Kari Chisholm has offered a nice rundown of how the process works, which I'll quote here verbatim. It describes exactly why trying to predict the post-Tuesday delegate count is impossible (a "fool's errand," as he calls it). After that, I'll predict the post-Tuesday delegate count.
Okay, clear enough? But that's too much to think about, so instead, I go back to my crude back-of-the-envelope math based on proportionality and old polls. (Mostly they're from February, but unless they reflect the last few moments of the voter's confused, indecisive mind, they're more or less useless.) Again we'll consult the chart, which by now my reader will have recognized has been riddled with errors and faulty assumptions--if pretty layout. I assume this one contains more of the same, but perhaps in a less ostentatious manner. For example, in the past I said the totals didn't include superdelegates, but in fact they did. (Superdelegates, of course, are even more impossible to predict, so I've been intending to just throw them out--thanks to this handy list, I've finally done so.)
In most states, roughly 35% of the delegates will be allocated based on the proportional vote each candidate gets in the statewide balloting. And roughly 65% of the delegates will be awarded based on the proportional vote each candidate gets in each congressional district.
But not all congressional districts get the same number of delegates. They've all got the same number of people - but not the same number of Democrats.... Every district gets between three and seven delegates....
For starters, you need at least 15% of the vote to get any delegates at all. Once the sub-15% folks are removed from the equation (sorry, Mike Gravel), here's how the math works:
* In a three-delegate district, if you get 50% plus one, you get two delegates. (Yes, that's a huge bonus for winning by a single vote.)
* In a four-delegate district, however, a 51-49% outcome leads to a 2-2 split among delegates. Even a 60-40 split is still 2-2. The winning candidate has to get to 62.5% to earn a 3-1 split.
* In a five-delegate district, unless the winning candidate gets over 70% of vote, it's going to be a 3-2 split.
* In a six-delegate district, it's going to be a 3-3 split, unless the winning candidate gets over 58.4%, then it's a 4-2 split. 75% of the vote earns a 5-1 split.
Based on these numbers (about which more will be said in a moment), the race will be a dead heat, with clinton holding a lead over Obama of roughly the delegates of Kansas:
884 - ObamaIf you look at the chart, you'll notice that the trend is almost uniformly toward Obama where data are available. This would suggest that, since the polls are lagging behind public opinion, these totals may be an underestimate. As you can see on Pollster's current trend line, that certainly looks to be the case:
913 - Clinton
However, there is some evidence that the trends may have flattened out just as Obama pulls even. Gallup, which has been tracking this each day, showed Obama pull nearly even only to drop back by five points:
So what's going to happen? My fear is that all this Obama hype will lead to inflated expectations and a Clinton rally like we saw before New Hampshire. Expectations are so high right now that if Obama falls 50-100 delegates short of Hillary, tomorrow's storyline will be that he's done. And, given the role of superdelegates and the punishing reality of proportional math, that may be the case.
If Obama falls back by 100 delegates, he'll be in a serious hole. After tomorrow, 1797 delegates of 3253 will have been allocated (excluding superdelegates). That leaves just 1456 remaining. Obama would have to win 779 of them--or 54%. And that's just to end up with the bare victory, leaving aside superdelegates. Today's results need to be close enough so that his momentum doesn't die. I'd put that at 50 delegates or fewer. I have yet to see a Democrat of my choice win the nomination for president, so my inclination is to fear the worst, nevermind Chuck's confidence. We'll know soon enough.