Got it. In fact, I am way beyond you on this.
I have repeated repeatedly that the justice system is a repeat player. That is, it has to calibrate its expectations so as to afford more resources and credibility to the most serious and plausible allegations, and to calibrate burdens of proof and going forward in a fashion that (roughly) balances false-positives and false-negatives in an appropriate way, over the long haul. We as armchair observers of the justice system are also repeat players, trying to figure out what seems to work or produce “justice” over a large number of cases as to which none of us can ever have “all the unique facts.”
You’ve missed my point that criminal justice isn’t concerned with flushing out all the unique characteristics of every alleged crime. Harsh but true. Prosecutorial discretion means that some borderline cases get dropped, and there can be no question that some of this discretion is being exercised on the basis of past patterns and past occurrences.
No one has denied (I think) that actuaries make lots of money (as repeat players) over the long haul by making a bet with <individual> white smokers who like skydiving, but also have “any number of characteristics” that may distinguish them from other white smoking skydivers, that they will likely live X years, based on the past experience of other white smoking skydivers, despite the fact that any particular white smoking skydiver may have “any number of characteristics” distinguishing his situation from that of the prior risk cohort.
In observing and suggesting policy for the justice system or for our amateur interest in how it might be administered, we are indeed repeat players, just like the actuaries, and it would be going against human nature (and would be a money-losing proposition) to ignore willfully any statistically-corellated data pattern in doing so.