funds raised to date. Just try searching for his website, for instance, you'll come across this before anything from the campaign. Kennedy is also not especially conservative, having recently changed his party affiliation, whereas Landrieu is one of the more conservative Democrats in the Senate. Where there is little difference in the ideology of the two candidates, the incumbent is usually going to win.
12. New Jersey (D-Lautenberg). This is a "BREAKING!" result based on the new Rasmussen poll that was released just moments ago, which shows Frank Lautenberg with a trivial one-point lead over Dick Zimmer. There may be a little bit of a "Hillary Effect" here as Lautenberg just survived a relatively vigorous primary challenge whereas Zimmer has not really been defined. New Jersey is notorious for teasing Republicans early in the election cycle and this may just be the latest example of that, but at the very least Lautenberg may need to spend from his Uncle Scrooge-like pile of money.
13. Texas (R-Cornyn). Rick Noriega is a fascinating candidate and may be able to make it competitive based on his charisma if he is able to turn out the Latino vote in greater numbers. But at the end of the day, it is probably too much to be asking for an authentic progressive to win a statewide race in Texas.
14. Maine (R-Collins). Susan Collins has shown a robust lead in the polls that has surprised some observers. But ultimately this one shouldn't be too hard to figure: she's a popular and effective Senator whose centrism, particularly on values issues, is a good fit for independent-minded Mainers. Tom Allen might be able to make a little bit of a push based on the war in Iraq, as Collins has not really distanced herself from the President on that issue, but his name recognition is already near 100 percent so the race is relatively far advanced.
15. Idaho (Open, was R-Craig). Certainly starts out in the Republican column, but it's very early as neither Jim Risch nor Larry LaRocco are all that well known.
16. Nebraska (Open, was R-Hagel). Scott Kleeb is an appealing candidate who is trying to make up for a big fundraising deficit with help from the netroots, but Nebraska is a red state and the prudent Mike Johanns might need some kind of Macaca moment to open Kleeb's door.
17. Kansas (R-Roberts). This race may tighten by a couple of points as Jim Slattery's name identification improves, but Pat Roberts is reasonably popular. A reach, even in a Democratic year.
18. Georgia (R-Chambliss). Could have been interesting, but the Democrats are taking too long to settle on a candidate.
19. South Dakota (D-Johnson). Could have been interesting, but the Republicans took too long to settle on a candidate.
20. Oklahoma (R-Inhofe). Democrat Andrew Rice would be an interesting candidate in a different state.
21. Wyoming-A (R-Barrasso). Democrats didn't mount serious opposition in what was effectively an open seat.
Live Boy, Dead Girl
22. Massachusetts (D-Kerry). John Kerry's approval ratings are surprisingly tepid, but Jeff Beatty is too conservative for Massachusetts.
23. Alabama (R-Sessions). Democrat Vivian Figures is too liberal for the state.
24. Mississippi-A (R-Cochran). Not the race to pay attention to in Mississippi.
25. Michigan (D-Levin). No polling yet, but Jack Hoogendyk is too conservative to get more than 40 percent of the vote against a well-liked incumbent.
26. Tennessee (R-Alexander). Democrats will have some kind of candidate but not a compelling one.
27. South Carolina (R-Graham). Democratic opposition never organized.
28. Montana (D-Baucus). Baucus is popular and opposition is poorly organized.
29. Iowa (D-Harkin). Christopher Reed is not a serious opponent.
30. West Virginia (D-Rockefeller). Rockefeller is popular and Democrats retain overwhelming party ID advantage in West Virginia.
31. Illinois (D-Durbin). Not in Barack Obama's year.
32. Arkansas (D-Pryor). Running effectively unopposed.
33. Rhode Island (D-Reed). Running effectively unopposed.
34. Wyoming-B (R-Enzi). Running effectively unopposed.
35. Delaware (D-Biden). Running effectively unopposed.