Hacker News Reader

Comment by mschwaig | Hacker News Reader

'So we have these fingerprints, and we think they belong to this guy we already have prints on file for. Can you give us a yes/no answer if they match up?'

seems like a pretty low bar for evidence. Seems like the kind of thing that could heavily skew towards telling you what you want to hear. Maybe someone else knows if it actually works like that, the writeup made it sound like that to me.

I'm just some guy who saw a tv documentatary at some point about how forensic techniques that worked like that got called into question when conflicting DNA evidence started turning up.

View on HN

My understanding from skimming the article is, they only identified the suspect from fingerprints. After identification they did surveillance and gathered additional evidence.

> The FDL [HSI Forensic Document Laboratory] returned the request after conducting a comparative analysis of the friction ridge detail of the fingerprints from the Imgur album and the fingerprint samples taken after police had arrested Porras for a different crime. The fingerprints in the Imgur album matched the prints they already had on file for Porras.

It doesn't sound like that to me, but maybe I am misunderstanding what a comparative analysis would entail.

It doesn't say they arrested him immediately after the match. If you read the article they say they placed multiple orders and would surveill him after every order. That's how they gathered evidence.

I did read the article. I am interested in how forensic evicende like that is gathered in general and how reliable it is in general. It's debatable what role that fingerprint played in that investigation, but I did not want to call into question that particular outcome.

For example my impression is that DNA evidence is very reliable while for example optically matching hairs or matching bite marks, which I think was done in the past in a similar 'does A match B setup' is fairly unreliable.

It's interesting to me both from both the 'what bar does evidence have to meet to make sure there are no false convictions' side as well as the 'what are the privacy implications of posting pictures with fingerprints in them' side.

Although not mentioned clearly in the complaint it's very possible the initial matching was done through the FBI's or another automated query system. This is a standard practice used in law enforcement investigations - capture fingerprints however you can and submit them to the FBI, which returns a report of possible matches. The automated matches would need to be followed up on with manual comparison, esp. due to the unusual nature of the fingerprint capture here, and this manual comparison is going to be the part mentioned in the complaint because it most clearly establishes a link.

> Law enforcement made a number of controlled purchases during the investigation into Porras and his co-conspirators. The purchases and subsequent surveillance followed the same pattern every time: make a purchase; watch Porras drive to a storage facility where he stored product; follow Porras to the Post Office; talk to Postal Inspectors about the package Porras or his co-conspirator had dropped into a USPS Blue Box.

Reply on news.ycombinator.com