For my statistically-minded colleagues out there:  is the Florida recount supposed to catch random errors and omissions?  If so, wouldn't changes or modifications be randomly distributed between the candidates in some proportion?  Since this was nearly a 50-50 race, shouldn't the "errors" be somewhat equal for both sides....hmmmm...... 6 to 1 errors found....Bush's Florida lead goes from 1800 to 300....is that statistically significant?

Dan