In Friday's New York Times, an article about Republican efforts to lure black voters contained this head-scratcher: "Although Mr. Bush won only 11 percent of the black vote nationwide in 2004, hardly denting support for his Democratic rival, strategists in both parties agreed that he made progress in states like Ohio." Sure, the exit polls weren't spot-on last November, especially intraday. But must the Times really turn to bipartisan strategists to know how Bush did with blacks in 2004?

Instead, the Times could have cited exit polls like the one on CNN.com that shows Bush won about 16% of the black vote in Ohio, up from 9% in 2000. That's a lot more convincing than anonymous strategists.

As for "states like Ohio," Gelf has no idea what that means. Perhaps Oregon and Oklahoma, due to alphabetical proximity? Bush got N/A and 28% of the black vote in those two states, respectively. Or more likely, the author meant close states, like Florida (13%, up from 7%) and Iowa (N/A). Sure, there are no results for states with too few blacks; Gelf will take that statistical uncertainty over vague statements from strategists any day.