View this email in your browser
We wrote a while ago about false positives, base rate neglect, and the fatal problems that creates for a system designed to predict who shouldn't have a gun. This week, Cam Edwards wrote up an object lesson on that from New York. A man got arrested for possessing a pistol without a permit, and for firing it unsafely (he hit a parked car). Then this happened:

"After King was arraigned at New Lebanon Town Justice Court, his spouse says he made a comment in the parking lot about how it would be easier to kill himself than deal with New York State laws. She says the comment was taken out of context, and he wasn’t remotely serious about it.

"Troopers took the comment seriously, later filing the Red Flag documents against him. It’s the first Red Flag case in Rensselaer County since the law took effect late last month."

Cam continues,

"King was back in court on Friday, where he agreed to give up his guns for a year, based on the judge’s decision that there was 'clear and convincing evidence' that King posed a danger to himself and others."

Also this week, data came out that since New Jersey's red flag law took effect on September 1, an average of one gun owner per day has been red flagged.

This is easy and natural for people to shrug at. "Some people shouldn't have guns. So we do our best to figure that out, and if we get it wrong, we can fix it. Yes it's messed up that they took this guy's guns for mental health reasons without involving any mental health experts, and yes this will discourage people who actually need treatment from coming forward. But that's a kink to be ironed out. It doesn't invalidate the whole concept."

Check out the base rate neglect article we linked up top to see the exact math on why this very natural line of thinking is provably incorrect. The counterintuitive math is that if:
  • You have a red flag system that only has a 5% false positive rate (which is unheard-of-ly excellent, but let's be maximally generous)
  • A 0% false negative rate (again, generous)
  • The population you're screening is 2 million people
  • And let's say 5000 people in that population are able to pass a NICS check but should be red flagged (again, likely generous)
If your system red flags somebody, what are the chances that that person is one of the 5000 true positives, i.e. that you don't have the wrong person? Intuition, given the numbers above, is that there's a 95% chance you're right. But the truth is surprising: run the numbers and there's a 95.2% chance the system is wrong.

Here's the math on why.

This week's links

Study finds fame-seeking mass shooters tend to receive more media attention
More research on contagion theory.

National Review just released a whole issue about guns
Bunch of good stuff in there from some of the smartest folks on gun rights Twitter.

T.Rex Arms with that  c o n t e n t
Their YouTube channel has been on fire lately.

NYT: an assault weapons ban just failed to get enough support to get a floor vote in the House
We should take a second to acknowledge this. In 1994, an AWB became law. In 2019, in a Democratic-controlled House, an AWB didn't have enough support to get put to a vote.

It's easy (and understandable) to forget, but in a lot of ways, gun rights are actually making extraordinary progress. If we want to win, step 1 is to act like winners. We wrote this up a little while ago:

Lots of work left to do and battles to win, but if we keep doing the work diligently, there is a lot of cause for optimism.

The truth about "assault weapons"
This classic was filmed around 1989, when California was passing its first AWB. A friendly mustachioed retired cop straight out of central casting patiently explains the absurdity of an AWB, and transforms a Mini 14 from a legal config into a banned one, just with some ergonomic changes. It's a good all-audience video to share with normies who want to learn more about the issue without being preached at.


The complete archive of Open Source Defense newsletters.
Copyright © 2019 Open Source Defense, All rights reserved.

Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list.

Email Marketing Powered by Mailchimp