First off, the majority of seats are pretty much decided already. Democrats stand no chance (essentially) of winning Idaho, Wyoming, Mississippi, Texas, Alabama (obviously), Maine, Oklahoma, South Carolina, etc. The same goes for the GOP winning Massachusetts, Delaware, Oregon, Rhode Island, Virginia, Hawaii, and so forth.
Below is a list of the seats that I would consider the remaining competitive seats. These are races where the training party has a better than 15% chance of victory.
- Iowa- Likely to be the closest race in the country. After having a lead for most of the cycle, the Democratic Congressman has stumbled into a close second behind the state Senator Joni Ernst. While the race has tightened again after the September surge by Ernst, she still maintains a slight edge.
- Colorado- This race has seen the biggest shift over the past couple months. After looking like Udall was doing fairly well, his opponent has had a streak of good polling. Though there have been a couple polls since this shift began that suggests Udall still have a very decent shot at winning.
- Arkansas-If you asked me in the spring, I would have said that of all the red states, this is the one that the Democrats would have had the best chance in. Well, aside from North Carlina and Kansas, that has mostly remained true. Mark Pryor is one of the more moderate members of the Senate Democratic caucus. He also, from time to time, sounds more like a fundamentalist christian Republican than a liberal Democrat when he makes comments stating his belief that the earth is only about 6,000 years old, and his score from the LCV is not all that high. His score is lower than that of Susan Collins (R-ME) and about the same as Mark Kirk (R-IL). From the ACLU, he gets only a 54% rating. For comparison: Lisa Murkowski (R-AK) 81%. John McCain (R-AZ) 72%. Even more telling is that he has only voted with the Democrats 78.3% of the time this session. The only currently serving US Senator with a lower percentage of votes with their party is Joe Manchin (D-WV). (Of course this doesn't include Angus King or Bernie Sanders). Also, Arkansas isn't the reddest state the Democrats are defending this year. Those two things together, while still overall problems for Democrats, are less of an obstacle than other races. I would still say the Democrats probably are more likely to win this seat than Kentucky, Georgia, Louisiana, or Alaska at this point.
- Alaska-The "accidental Senator." After narrowly winning in 2008, the polling in 2014 has been trending away from Mark Begich. This session, he has voted with the Senior Senator from Alaska 64% of the time. Murkowski (R) has an above water approval rating with the Independents and Democrats of the state while having a below water approval rating with Republicans in the state. It has been repeatedly said that Begich has run the best campaign of any Democratic incumbent in a red state this year (I would disagree though) and that polling here is poor and scarce. If I had to place a bet, I would say Sullivan takes the seat. But then again, this one would be the most likely race to shock us on election night.
- Kentucky-After holding a brief advantage this spring, Grimes has fallen behind and pretty much not improved much until the last couple weeks. While still a steep underdog, this race could be won if turnout is high enough. If there was one state that Democratic turnout could exceed expectations: it would be Kentucky. McConnell is to Democratic and Independent voters in Kentucky what Obama is to Republican and Independent voters in the state. What will likely help McConnell win this one will be his money advantage and a Democrat that has been put on the back burner recently by the national party. While the Democrats shouldn't pull out of here entirely, they would be spending there money better in Iowa, Colorado, and Georgia. (The same thing I would say to the GOP about North Carolina)
- Georgia-This (aside from Kansas) is the best chance the Democrats have at picking off a seat from the GOP board. This race, as it looks now, will go to a run-off. The question will be: has the Senate already been decided? If the Senate is already decided (especially if the Democrats have 50+ seats already), the GOP would be less likely to mount an overwhelming GOTV campaign. The Best GOP turnout would come from the slogan: "Vote for Purdue to assure a Republican majority in the Senate." The question here might be: if the Senate is already decided, how badly do the Democrats want this seat?
- Louisiana-Mary Landrieu is no stranger to close races. She may still be able to win this year. This race is almost certain to go to a run-off. If the Senate has already been decided, the GOP might have a hard time getting Maness supporters to turn out for Cassidy. This might be a difficult point anyhow.
- Kansas-This might be the strangest race we have this year. After the withdrawal of the Democrat from the race, Orman has maintained an advantage over the unpopular incumbent. Senator Roberts (R), has residency problems and only became fully (still debatable how much so even now) engaged in this race in the last few weeks. The biggest wildcard here might be whether Milton Wolf will ever endorse Pat Roberts, and if Wolf's supporters from the Primary will turnout and for Roberts. The winner here wouldn't be relevant if the GOP already had the Senate majority. Orman has said he would caucus with the majority, so if the GOP had 51 seats without Kansas, they would end up with 52 with Kansas regardless of the winner of this race.
- New Hampshire- This race has been an uphill climb for former Massachusetts Senator Scott Brown. Since the primary he has picked up quite a bit of ground and even held an advantage in his first couple polls. This race is still quite unlikely to go the GOP's way at this point. I still have Brown at a 4% chance of winning here. If the GOP takes this one, they would likely have 54-55 seats.
The point I am making here is than most of these races that are close are already in the hands of the GOP in my projection. Therefore, there is really only one more seat of these that the GOP could gain. Since these all remain close, it is conceivable that the Democrats could still win a couple of them. Since almost all of the close races are currently in GOP hands, and there are so many of them, coupled with the Orman situation: This is why the Senate outlook is still a toss-up. I still have the range for this cycle at R +2 - R +9. Though in reality, I would narrow that down farther for R +4 - R+8. We are currently at R +8. A couple of these races that the GOP holds the advantage in have tightened in the last month. Therefore, it is likely that at least one of the 8 they currently have flips back and becomes no better than R +7. If the Democrats pull back 2 (likely IA and CO) this becomes R +6 if Orman caucuses with them and R +5 if Orman caucuses with the Democrats. Presuming Orman caucuses with the Democrats in a split Senate, the Democrats would maintain an effective 51-50 majority with Joe Biden.