In Ohio's 2nd CD race, Paul Hackett graciously accepted his close 52%-48% defeat.

But bloggers and activists who watched the GOP steal Ohio for Bush in 2004 think they may have stolen this race too.

Billmon lays out some of the early statistical evidence, which can be suggestive but is rarely conclusive.

The best way to evaluate statistical claims like these is to analyze the precinct-by-precinct results:

1. Compare the precinct vote tallies (how many votes each candidate received) with the sign-in sheet tallies (how many voters signed in). These should be identical - if not, any significant discrepancies would be a red flag.
2. Compare the precinct vote percentages by candidate in this election, with the same precinct percentages by candidate in November 2004. These would not be identical, but they should be pretty similar.
3. Examine the sign-in sheets to make sure the voters who voted were true local residents, not outsiders bussed in (or mailed in) by the GOP.

If you see any work like this posted on the Internet, please link to it below.