Good Point, Bad Math: DNA Database Statistics Misunderstood (Again)

An op-ed article in yesterday’s New York Times speaks of DNA database hits as “high risk.” That police, prosecutors, defense lawyers, and juries should not accept every hit as conclusive proof of guilt is quite true. To prove the point, the article calls attention to a false arrest in the ongoing California case of the murder of millionaire investor Raveesh Kumra — a matter that made the news last month. Although the facts of the case are atypical, it is a warning to overly credulous police and prosecutors.

The article, by Hastings Law Professor Osagie K. Obasogie, also rehashes two dated — and previously refuted (or so I think) incidents — that are supposed to show that DNA databases are producing lots of false hits. Some of the factual and statistical errors in the article are identified in the Forensic Science, Statistics, and the Law Blog (