The magic number in the Senate

As most of the polling news continues to get worse for the Democrats, some Republicans are beginning to believe the party can regain control of the Senate. It’s a tall order, requiring a pick-up of ten seats. While a plausible scenario exits for such a victory, it requires that nearly everything break the Republican’s way.
For me, though, there’s an alterative magic number in the Senate – four. That’s the number of net conservatives we probably need to add to deny liberals a filibuster-proof majority.
The way I figure it, the Senate currently has 38 reasonably conservative members – all 41 of the Republicans minus the two Maine Senators and Scott Brown. Thus, we need a net gain of three conservatives to deny the Dems 60 votes. But let’s add one more, to create a margin of error in cases when someone like Lindsey Graham wanders off the reservation.
What are the prospects for reaching this magic number?
Republicans, I believe, have four virtually certain pick-ups in November – North Dakota, Arkansas, Indiana, and Delaware. Three of those pick-ups would bring conservatives into Senate seats held by non-conservatives. The exception is Delaware where Republican Mike Castle is a moderate and probably left of center.
This gets us to 41, but doesn’t give our one seat cushion. However, there are three more races in which the prospects of adding a conservative are pretty good. They are Pennsylvania, Nevada, and Colorado. Illinois is another very possible Republican pick-up, but I’m not convinced that the Republican candidate, Mark Kirk, qualifies as a conservative (his ACU rating tends to be in the 50s or 60s, according to Real Clear Politics).
So the chances that four to six conservatives will replace non-conservatives are good. But we must also consider the danger that some non-conservatives will replace conservatives. I think this might well happen in Florida (with Charlie Crist), Ohio, and Kentucky. Collectively, these races seem roughly as losable as Pennsylvania, Nevada, and Colorado are winnable. Thus, “four” might be both the magic number and the “over-under” when it comes to conservative pick-ups.
But very often Senate races break strongly in one direction. Thus, it may be a bit unrealistic to think of conservatives gaining in states like Pennsylvania, Nevada, and Colorado only to give back the gains in Florida, Ohio, and Kentucky. If this is the strong Republican year many think it will be, the conservative candidate might well prevail in all (or maybe five of six) of these races and a few more, e.g., California (I assume here that Carly Fiorina really is a conservative) and Wisconsin.
In any event, I think we should analyze the upcoming battle for the Senate not just on the Democrat-Republican continuum, but also in terms of conservative vs. non-conservative.