Let’s say you are in the position to help one of the campaigns out–by donating money, or running an independent-expenditure ad campaign, for example. Under those circumstances it may well make sense to concentrate on Senate races rather than the presidential race. The question isn’t whether the presidential race is “over” or whether an improved Election Day performance by McCain might help Senate candidates; the question is where the marginal dollar would be likely to have the most effect. But of course if you’re making that kind of calculation not all Senate races are equal.

Devoting resources to the Virginia Senate race would be crazy: Republicans are not going to keep that seat (and as Mark Steyn observed, McCain’s veep pick has nothing to do with that). Spending money in Maine would be crazy too: Republicans will probably keep the seat, and keeping it may not help all that much next year. Spending money in Alaska makes no sense either: The outcome there depends on a jury. It seems to me that the places where marginal dollars would have the most impact are Minnesota and North Carolina, where Norm Coleman and Elizabeth Dole, respectively, are fighting strong challengers.