The votes are still being counted as DR works on this post, but it certainly has been a very good day for John Kerry. He won 5 out of 7 contests, only losing South Carolina and Oklahoma, the two most "southern" states and two states the Democrats are highly unlikely to win. But he won two critical southwestern swing states (Arizona and New Mexico), as well as the great midwestern swing state of Missouri, which are more important states to the Democrats. And he won them easily, as he did Delaware and North Dakota.

On the other hand, the fact that Edwards won a state (South Carolina) and so did Clark (Oklahoma: with 100 percent of the precincts reporting, he is ahead by about 1,300 votes, though CNN has not as yet formally declared him the winner) suggests that one or both of these two candidates may well remain around for awhile. Of the two, Edwards had the better day. Looking at the top four contests (AZ,MO,OK,SC), Edwards' winning effort in SC an easy victory, while Clark's in OK was a squeaker. Then Edwards had two seconds, including his almost-winning effort in OK, plus a solid second in MO. Clark, on the other hand, had only his solid second in AZ to go with his squeaker win in OK.

Edwards has another thing going for him that DR finds kind of interesting: in contrast to Kerry, who did substantially less well among independents in each of these states, Edwards generally did substantially better. Here's Kerry's data in these states, with Democratic support listed first: AZ: 45/33; MO: 58/34; OK: 29/18; SC: 31/21. Now here's Edwards: AZ: 7/7; MO: 18/24; OK: 28/34; SC: 43/50.

So Edwards may have an easier time reaching out to independent voters as the primary contests move toward a possible one-on-one stage, as well as an easier time with these voters if he succeeds in getting the nomination (or, more likely, becomes Kerry's running mate).

As for getting the nomination, it's going to be very tough indeed to beat Mr. Electability. In each one of these states, including the two he lost, Kerry beat his closest rival easily among those voters who selected "can beat Bush" as their top candidate quality: AZ: 65/18; MO: 75/12; OK: 55/21; SC: 57/29.

Oklahoma and South Carolina, where Kerry's margins among electability voters were less, were also the two states where there were the fewest electability voters. In neither state did the proportion of electability voters break 20 percent and in neither state was it selected by the most voters as the top candidate quality they were looking for.

So that's the secret, DR supposes, of beating Mr. Electability: the less the contest is about that, the better off the other candidates (which at this point really means Edwards and Clark and possibly just Edwards) will be.

But the primaries, as we move forward, seem likely to be more, not less, about electability. Which means their only chance may be to steal that mantle from him. Good luck. It ain't going to be easy.