Hacker News new | ask | show | jobs
by fauxpause_ 1145 days ago
This butchering of statistical reasoning is so poor I have to assume it is a troll

But in case it’s genuine, consider vast differences in gun ownership, homicide rate, and… honestly so many other things it’s just annoying to list them.
1 comments
nawgz 1145 days ago
> consider vast differences

I did.

Gun ownership is about 32% of individuals in USA vs 25% of individuals in Canada.

Homicide rate is about 2.5x higher in the US, which means US cops are outperforming this statistic vs US civilians.

> so many other things

It's not just annoying to list them, it's ridiculous; these countries are the exact same place. Across the board, everything you can think of has a lot of statistical homogeneity between these places. But Canadian cops are way better educated, and kill way less.

Around the world, cops that are better educated kill way less.

So why exactly are people excusing reductions in education requirements for police? It makes no sense. There's no supporting evidence that that's a good choice, and literal mountains showing that it's a bad one.

This thread is bizarro world.

link
fauxpause_ 1144 days ago
Your analysis is just dumb. You’re implying this is the only difference. You’re implying the entire outcome difference could be due to this one variable. You have shown no evidence to suggest this relationship exists at all. Like, for example looking at other countries. You’re implying that one thing has to account for the entire effect as a linear relationship by saying homicide rate being 2.5x is not important or that they’re “over performing in spite of it”. But you want one? Fine. Cops killed on duty. About 10x in the US. That settles your desire for a linear relationship on a compelling alternative story. Is that actually a sufficient analysis? No.

You get an F for intro to stats. And an F- for being confidently wrong.

There is no “excusing” reductions in education. You have not demonstrated a relationship between these two things as important.

link
augment003 1145 days ago
Have you considered handguns as a factor?
link
nawgz 1145 days ago
Handguns are the most interesting statistic here, with the US being a serious outlier (58%! vs 12% in Canada), it could be related

I still don't see how you deny that more educated officers are less violent when I've made multiple links supporting that argument

link