News for hackers. Stuff that matters.
Re:Washington Post publishes dataset of 52,000 criminal homi... by a_bonobo on Wednesday June 06, 2018 @09:45PM (#17252603)
$ curl https://raw.githubusercontent.com/washingtonpost/data-homici... | grep 'Closed by arrest' | wc -l

25667

$ curl https://raw.githubusercontent.com/washingtonpost/data-homici... | grep -v 'Closed by arrest' | wc -l

26506

Wow that's depressing. All possible outcomes via

$ curl https://raw.githubusercontent.com/washingtonpost/data-homici... | cut -f 12 -d ',' | sort | uniq -c

          1 Closed by arrest
      25673 Closed by arrest
       2922 Closed without arrest
          1 disposition
      23583 Open/No arrest
Pretty sure that 'Closed by arrest' also regularly results in a 'Not guilty' down the pipeline, so the 'real' number is probably much lower.
Re:Washington Post publishes dataset of 52,000 criminal homi... by Pywarrior on Thursday June 07, 2018 @03:26AM (#17253953)
It's Closed without arrest and Open/No Arrest that are most concerning. Closed by arrest indicates effort was probably spent during an investigation, the bare minimum assumption. An assumption that can not be made about the 26,505 Closed without arrest and Open/No Arrest murders.