Firearms by the Numbers

This is one of those articles where a data scientist hobbyist appears to completely ignore the experts of the field, make a few graphs, and think they’ve come to some new insight to an extremely complicated problem. There are people who spend their entire career studying this issue. I am sad to say I do this sometimes too - my girlfriend works in healthcare, and sometimes I’ll say things that are clearly wrong, but since I don’t work in the field, I miss a lot of institutional knowledge. And as a professional scientist, I see many, MANY people making similar errors when it comes to my field of expertise. I don’t want to stop people from being curious - on the contrary, being curious is great! But, recognizing that just being an educated individual does not mean we are qualified to talk about things that are not our area of expertise is also extremely important.

This ad hominem response doesn't refute a single data point or interpretation the author put forward. If the institutions that historically control the narrative are the only ones authorized to interpret the data then we may as well not have access to the data; we should close our eyes and allow them to lead us by the hand.

I think the issue is which “experts” are talking. The reality, is that there are no “firearm data science” experts, unless you decide to count the NRA or something.

My overall point here, is that even the narrative given “guns cause violence” is objectively not true. We can prove that. Similarly, we have substantial evidence to the contrary. In fact that the firearm homicide rate simply correlates with overall homicide rate.

That is much higher in certain population(s) of the U.S. which leads to a dramatic skewing of any analysis. WHICH btw I’d argue the experts are intentionally neglecting discussing.

This analysis is willing to play footsie with the third rail. No one should be unwilling to engage with verifiable data.

I’m not convinced about the usefulness of homicides per gun. You’d expect dilution, especially as the number of guns approaches the number of people. Gun deaths per gun owner is more useful, and plotted against the number of guns per owner might be useful. Context is important- would the same homicide have been committed if a gun was not available?

Data science here would also see how independent gun ownership is from other variables associated with homicide and suicide. I didn’t get much of that.

Parkland, Columbine, Las Vegas, Pulse all involved easily accessed guns. The data shown here does not discuss mass attacks, ie. number of people killed per event. They could have been accomplished with other means, but likely wouldn’t have.

Of course, the best numbers available for ownership is legal ownership. Illegally possessed firearms used in crimes are a real problem, and it’s not likely we’ll remove that problem without removing most all guns. Yes, if guns are illegal then only criminals will have guns, but the same is true for cannabis, ice cream and blue jeans.

I don’t have any use for guns but I don’t mind other people who do. I would rather not be shot at (pretty sure it happened once) but I have more to fear from distracted drivers killing me on my bike.

The social reasons that people kill each other are more valuable to address. For example, don’t fill people with rage at being outcasts.

“ It’s my opinion, retaining human rights is why firearms are important.”

He explicitly states it’s an opinion, e.g. not backed up by or refuted by data...