Not a lot of changes on this final prediction. 22 of the 37 races are listed as "strong Democrat" or "strong Republican." These races are foregone conclusions -- I am certain of the winning party and the margin will almost certainly be double-digits. 6 of these races are safe for the Democrats and 16 for the Republicans. Several of the "safe Republican" races are races that were expected to be quite competitive just a few short months ago -- Florida, Louisiana, and Ohio. (I am going out on a limb and predicting that Marco Rubio will win the three-way Florida race with more than 50% of the vote.) The fact that the Democrats are no longer competitive in these races speaks volumes about what kind of election year it is shaping up to be for the GOP.

7 more seats are in the "Leans R/D" category -- 5 leaning toward the GOP and 2 leaning toward the Democrats. These are races where I am confident of the winner but acknowledge that the contest is competitive and am not willing to completely rule out a close final margin of victory -- or even an upset. Here are some quick snapshots:

Connecticut (Leans D) -- I expected this race to be very close up until Election Day, but Democrat Richard Blumenthal has confounded my expectations and pulled out to a 10-point lead despite being outspent more than 5-to-1. I think in the end, McMahon's controversial past and the state's Democratic lean are just too much for her to overcome. While I expect Blumenthal to post a clear win, I wouldn't be surprised if McMahon outperforms her poll numbers a bit.

Delaware (Leans D) -- I was pretty certain Christine O'Donnell would lose from the day that she won the GOP primary in mid-September, and I continue to believe that. However, O'Donnell has actually out-raised and out-spent Coons and is polling a bit better than one might have expected (two recent polls have her at or above 40% and 10-11 points behind). I wouldn't be surprised to see her lose by less than 10 on election night.

Kentucky (Leans R) -- I really thought Rand Paul would have put this race away awhile ago, but Jack Conway has kept hanging around. Only in the past week or so, with significant negative fallout from Conway's outrageous "Aqua-Buddha" attack ad against Paul, has Paul finally started to put some real distance between himself and his opponent. The video of a Rand Paul supporter stomping on the face of a Conway supporter outside the debate hall last week may have broken Paul's momentum slightly, but shouldn't have any significant impact on the outcome. Still, it's disappointing to me that a nasty, negative, left-wing trial lawyer like Conway could come this close to winning in a state where Obama's approval rating is in the low 30's.

Missouri (Leans R) -- This was expected to be an epic ideological clash between representatives from two well-known political family dynasties, but in the end it looks as though Republican Roy Blunt will end up winning comfortably. The Democrats have spent a lot of money here, but decided a couple weeks ago that the race was likely unwinnable for them and pulled out most of their remaining funds. There's little doubt Carnahan has been severely hurt by her association with Obama and his policies, which are toxic in this state.

New Hampshire (Leans R) -- I continue to be impressed (and a little surprised) with the size of Republican Kelly Ayotte's lead in the polls (the last three have had her up by 15). After all, New Hampshire has been trending Democratic of late, and Obama's approval rating is higher here than it is many other places. But even here, ObamaCare is not playing well with voters and is dragging down Paul Hodes. Ayotte clearly has more appeal to independents and looks likely to win comfortably.

North Carolina (Leans R) -- Republican incumbent Richard Burr looks likely to win, but his lead in the polls is a bit underwhelming (and has been all year). Even at this late point in the campaign, a couple of polls have shown him under 50%. In the end, Burr will almost certainly win comfortably, but he is lucky to be facing a relatively weak and severely underfunded challenger in an extremely favorable political climate. If it weren't for these advantages, Burr would likely be in serious trouble.

Wisconsin (Leans R) -- This race says a lot about how toxic the Democratic label is in 2010. Russ Feingold is a long-time Democratic incumbent who has carefully cultivated a squeaky-clean reputation. He hasn't made any serious campaign gaffes missteps and is running in a state that hasn't voted for a Republican presidential candidate in over 25 years. And yet, he is badly trailing in the polls to a Republican who has never before run for public office. The anti-Democratic wave is as strong in Wisconsin as it is anywhere in the country, and it looks likely that it will sweep Ron Johnson into the Senate.

Finally, I have 8 seats in the Tossup category. These are seats that could go either way on election night, and the final victory margin for each of these races will probably be no more than 5 points or so. I currently have 6 of them going Republican, with just 2 going to the Democrats. This is consistent with past wave elections, where most of the close races end up going to the same party.

Alaska (Tossup Republican) -- For purposes of this ranking, I am considering Lisa Murkowski an independent. I certainly believe Murkowski has behaved very selfishly and has allowed her ego to put a Senate seat that should be safe for the GOP at risk. Nevertheless, it seems clear to me that she has a real shot at winning as a write-in candidate. Her high name recognition in the state makes her chances much better than most write-in candidates. Also, everyone seems to agree that Joe Miller has run a poor, gaffe-prone campaign. I still think Miller's chances are better than anyone else's in this heavily Republican state, but I can't rule out a Murkowski win. Even the Democrat has a chance in this unpredictable race. If it's as close as polls seem to indicate, it's likely we won't know the outcome for several days after the election.

California (Tossup D) -- The polls right now are basically in the same place they have been for the past three or four months. Boxer is slightly ahead and slightly under 50%. This is not a great place for an incumbent to be a few days before an election. But California is such a liberal state that I still think her chances are a bit better than 50-50 to win. Carly Fiorina has a chance too, but she has no margin of error. She has to hope for a big turnout gap between Republicans and Democrats in the state, coupled with a big win among independents and late-deciders. Even then, she would likely only win by a couple of points. Carly is a great candidate, and has put the GOP in position to win here, which is all anyone could ask. She is also outperforming her more moderate GOP counterpart in the governor's race, Meg Whitman, which belies the naysayers' claims that a more moderate candidate would have been more likely to win.

Colorado (Tossup R) -- Polls over the past couple of weeks have shown a virtual dead heat between Ken Buck and Michael Bennet. The Democrats have been spending big in this race and making every effort to salvage this seat. Most observers feel Buck has made a lot of mistakes and squandered a lead with some ill-timed comments on social issues. Nevertheless there are good reasons why I think Buck will ultimately prevail. While Buck's numbers have moved down a little in recent weeks, Bennet's have been mired in the mid-40's for months -- not a good situation for an incumbent, even an appointed one. Obama and his agenda are quite unpopular here, and Bennet is tied at the hip to that agenda, like it or not. Also, Republicans have a registration advantage and have a turnout advantage in early voting. I think these factors add up to a narrow Buck win.

Illinois (Tossup R) -- Most recent polls have shown Republican Mark Kirk with a narrow lead of 1-4 points. There're still a fair number of undecided voters, but many of them are independent voters who I think will eventually break for Kirk. Kirk also has a financial advantage and a reputation as a centrist -- two factors that are working to his advantage. And Obama is more unpopular in his home state than you might expect, so he may be unable to help Gianoullias much. There's no doubt this is a tossup race that could go either way, but I have a strong hunch that Kirk will pull this one out.

Nevada (Tossup R) -- After months of pessimism about Angle's chances, I now see enough positive signs to predict a GOP victory. Over the past month, Angle has clearly built momentum and the polling trend is definitely in her favor. She significantly outperformed Reid in their debate, and she has actually raised more money in the final stretch as well. Early voting trends are inconclusive, but it's clear Republicans are turning out at slightly higher rates than Democrats. That coupled with a strong advantage among independents should be enough to put her over the top. Still the race is close, and Reid still has a chance, given the big registration advantage the Democrats have. There is certainly no concession speech I would like to watch more than Harry Reid's!

Pennsylvania (Tossup R) -- Polling shows a tightening race, but given the Democrats' significant registration advantage this was probably inevitable. Most polls still show a small Toomey lead, which is pretty impressive given the millions and millions of dollars that the national Democratic party has poured into this race. While Toomey is definitely more conservative than most Pennsylvania Republicans who have won statewide, Joe Sestak is far from a centrist himself. In the end, I think Obama's 60% disapproval rating in the state will prove decisive, and Toomey will win by 4-5 points due to some Democratic crossover support and a large advantage among independents.

Washington (Tossup R) -- This was a last minute decision based on a couple of recent polls indicating Republican momentum. I am far from confident about this prediction; however I am very confident that this is a close race that will come down to 1 or 2 percentage points either way. Patty Murray is a weak incumbent who is under 50% in most polls. Also, we know Republican enthusiasm in the state is running high, because more Republican than Democratic primary votes were cast earlier this year. In the end, I think the wave will put Rossi over the top. He has run a strong campaign and has been able to compete financially with Murray over the past few weeks. If the race is close, however, Republicans have to be concerned about fraud in King County, which robbed Rossi of a victory in the governor's race back in 2002. It's possible Rossi could have the lead on election night but "lose" as a result of recounts and absentee ballots.

West Virginia (Tossup D) -- This is another very close race which I have gone back and forth on. I have decided to move it back to the Democratic column, since recent polls have indicated momentum for Joe Manchin. If Manchin wins, it will be because he has gone to extraordinary lengths to separate himself from Obama and the national Democratic agenda. Honestly, I expected Manchin to win easily all along, and have been very surprised by how competitive John Raese has been. Win or lose, it is impressive that the GOP has gone toe to toe against a governor with a 70% job approval rating. This race is very close, and if the GOP is able to win here on election night, it will be a possible early indicator that they will take the Senate.

My overall prediction remains the same: the new Senate will be tied, with 50 Democrats and 50 Republicans. Of course, Joe Biden will be the deciding vote which will enable the Democrats to keep their majority. It's possible the Republicans could gain 8 or 10 seats instead of 9, but I really doubt they will gain less than 8 seats. Anything less than 8 seats would be a big disappointment.