Link found on slashdot. While I can’t speak to the accuracy of the analysis, it is an interesting illustration of how election fraud could be discovered using statistical analysis. (In particular, there might be good reasons why larger precincts favor a particular candidate — smaller -> rural, larger -> urban.)
https://docs.google.com/file/d/0ByJAC-sfXwumZzI2bVlON2VTMnFyYVZZSnpDYnNyQQ
Makes the hullabaloo about voter fraud that can be prevented using ID cards rather “interesting”…… if these results are true, the nominations have zero legitimacy. And it was all done inside the machine.