Another primary, another exit poll, and more pundits proclaiming the what ifs that have yet to become reality.  Mississippi voted yesterday and its sizeable black population gave Barack Obama a sizeable win.  But his modest turnout of white votes are once again giving pundits pause.

I ask why?  Time.com’s Michael Duffy writes "Only a Democrat who could win 35 to 40 percent of the white vote, while holding onto a lopsided percentage of blacks, could put the state in play in a head to head match with a Republican in the fall" saying Obama’s 30 percent of the white vote falls short.

But politics is not a zero-sum game.  White votes for Clinton do not instantly transfer to John McCain in the general election, just like Clinton’s wins in swing states Ohio do not mean only she is allowed to win it come November.  Exit polls in Ohio showed 44 percent would be satisfied with either candidate.  50 percent of them voted for Clinton.

Also, exit polls are showing just how hard it is to predict some states and voting groups.  Obama faired poorly with whites in Mississippi and Ohio but won the white vote in the larger Virginia, a once solid Republican state that will be a major swing this year.  And how many black people voted in Iowa and Wisconsin?

This election cycle has simply proved how little all the pundits (including myself) can know and predict.  But that hasn’t stopped the predicting and analysis which is captivated by the fight at hand without considering how things we didn’t predict will change everything else.  And remember votes don’t disappear after they’re cast.  We get to do this all over again in November.  I can’t wait.