The conventional wisdom forming among reporters is that Clinton can win by anywhere from 3 to 12 points in the popular vote but come out tied or even behind in the delegate count. In fact, a local congressman who supports Obama told the New York Times, "At the end of day, if we can carry more delegates and not have her win in the double digits, that would be great." The fact that such an outcome would be unjust doesn't stop it from being a serious blow to her candidacy.

Here's the state of current polling. Quinnipiac has Clinton up 7 points, 51% to 44%. Suffolk University has Clinton leading by 10 points, 52% to 42%. Mason-Dixon has Clinton ahead by five points, 48% to 43%. ARG has Clinton by 13 and Strategic Vision has her up seven. A random PPP poll has Obama winning by three.

The differences between the polls are products of the way they sample demographics. If tomorrow's turnout is heavy on young people and African-Americans, the polls that show a close race will be right. If turnout is heavy on working class whites, the blowout projections will be closer.

The average of all the polls? A difference of 6 points, 49% to 43%. That may well be the mark the media uses to determine victory, roughly speaking. Over 7, victory for Clinton. Under 5, victory for Obama. Anywhere from 5 to 7 and we call it a draw and do it all over again in Indiana and North Carolina.