3 ms·

It's not easy to work out the statistics because the vast majority of crimes go unsolved. The sensational headlines are the ones where criminals managed to get

by ggreer 4y ago

It's not easy to work out the statistics because the vast majority of crimes go unsolved. The sensational headlines are the ones where criminals managed to get caught. The ones who were out on bail (or released on their own recognizance) but didn't get caught won't show up in the stats.

ravi-delia 4y ago
A cursory google suggests that 50% of murders go unsolved. Obviously people out on bail are far more likely to be investigated than those who aren't (especially considering that the intersection of "criminals who haven't gotten caught" and "people out on bail" is empty, so we're looking at the finer category of people who have been caught once), but lets pretend that the opposite is true, and that every single unsolved murder is committed by someone out on bail. That utterly made up statistic would still have more value than every such fearmongering headline combined.
- ggreer 4y ago
  There are many more crimes than murders, and the chances of getting caught are much lower. For example: in San Francisco, the clearance rate for reported larcenies is 2.9%.[1] So a thief can expect to steal another 23 times before having a 50% chance of being caught. SF's 6.4% clearance rate for vehicle theft means that a car thief can expect to steal another 10 cars before getting caught. 1. https://www.sanfranciscopolice.org/stay-safe/crime-data/clearance-rates-dashboard https://www.sanfranciscopolice.org/stay-safe/crime-data/clea...
  - robocat 4y ago
    I presume the actual distribution is highly skewed, so “expect to steal another 10 cars” is likely better said as something like “expect to steal twice as many cars than they have already stolen” plus a clearance rate for a first time offender. Obviously difficult to get reliable information on the true distribution - but perhaps can be predicted by the outlier fat tail? More simply put: much more than 50% of people that have stolen 1 car will get caught before they have stolen their tenth car, and if someone has stolen a hundred cars then their chances of getting caught before stealing 110 are much less than 50%. Edit: note that my figures don’t match your numbers - they are only examples.
- dragonwriter 4y ago
  > the intersection of "criminals who haven't gotten caught" and "people out on bail" is empty This is not likely true, and it is definitely not necessarily true. There are quite likely people out on bail who have (1) committed at least one crime, (2) never been identified and arrested or charged for any crime that they have committed.

by ggreer 4y ago

ravi-delia 4y ago

A cursory google suggests that 50% of murders go unsolved. Obviously people out on bail are far more likely to be investigated than those who aren't (especially considering that the intersection of "criminals who haven't gotten caught" and "people out on bail" is empty, so we're looking at the finer category of people who have been caught once), but lets pretend that the opposite is true, and that every single unsolved murder is committed by someone out on bail. That utterly made up statistic would still have more value than every such fearmongering headline combined.

ggreer 4y ago
There are many more crimes than murders, and the chances of getting caught are much lower. For example: in San Francisco, the clearance rate for reported larcenies is 2.9%.[1] So a thief can expect to steal another 23 times before having a 50% chance of being caught. SF's 6.4% clearance rate for vehicle theft means that a car thief can expect to steal another 10 cars before getting caught. 1. https://www.sanfranciscopolice.org/stay-safe/crime-data/clearance-rates-dashboard https://www.sanfranciscopolice.org/stay-safe/crime-data/clea...
- robocat 4y ago
  I presume the actual distribution is highly skewed, so “expect to steal another 10 cars” is likely better said as something like “expect to steal twice as many cars than they have already stolen” plus a clearance rate for a first time offender. Obviously difficult to get reliable information on the true distribution - but perhaps can be predicted by the outlier fat tail? More simply put: much more than 50% of people that have stolen 1 car will get caught before they have stolen their tenth car, and if someone has stolen a hundred cars then their chances of getting caught before stealing 110 are much less than 50%. Edit: note that my figures don’t match your numbers - they are only examples.
dragonwriter 4y ago
> the intersection of "criminals who haven't gotten caught" and "people out on bail" is empty This is not likely true, and it is definitely not necessarily true. There are quite likely people out on bail who have (1) committed at least one crime, (2) never been identified and arrested or charged for any crime that they have committed.