If they reflected larger trends, sure, there's probably precedent. There can be a sudden bump in count for a number of reasons, but it is statistically unlikely that it could be large and yet also buck the general trend. For those late-night ballots to have a chance at being real, they should have reflected close to the 50/50 split that was playing out in the rest of the states/country - or at least towed the line on a percentage basis compared to local results. (In other words, the odds say they should have just been more of the same.) And, when you consider that, in many states last election, Republicans requested more absentee ballots than Dems... yet the results were 99.9999% Democrat... nope. Then factor in the leaked JSON files that pretty much show how the data was manipulated... You'd have to be hockey-helmet-wearing retarded to accept the results of the last election... Yet, talk to 90% of people around you... Not enough helmets to go around.