Hacker News new | comments | ask | show | jobs | submit login
How DNA Transfer Framed Lukis Anderson for Murder (themarshallproject.org)
98 points by cafogleman 9 months ago | hide | past | web | favorite | 10 comments

This is one of the reasons I'm categorically against any kind of DNA databases. Yes, they can be useful. But if a crime is a bit harder to solve and there is a nice & convenient match against a DNA database of for instance known offenders (or babies whose blood was sampled right after birth...) then policework will shift from 'whodunnit' to 'howdidtheydoit'. Rationalization after such a detail comes up is a path that is a lot more tempting than to come up with suspects for which there is no evidence at all.

> a limp handshake relays less DNA than a bone-crushing one

Bowing seems to make good sense in this context.

Finally, shocking he wasn't compensated for spending a long time in jail.

'Hard cases make bad law', as the saying go. We know they won't lead to a massive crime wave of false convictions because DNA databases generally aren't abused like that, and the effects on crime are enormous: "The effects of DNA databases on the deterrence and detection of offenders" http://jenniferdoleac.com/wp-content/uploads/2015/03/DNA_Den... , Anker et al 2017.

Part of the problem is that as the size of the database grows -- i.e., as every police encounter turns into a mandatory sample collection -- the false-positive rate will also go up.

Remember: DNA "matches" are probabilistic, and most of their utility has been based on the low probability of a sample taken from a crime scene "matching" a sample taken from a suspect. Prosecutors (committing the prosecutor's fallacy, of course) love to tout the "one in a billion" type statistics for a match being a false positive, but it doesn't take a very large database at all for "one in a billion" coincidences to start occurring.

These laws make hard cases easy cases. The only reason it didn't work here is because there were some very exceptional circumstances. Even the suspect wasn't sure he had not done it and the combination of DNA evidence + plea bargaining skews terribly against suspects, even innocent ones.

DNA evidence is like a magic bullet for crime, DNA found -> guilty is the norm, not the exception.

As far as the statistics are concerned: we do not actually know the false positive rate.

DNA databases, that haven't been around that long and haven't been especially useful because of not having a lot of data in them in comparison to all the possible DNA data, generally aren't abused in the way they would be if they had all the DNA data and were a common tool.

One the other hand, the DNA database was what found out the real murderers.

It is what first found them out.

Nevertheless, it is likely that a traditional investigation would have found them out too. Especially through the link between the victim and the sister, coroborated by phone records. Finally, their culpability could have been put beyond doubt by using a DNA test.

A sampling of nice good articles about Forensic Pseudoscience, courtesy of Google search:


This is a broad-based problem.

> The participants unwittingly brought with them alien genes, perhaps from the lover they kissed that morning, the stranger with whom they had shared a bus grip, or the barista who handed them their afternoon latte.

Another conspiracy theory that became true. Is there a list already?

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact