That is a very good point, and the result of that feeds back into the prosecutor's fallacy. If you have plotted the routes of a million people traveling through

That is a very good point, and the result of that feeds back into the prosecutor's fallacy. If you have plotted the routes of a million people traveling through a city, and there is an incident (let's say a bombing at a stadium), then you could "cold hit" someone from that database of routes to find someone with a movement pattern that appears to suspiciously correlate with the bombing, but is in fact just one of a few hundred people with similar movement patterns (only one or two of whom were actually involved). Then you link that to the person having demonstrated at a couple of, say, animal rights protests, get a few neighbors who say that the guy was "weird" and "acted strangely", and bam - you have what looks like an open and shut case of a radical animal liberation terrorist, but is in fact just a vegan who happened to pick up some gardening supplies at a Home Depot and a Pike's Nursery two days before going to a baseball game.