So, how does one account for the 2% that didn't correlate? I get that the similarity's pretty striking, but what are the odds that they might both have most of the same data? It was a pretty darn large number, and isn't it feasible that the FBI may have also acquired their data similarly?