Sen. Hillary Clinton had a good night last night in West Virginia, winning by nearly 150,000 votes over Sen. Barack Obama. Of course a week ago the same candidates had quite different results in North Carolina and Indiana, where Obama won those states collectively by some 210,000 votes.

What happened to change the contest so dramatically?

Nothing, really. It's just like nothing had changed in Indiana's electorate when, long after the polls closed, Obama had a surge of support in votes counted in Lake County that for a time made it unclear who would carry the state.

What we've been seeing since February 5th are results that largely track what the Obama campaign projected would happen in a memo mistakenly released to Bloomberg News on the morning of the 6th. A lot has happened since then on the campaign trail, from controversies over Clinton facing sniper fire in Bosnia to statements by Obama's former pastor Jeremiah Wright, but the Obama campaign's memo projects most state winners and their victory margins with uncanny accuracy.

In a nationally competitive race, as this one certainly is, some states will go heavily toward one candidate, others toward another. Within states, one candidate typically piles up support in some counties, while losing heavily in others. Political analysts jump to certain conclusions about momentum or leap to hype themes like Obama's seeming trouble with the white working class-overlooking that in other areas of the country Obama has done very well in that demographic and currently is favored in upcoming heavily white states like Montana, South Dakota and Oregon.

What we are seeing is essentially a national primary unfolding state by state, county by county. Having different winners in different places doesn't change the basic dynamic of who is ahead overall.

What is unfortunate for Hillary Clinton is that the primary schedule – one erratically created with little national coordination– poorly serves such a national campaign. She has almost certainly lost among pledged delegates and the popular vote, and indeed that was clear weeks ago. She's like a team like playing the 7th game of the World Series after losing four of the first six games – she can win that 7th game, but can't win the series.

What would have made more sense is to have the campaign build to a crescendo where the final states really did determine the outcome even as the early states established momentum and began to clear the field. Developing such a schedule will take coordinated national party action – something we strongly support. I urge you to review different proposals at our FixThePrimaries site.

At the end of the day, voters matter and their decisions matter. Their predictability cannot be taken for granted. Yet their general predictability should be encouraging to defenders of representative democracy. Voters are grounded in certain beliefs and attitudes and over-hyped stories about campaign gaffes and big money spending isn't necessarily going to shake those views – indeed it often doesn't at all.

Voter predictability does mean that we should establish electoral rules that recognize the geographic realities now defining the impact of this predictability. The primary schedule should be made more rational, obviously. More fundamentally, if we want voter choice in legislative elections, we must move beyond winner-take-all rules to some kind of proportional representation – at a minimum, an Irish-style system of ranking candidates in districts of three to fives seats, where like-minded voters will be well-positioned to earn representation of their views.

And a final jab at some (not all) of our talking media heads: please stay calm. Jumping up and down over the profound horserace meaning of certain results may mean better ratings, but it can make you look pretty darn foolish over the glare of an objective spotlight.