I haven't said that I agree with analyses but I'm really mean it when I say that there are more good people than bad people who use bitcoin. Don't you really think that's real? I don't think their calculation of number of coins used for bad activity is correct, that's probably wrong because they often fail with the correctness of their AML score and we don't exactly know how they mark coins bad and wrong. Probably, every coin that went through mixer is marked as bad by them and that gives us a very unfair result.