Although the Indiana race is going to be close the exit polling, which is almost *always* correct, indicates that Hilary Clinton will win Indiana by a narrow margin:
Male (45% of vote ) 51% for Clinton 49% for Obama
Female (55% of vote) 53% for Clinton 47% for Obama
The Male and Female stats cover all the bases, and therefore the likely outcome, assuming the exit polling was done fairly well, is that Clinton will get (.45 x .51 .2295) + (.55 x .53) = 52.1% of the total Indiana vote and Obama will get 47.9%. There is enough error in polling that this outcome is not certain, but it is very probable.
Contrary to what many think the exit polling in Florida was correct to suggest that Al Gore won the state. The key factor in Gore's "loss" in Florida was that the butterfly ballots of Palm Beach County which led to thousands of spoiled votes due to double voting. The exit polls can't measure this spoilage - they have to assume that people's votes actually *counted* where in the case of Florida the cast ballots were destroyed.
An important tangent to that is that the chad situation - even if Gore's Supreme Court challenge had moved ahead with a recount - would still have decided the election for Bush. Only certain unusual treatments of hanging chad counts, or the calculation where the spoiled palm beach ballots were mathematically allocated to Gore would have given him the edge. In a sense most of the analyses there were wrong. Republicans did not steal the election, but Gore did in fact win it. Blame our foolish Democracy and ballot procedures which fail to do a good job in some close elections.