Can you really tell? Choose the perpetrator from the lineup. | Hacker News Reader

Can you really tell? Choose the perpetrator from the lineup.

picktheperp.com

11 pointsShamiq12 comments

I think this upsets me. How is this at all fair or good to the people involved? What is the intended message?

Edit: It has full names as the alt text for each photo. That is completely unnecessary.

Here's the open directory of all the photos: http://picktheperp.com/mugs/

Good questions. I don't agree with the practice of releasing mugshots for charges. If you're found innocent by the court, your mugshot is still out there on the net making you look like a criminal. It also seems to me to bias trials in advance. I don't understand why this practice hasn't been ended by legal challenge.

agreed. this seems somewhat dickish.

So I'm hitting 40% over 30 tries, twice what random chance would account for. What does that imply?

I asked myself the same question. Could be just a lucky run though. Unfortunately, I can not remember right now how to compute the probability that such a streak is a lucky run.

That profiling works. I was getting a bit better than 50%, surprised myself.

I'm not sure how you're defining "works". I couldn't get anything close to what I would call reasonably acceptable when trying to exploit known and my own stereotypes. Perhaps "may marginally increase % correct".

If I can consistently get the perp right half the time that's statistically significant which in my book qualifies as "works". That's all profiling is supposed to do, skew the odds in your favor.

After doing it a few times, it seems to "freeze" (clicking does nothing) in Firefox.

Are all these people criminals? That's different from a normal lineup, and makes it less interesting.

Given a bunch of criminals, I don't think I could pick who committed a specific crime.

What would be interesting would be to see if I could pick out a criminal from a group of non-criminals.

There's no reason to believe any of these people are criminals. They've all been arrested, but that doesn't make them guilty! That's why this site is disturbing.

Most real lineups consist entirely of criminals too. One person in the lineup is the suspect, and the rest are inmates from the local jail. You implicitly identified the reason for this: everyone in the lineup needs to look equally sketchy, or else picking out the suspect is too easy and weakens the evidential value of a correct identification. It's also just the easiest way to find volunteers to participate.

Reply on news.ycombinator.com