Dan Balz writes in today's Washington Post:
Every political junkie around the country is spending hours with Slate's delegate calculator on the Internet or with more complex spreadsheets that are being passed around by e-mail. Plug in what you think the results will be in Ohio or Texas or Pennsylvania or Puerto Rico and see how the delegates divide between Clinton and Obama.
A note of caution for political junkies: Slate's delegate calculator doesn't give you the straight dope. It assumes that all of a state's delegates are awarded proportionally according to the statewide popular vote. This is wrong. About two-thirds of pledged delegates are awarded proportionally according to each congressional district's vote, while the rest are awarded proportionally according to the statewide vote.
For example, Ohio will award 92 delegates by congressional district (click here to see a map), while its remaining 49 pledged delegates will be awarded according to the statewide vote. Now, this doesn't make things easier for Clinton. In theory, she could win the seven congressional districts with four delegates apiece by 20 points, and each of those district's delegates would split evenly between Obama and Clinton. Obama could win five delegates by beating Clinton by just one vote in each of the five districts with five delegates.
So it might be even harder for Clinton to gain on Obama's pledged delegate lead than Slate's delegate calculator would lead you to believe, but you shouldn't focus solely on pledged delegates.
I still think if Clinton wins at least one state tonight and wins Pennsylvania on April 22, this race will continue at least until June and probably head to the convention. According to Real Clear Politics, Obama leads Clinton by 113 total delegates. With 611 pledged delegates unallocated and 359 superdelegates uncommitted, she can catch Obama or come close enough to catching him. Remember, "pledged delegates" are not actually required to vote for the candidate to whom they are pledged.
