J. McIntyre at RealClearPolitics has a nice essay on the way he thinks the Electoral College sits today. Kerry has made up some ground, especially in Ohio, Pennsylvania, and Iowa, he thinks. But, Kerry still has an uphill battle; let McIntyre explain it. Note that Kerry hasn’t moved up in Florida and that Bush seems to be ahead in Wisconsin. Is it possible that Kerry could win Ohio and still lose the election? Also note: "Pennsylvania and Florida are must wins for Kerry and Bush respectively, a Bush loss in Florida or a Kerry loss in Pennsylvania means the election is over." A good read. 
Kerry has no chance of taking Ohio.
Kerry *does* have a chance of taking Ohio, mostly because he just keeps spouting the same job-loss statistics and telling people that he will bring them hope. (He doesn’t clearly explain how he will bring them hope, but for many voters, that appears to be irrelevant.)
He won’t take Ashland County, of course, but he is in a statistical tie in the state as a whole, largely because of his strong support in the counties hit hardest by job losses.
Stark County, Ohio (where Canton and Massillon are) is rumored to be the political bellwether of the country; if that is indeed true, I think there may be more recounts this year than there were in 2000, and the electoral college will be a mess.
I’m sorry. I just don’t agree that being down 5 points is a "statistical tie." It seems pretty convenient to me that any state favoring Kerry by more than two points is written off as Democratic, but every state where Bush’s lead is less than double-digits is a "battleground" or a "tie." You just can’t have it both ways.