For people wondering why this is bad: Let’s say you have a DNA test that’s 99.9% accurate. If you find a suspect through normal investigative means and then test his DNA and he matches, you‘ve almost surely got the right guy.<p>If you geofence the crime scene and test/Ancestry-scan the 10,000 people who went through the crime scene that day, you’ll get a DNA match for 10 innocent people, and a lay jury won’t have the statistical fluency to understand how the two scenarios are different.