De-Obfuscating the Statistics of Mass Shootings

http://www.charlespetzold.com/blog/2015/07/De-Obfuscating-the-Statistics-of-Mass-Shootings.html

by FigBug • 11 years ago

188 166 11 years ago

aftbit• 11 years ago

Interest groups lying with statistics is a specific example of confirmation bias. It's sadly so prevalent that I generally assume that any statistics I find are intended to lead me in a particular direction.

I am skeptical of the final implication of the blog entry, that improved gun-safety legislation would solve this issue. For someone taking such a strong approach to statistics, it's a shame that they focus on only one facet of such a complicated situation. American culture, mental health treatments, social programs, and gun laws are all different from most other countries on their list - it seems disingenuous to mention only the last.

ta929292• 11 years ago

I know this is just pissing in the wind (and you just used the term from the article), but "gun safety" already means something. <https://en.wikipedia.org/wiki/Gun_safety> This type of language engineering annoys me.

apolretom• 11 years ago

"This article disagrees with my priors, therefore it is wrong."

vanderZwan• 11 years ago

The blog concludes that the abuse of statistics is politically motivated by a vested interest in keeping gun-safete legislation minimal. This is not the same as implying that there should be more gun-safety legistation.

kromem7• 11 years ago

Your comment is actually a perfect example of why I have issues with using per capita as the basis of the analysis.

I adjusted the OCED numbers to be on a per-gun basis as opposed to a per capita basis and non US OCED countries combined were more likely to have rampage shootings per gun than the US.

There are not issues in the US with social programs, mental health, etc as compared with other countries when looking at mass shootings. The only number that plays any significance is the number of guns per capita to influence the number of shootings per capita.

If other countries had as many guns per person as the US, the data indicates they'd proportionally have the same amount of shootings per person or more.

lagadu3• 11 years ago

The problem here is that you're assuming there's a correlation between guns per capita and shootings per capita, which is similar to the (potential) fallacy that the OP made: that there's a correlation between gun restriction laws and shootings per capita.

jcromartie• 11 years ago

In the USA, people who own guns are likely to own more than one gun. They may be hunters or collectors. Per-gun is fairly meaningless.

gadders• 11 years ago

Do these counts of the number of guns include the number of illegally held guns as well?

wongarsu• 11 years ago

To a certain extend it makes sense that more readily available guns cause more shootings, but I highly doubt that it's a simple linear relationship.

If I don't have access to a gun I can't go on a rampage shooting. If a friend of mine has a gun I could go on a rampage shooting, but I'm probably less likely to do it on impulse. If anybody in my household owns a gun I can likely go on a rampage shooting whenever I like. But if I own 10 guns I don't see how that makes a rampage shooting any more likely.

The proper statistic would be rampage shooting per household with at least one gun. But even that's probably still very flawed (are dangerous people more likely to own guns?)

eli_gottlieb• 11 years ago

Unless social programs turn out to be unexpectedly cheap or gun law enforcement unexpectedly expensive, gun laws are probably the cheapest, smallest-delta intervention against mass shootings.

hga22• 11 years ago

Gun control cost the Democratic party control of the Congress in 1994 and the Presidency in 2000. That's about as "expensive" as policies get in the US....

And as noted in this discussion, any attempt to squeeze tight enough that there might actually be an effect on mass shootings of the current type would replace them with fantastically more "mass" shootings of the civil war type, which tend to be only exceeded by foreign invasions as "expensive".

vacri2• 11 years ago

Spree killings/mass murders/rampage shooters/whatever are all a canard anyway. A canary in the coal mine. Add up all the spree killers in the US and they'll not break 1% of the murders. Just like the debate over automatic weapons is a canard, since they're very rarely used in murders.

The US has a murder problem. It's murder rate is four times that of it's industrialised contemporaries. Most of those murders (~75%) are done with firearms, and most of those (>75%, can't recall exactly) are done with handguns. And by someone who knows the victim closely.

The stereotype of the inner-city gangsta gunning down a gangland enemy with an automatic weapon is an outlier. And the same is true of the spree killer. As long as debates on either side center on these kinds of outliers, little will be achieved. The murder rate isn't going to be affected by open carry, since most murders aren't of strangers. And banning automatic rifles isn't going to do anything either, because it's handguns that do the dirty work.

The problem is that these details aren't sensationalist. Being gunned down by a stranger on a rampage makes the news. Being killed by a crazy relative doesn't.

151553• 11 years ago

because it's handguns that do the dirty work

I'd also like to point out that the Supreme Court of the United States of America affirmed that individual ownership of semi-automatic handguns (cited example, but, "common use" is the test) is a protected right.

We have a financial inequality and mental health problem (somewhat stemming from said inequality) in this country, not a firearm problem. These are the problems that need addressed to reduce crime.

x0x0• 11 years ago

The justices, as aptly demonstrated recently (viz scalia, whose textualism has no meaning besides whatever his personal politics prefer) are politicians, no more. Or one could ask why gay rights have done better than reproductive rights, and the answer comes down solely to Kennedy is more sympathetic.

At this point, we are so inured to gun violence that my SO was at a community college a couple years ago while it had a lockdown because of a shooter on campus. She was ushered into a classroom by swat with them standing outside the door and it didn't even make the tv news that night. I grew up shooting guns with my father and have treasured memories of going to gun shows. I'd vote for an amendment for Japanese gun laws.

ps: in 2008, the us had 12k+ firearm related homicides while japan had 11. The population ratio was 2.39, so the US would have experienced (rounding up) 27.

wisty3• 11 years ago

Yep. Banning guns will reduce mass shootings, and gang shootings (somewhat - well organised criminals will still have guns, but the stupid ones who cause the most damage won't). It can also reduce suicide rates (since guns don't give people time to change their mind).

However, a crazy relative can still kill you without a handgun. In countries that introduce gun laws, they get more murders with knives, blunt objects, etc.

US murder rates may be partly driven by guns. But they're mostly the result of education, prison sentences, and other social factors. The US is a developed country with the welfare system of a developing country (and a bunch of very expensive bandaids to catch the people who have fallen too far to really recover).

bmelton14• 11 years ago

I hate to be pedantic in discussions like this one, and please know that I certainly don't mean any offense beyond correcting a technical error that perhaps nobody really cares about.

"Just like the debate over automatic weapons is a canard, since they're very rarely used in murders"

There is no debate surrounding automatic weapons, or, at least there isn't one on the national stage. Automatic firearms are dramatically different from semi-automatic firearms, and it's been a (specifically expressed) political maneuver to confuse/conflate the two.

The number of deaths attributable to automatic firearms is effectively nil, as they've been highly controlled since the days of Al Capone and his ilk.

When people are referring to "automatic firearms" in today's political arena, they almost always mean to say "semi-automatic firearms", which most closely maps to the sorts of weapon used in the Newtown and Aurora shootings.

It's a subtle, but important distinction, but as those who are trying to conflate the terms generally intend to do so to confer additional scariness to semi-automatic firearms that are simply not there.

vacri2• 11 years ago

It's fine to be pedantic against someone who is claiming the issues are not what they seem. Even encouragable :)

You are correct of course, regarding auto vs semi-auto, but I mean that the anti-gun lobby makes a lot of noise over 'big scary rifles' when it's actually handguns doing the most damage.

I'm for gun regulation myself, but even if do rifles get heavily regulated in the US, it'd do very little to the stats. It'd actually work against the anti-gun lobby, because the pro-gun lobby would be able to point to the relatively unchanged stats.

roel_v• 11 years ago

"The number of deaths attributable to automatic firearms is effectively nil, as they've been highly controlled since the days of Al Capone and his ilk."

Well, since we're going to be pedantic, this should really say "the number of deaths attributable to legally owned automatic firearms", since there is quite a bit of criminal violence (as in, violence to support criminal activities, like drive bys, rip deals resulting in shootings etc) and deaths involving automatic weapons (although it's quite hard to find statistics on this, to my surprise). MAC 10's and Mini Uzis are rather popular for that sort of shenanigans, because of their small size (thus easily concealable). An AR15 style weapon (which is usually the type of weapon people talk about when the discussion is about 'assault rifles') is very impractical for such a thing, as per your point.

sudioStudio648• 11 years ago

I have friends who have class 3 licences and they say that the strictures in place on automatic weapons are why you don't see then used in crimes very often. They are heavily regulated and tracked.

As to your point, I don't fault the antigun crowd for using any tactics that work. The progun crowd is certainly guilty of he same thing.

sudioStudio648• 11 years ago

That's absolutely true. The discussions on this site tend to the abstract and miss this point.

pjc503• 11 years ago

The US has a murder problem

The US has a gun violence legitimization problem. Yes, other countries have guns. But America has a particular culture surrounding the idea that shooting someone whom you percieve to be a threat is a reasonable thing to do, and moreover you're likely enough to encounter this situation that it's worth arming yourself beforehand.

Whenever one of the controversial police shootings happens, there are plenty of commentators on hand to say how the victim deserved it (and therefore his grieving relatives should shut up and stop complaining about the police). But this attitude isn't limited to the police, it's part of wider American society. So whenever two people get into the kind of stupid altercation that in other countries results in black eyes and lacerations, in the US somebody might pull a gun. Undoubtedly this makes a lot of domestic violence situations worse.

The US concept of "right to life" is applied really narrowly, leaving both judicial executions, police shootings, and wider acts of negligence in public safety untouched.

Not to mention the racism problem. You don't need guns for this, other forms of attempted mass murder are available: http://time.com/3942688/black-church-burning-mount-zion-ame-...

Houshalter• 11 years ago

But no one is afraid of being murdered by a relative. Mass shootings are what people are afraid of. Mass shootings are what make the news and get people's attention.

The issue is entirely about fear. If people really cared about the most effective policies to decrease premature deaths, maybe we'd all be arguing about traffic safety issues or banning swimming pools or something.

rayiner3• 11 years ago

The interesting thing is that we've always had a murder problem: https://en.wikipedia.org/wiki/List_of_countries_by_intention.... We've been at 5-10x the per capita murder rate of England and Wales going back to at least the 1910's. It looks like between 1910 and today the lowest rate in the U.S. was 4.0 murders per capita, in 1957. That same year, England and Wales had only 18% as many murders as the U.S. per capita.

I think the scale of violence here is really hard for Europeans to understand. New York is one of the safest large cities in the U.S., and the murder rate there is more than 3x that of similarly-sized London.

hga22• 11 years ago

Then there's the flip side about how many people died from state sanctioned violence in Europe since 1910 (and going back a long ways; see especially the Thirty Years War). We in the US only understand that from immigrants who wanted to escape that scene, as well as the usual media, and it is one of the primary motivations of a bunch of us pro-gun types.

Neal Knox, one of the top figures there in the last century, got his start at boot camp, where he heard an eyewitness story from a recent immigrant from the Netherlands. The Nazis went to a family's house where a handgun was registered to the homeowner but since he couldn't find it, they took the whole family and machine gunned them in the village/town square.

caf2• 11 years ago

I believe the statistical lesson here is widely applicable, and may well be of use if data analysis - particularly of rare events - is part of your job.

praptak• 11 years ago

This statistical lesson is actually a textbook error, similar to taking a p=0.05 test on multiple subjects separately.

What makes this interesting is the political context.

caf2• 11 years ago

Certainly, but I have seen many people tasked with analysing data in a work context who have never picked up a stats textbook.

platz• 11 years ago

What is the formal name of this error?

kromem7• 11 years ago

It's disingenuous to compare shootings purely by a per-capita basis. It should be per-capita among gun owners.

The data as shown indicates the US has a cultural mass shooting phenomenon, but it's simply not accurate. The US just has more guns, and mass shootings happen at extremely rare rates among gun owners everywhere.

This approach would be like comparing drunk driving fatalities in Los Angeles vs Venice, Italy.

Edit: I was curious and just ran the numbers with the full OCED countries. The US had 0.136 rampage shootings per gun (compared to 0.120 rampage shootings per capita), and non-US countries had 0.177 rampage shootings per-gun (compared to 0.246 rampage shootings per capita). I used the author's numbers and the Wikipedia page data for guns/100 residents.

habitue• 11 years ago

This is exactly the point. The variable in question is whether to try to reduce the gun-owning population to get shooting rates down. It sheds no light on that problem (and makes no sense) if you restrict the shootings to the gun owning pulations.

girvo2• 11 years ago

It's disingenuous to compare shootings purely by a per-capita basis. It should be per-capita among gun owners.

First off, I don't live in the US and I don't have a horse in this race.

You may have a point. However, wouldn't the fact that having a large number of gun owners directly correlating to mass-shooting fatalities still come down to the same conclusion: that guns should be regulated further? Or do you think that if the analysis was done it'd show something different to what the analysis in the original article shows?

bmelton14• 11 years ago

First off, I don't live in the US and I don't have a horse in this race.

Not living in the US presumably subjects you to position bias. Those who live in countries where gun rights are strong tend to favor gun rights. Those who live in countries where gun control is strong tend to favor gun control. This is true of most things, as the moral compasses of most are drawn from what is currently normative to their existing environment. This is true of most places in which socio-political biases are codified into law, whether regarding religion, gay rights, whatever; Change in these areas tends to come slowly if at all.

guns should be regulated further

Perhaps they should, perhaps they shouldn't. I have my bias, as I'm sure does everyone engaged in the discussion, but, at the end of the day, the document of the highest possible authority suggests that the citizenry should have access to guns, and the only way to overturn that is by ratifying that document. Thankfully, the document itself offers prescriptive measures for such ratification, but at the end of the day, without a strong enough amount of support throughout the nation and states for such a measure to pass.

At present, such support does not exist, so our politicians play different games on what does or does not constitute "infringement".

In the past, the argument was that it wasn't infringing if the person who wanted to exercise gun rights wasn't a member of a militia. We can prove historically and grammatically that this isn't true, but it took the Supreme Court's agreement for the legislature to accept it.

Then, the argument was that sure, it's an individual right and not limited to those in the militia, but hey, the second amendment says you can keep and bear arms, but it doesn't say where, so instead, the government insists that the right is only applicable to within the home. Have as many guns as you like, and carry them anywhere you want, as long as it's inside your home.

It took Chicago v. McDonald getting that one overturned in 2012 by the Supreme Court.

So, legally speaking, we know that the right to keep and bear arms is an individual right, disconnected from membership in a militia, and it extends outside the home. To those who favor gun rights, the matter might settled, but to those who favor gun control, it's a challenge to come up with a new rule that doesn't directly contradict those recently affirmed rules by the Supreme Court.

In D.C., the argument is currently that the right to keep and bear arms is only available to those with a heightened need... but what other right in our Constitution merits such a high bar to exercise? Would it be okay to restrict voting only to those harmed by the current administration? Would it be okay to limit speech only to those who can show that they have been victimized?

Anyway, I'm obviously ranting, and there's certainly more of my bias in this post than I intended, so I should shut up. The point, that I'm sure I made poorly, is that the law here says that we can have firearms, and that not even the elected legislature has the power to overturn that without ratifying the Constitution... unless they can come up with something legally clever, or wait until there exists a more favorable court to their cause.

kromem7• 11 years ago

If your only concern is reducing the number of mass shootings, yes, that is exactly (and the only) correct answer from the data. The only reason why the US has more of these shootings is because we have more guns, and reducing the number of guns will reduce the number of shootings.

However, the fact that these are extremely rare occurrences needs to be noted. Gun control has a lot more factors that are probably much more important than the rare issue of mass shootings.

To put it in perspective, doctors' sloppy handwriting kills roughly 7,000 people annually. That's far more than combined spree shooting victims. Certainly mandating doctors to type out all communication would be much easier than the herculean task of ridding the US of guns.

And the broader gun control issue has other issues and data that would need to be examined. Unfortunately, we don't have particularly good data regarding guns in the US due to the effective ban on government funded gun data (thanks NRA!).

Personally, I'd like to see the first step in gun control being federal funding directed toward research. Do guns (concealed carry, etc) actually deter other forms of violent crime? How many homeowners have used guns in self defense during a home invasion?

For example, a study in the 1990s found between 800,000 to 2.4 million defensive gun uses annually (Kleck & Gertz). And apparently while in the UK roughly half of all burglaries are when homeowners are present, in the US it's only 13%. There's also some data that suggests increases in crime rates in the years after gun bans, but the data comes from strong pro-gun sources and I'm a bit skeptical as to the validity.

Basically the data I could find regarding general gun control is spurious at best. We need better data on the pros and cons of gun ownership, and our decisions should be driven from that data and not anecdotal or tragic but extremely uncommon (statistically) events. So in my opinion, the rallying cry when national attention is on guns should not be taking head conjecture from all sides, but rather a national plea for better data so we can have an informed opinion on the issue.

demallien• 11 years ago

Noooooo, that's the whole point! The question that is being asked is not "how to make gun owners less likely to shoot people" but rather "how to make it less likely that people get shot". Per capita is absolutely the right measure to use for the latter question.

kromem7• 11 years ago

In terms of rampage shootings, the way to reduce the number of shootings is to reduce the number of guns. That's what the data shows.

A per capita analysis starts discussions about how the US is different in multiple ways. Oh, well, in the US a person is 5x more likely to go on a shooting rampage than in other OCED countries. Maybe it's US media coverage? Maybe we don't have the background checks or psychological evaluations other countries have. Maybe it's our action movies?

No.

It's simply that we have more guns. In fact, on a per gun basis, other OCED countries combined have a 50% higher rate of spree shootings. We just have a LOT more guns than other countries.

Now, whether this is a good or bad thing is a much broader discussion. But the ONLY data driven answer to "why does the US have more spree shootings than the rest of the world?" is: "Because the US has more guns".

TL;DR: A per-capita analysis leads to sociological problem solving as the base metric is people-based. A per-gun analysis cuts directly to the significant factor.

kromem7• 11 years ago

Edit: It's been to long since I last edited and I think I've been locked out from an edit, but it should be noted that those numbers in my last edit are PER MILLION (for both per-capita and per-gun).

ajuc2• 11 years ago

Of course mass gun ownership is the source of problem. Nobody claims otherways.

Some people just want to solve the problem by changing the law so that people have fewer reasons to own guns.

kromem7• 11 years ago

Actually quite a lot of people claim otherwise. If you watch mainstream news after a massacre, you frequently see everything from video games, to the state flag, to national attitudes regarding mental illness, to bad parenting, to bullying, to nearly any other non-gun topic blamed for the tragedy.

As long as mass shootings are discussed as a uniquely "US-issue" and not simply as a uniquely "Gun-ownership issue" these red herring discussions will continue. Especially as the narrative that the US is different from the rest of the world is one we love to tell (even when negative). There is a difference, but that difference is only the # of guns. All other sociological differences don't seem significant.

Now as I've said elsewhere in this thread, the issue of gun ownership has a heck of a lot more nuance than the two sides seem to recognize, and personally I find that I don't have enough solid data to be pro-gun or anti-gun, and mostly just want for the country to finally agree this is an issue worth spending national dollars on studying.

But I think so far we're much more likely to end up with studies on video game violence, mental health treatment models, etc, as NO ONE on TV utters anything but rhetoric after these events, and ironically these extremely rare and statistically insignificant (but anecdotally significant) events are the only time it comes up at all.

In fact, the leading cause of gun deaths in the US is suicide. And there's actually some REALLY interesting data and discussion to be had there (such as extrapolating bridge suicide nets with successful suicides before and after), and with around 20,000 annually this is a much bigger public safety issue than a few hundred spree shooting deaths, but it's also a discussion we'll simply never have at a national level.

bluedino2• 11 years ago

Look how low the murder rate is in the United States right now. It's half of what it was in the 70-90's.

http://wmbriggs.com/post/5969/

hga22• 11 years ago

And something fascinating about that is that per capita gun ownership has at minimum stayed steady during that period, when very roughly both increased by 100 million from 200 million.

Since we have a (very rough) control for that variable, it suggests we need to look elsewhere. One I like is detailed demographics, younger men tend to be the most violent, and the demographic Baby Boom may explain most of that.

ekianjo• 11 years ago

Excellent article. I wonder what software was used for the computer simulation in the second part. Can one do that in R relatively easily ?

emj• 11 years ago

If you use R it's easy, if you use python it's probably easier to that simulation there. It's more about knowing statistics, all of those conclusions in the article are not trivial to come to.

CodeCube2• 11 years ago

He mentions that it's a C#/WPF program that he wrote :) The article includes a link to download the code and play with it yourself: http://www.charlespetzold.com/blog/2015/07/RampageShootingSt...

nabla92• 11 years ago

Intresting note:

Nordic countries (Finland, Norway) and Swizerland rank very high in gun ownership among OECD countries. http://www.nationmaster.com/country-info/stats/Crime/Gun-own...

anc842• 11 years ago

They are sparsely populated and wildlife is dangerous. Might simply be that and/or historical reasons.

kwhitefoot2• 11 years ago

That's the silliest thing I have seen for days.

The wildlife in the Nordic area is very rarely dangerous. In the whole of Norway (60% more land area than the UK) there are a couple of dozen wolves and about 140 bears. The commonest large animal is elk and that is what the Norwegian hunting rifles are for.

You are more likely to be bitten by an adder than attacked by a large mammal and even then fatalities are rare.

As for the sparse population that is an example of averages being very misleading. Where there are any substantial numbers of people the population density is similar to the average for the UK while the mountains and high plains are almost not populated at all.

I write as a long term resident of Norway.

nabla92• 11 years ago

wildlife is dangerous

LOL no. We have lots of guns because we hunt.

tomwilde• 11 years ago

Switzerland is biased because of the militan army approach. Every able-bodied Swiss male has an assault rifle at home, without any bullets, however.

at-fates-hands2• 11 years ago

> Switzerland is biased because of the militan army approach. Every able-bodied Swiss male has an assault rifle at home, without any bullets, however.

This was exactly the purpose of the text in the second amendment when they talk about having a "well regulated militia":

The phrase “well-regulated” was in common use long before 1789, and remained so for a century thereafter. It referred to the property of something being in proper working order. Something that was well-regulated was calibrated correctly, functioning as expected. Establishing government oversight of the people’s arms was not only not the intent in using the phrase in the 2nd amendment, it was precisely to render the government powerless to do so that the founders wrote it.

source: http://bearingarms.com/well-regulated/

Essentially this meant every citizen should be properly armed should they be called into duty for defending the country. Sure, a lot has changed since then, but most gun owners I know still take this wording very seriously and believe it is their duty to defend against government tyranny.

ta929292• 11 years ago

You can buy ammunition in Switzerland in a gun store with nothing more than the equivalent of a an American "background check", i.e. no criminal convictions and/or mental health history.

at-fates-hands2• 11 years ago

Some more interesting facts:

Nations with stringent anti-gun laws generally have substantially higher murder rates than those that do not. The study found that the nine European nations with the lowest rates of gun ownership (5,000 or fewer guns per 100,000 population) have a combined murder rate three times higher than that of the nine nations with the highest rates of gun ownership (at least 15,000 guns per 100,000 population).

For example, Norway has the highest rate of gun ownership in Western Europe, yet possesses the lowest murder rate. In contrast, Holland’s murder rate is nearly the worst, despite having the lowest gun ownership rate in Western Europe. Sweden and Denmark are two more examples of nations with high murder rates but few guns.

and

the study also shows that Russia’s murder rate is four times higher than the U.S. and more than 20 times higher than Norway. This, in a country that practically eradicated private gun ownership over the course of decades of totalitarian rule and police state methods of suppression. Needless to say, very few Russian murders involve guns.

source: http://www.theacru.org/harvard_study_gun_control_is_counterp...

cronjobber• 11 years ago

The reason we care about mass shootings is because people die. Petzold drops the one variable we care about and calls it de-obfuscation? Neat.

OscarCunningham• 11 years ago

If you care about people dying then mass shootings aren't an important issue. (Compared to almost every other issue.)

Spooky232• 11 years ago

This brings up more questions than answers in my mind. Do we care about mass killings, or whatever is qualified as a "rampage shooting"?

For example, Mexico has 1 "rampage shooting" in the data referenced in the article. Yet you have cartels killing dozens of people on a regular basis. The question in my mind is that if we don't have lots of "rampage shootings" in countries with lots of access to weapons, why does the US have so many? Why aren't developing countries with ubiquitous AK-47s experiencing this phenomenon?

nraynaud• 11 years ago

careful, these are OECD countries, so there are no third world countries.

But your remark on Mexico rings true to me, maybe they are excluded because the victims are mostly not chosen at random.

lagadu3• 11 years ago

Well I'd say it's a fair assumption to make that organized crime and "sporadic" shootings have wildly differing causes, unless you're willing to generalize to the point of saying "they're both caused by unspecified socio-cultural issues".

Spooky232• 11 years ago

Personally, I'm more concerned about why folks in the US are going nuts in that way.

dsmithatx• 11 years ago

And if you included all the killings in Chicago during the same period, it would probably make most of Mexico looks like Disney Land.

dopamean• 11 years ago

Chicago is bad and the recent news of what happened over the weekend is terrible but saying that the situation in Chicago is somehow worse than the violence in Mexico is absurdly hyperbolic.

GVIrish8• 11 years ago

Sure you could find a city in Mexico that would look like Disneyland next to Chicago. But compare Chicago to Ciudad Juarez and the murder rates are about the same (457 in 2014 in Chicago, 424 in 2014 in Juarez). And a few years ago (2010) Ciudad Juarez was the murder capital of the world with over 3000 murders.

Mexico as a country still has about twice as many murders per year as the United States while having less than one-third the population. Mexico has FAR stricter gun laws btw.

exelius2• 11 years ago

Actually, you're not wrong. [1]

Chicago ranks higher than any city in Mexico save Acapulco (which is a short-term spike due to an especially violent cartel turf war raging there). That said, most of the worst of the cartel violence in Mexico takes place in rural areas, and wouldn't make this list.

[1] https://en.wikipedia.org/wiki/List_of_cities_by_murder_rate

GVIrish8• 11 years ago

I think in the case of Mexico there are plenty of mass shootings but they're due to cartel violence so they're not classified the same way.

As to why the US has so many mass shootings, I have to think there is cultural component to it. The idea of randomly killing a bunch of people has taken root amongst violence-prone mentally ill in this country. When you look at all of these shootings the vast majority involve someone who is struggling with mental illness and possibly suicidal feelings who tend to be loners. These people in their misanthropy look to take the lives of others in their last act in a blaze of violence.

For a different example of culture and violence, look at South Korea. Murder is rare, but South Korea leads the world in suicide rate (more than double the US). There, social and cultural pressures that lead people in pain to taking their own lives. It's not the gun laws, or drug laws that are causing that situation. But in that society the idea of taking your own life is more widespread for whatever reason.

As for developing countries with ubiquitous AK-47's, I can imagine most of the violence is attributed to civil unrest and state-sponsored violence so it's not tallied the same way.

hga22• 11 years ago

possibly suicidal feelings

Given that almost all of these people commit suicide before being apprehended, yep. It's rare that we have case like the Charleston church, and the Colorado movie theater, where we know the guy was seriously mentally ill, for his university psychiatrist was so concerned she broke doctor-patient confidentiality to report him, something you're only allowed to do if violence to self or others seems likely.

And, yes, "the authorities" dropped the ball, once he withdrew the security group of the university, which included that doctor, dropped it instead of referring it to local authorities. Something similar happened with the Virginia tech shooter; it's clear to me that if partially reversed the "reforms" in institutional commitments we'd have fewer of the mass shootings. Hmmm, another case like that, where the shooter just didn't get a chance to commit suicide, was of the Arizona congresswoman. From reports of his behavior at a local college, in times past he would have gotten serious attention, if not commitment.

If you support our current policies in both, you might say that freedom is enhanced by limiting commitments and maximizing gun ownership, but I suspect few but libertarians support both.

exelius2• 11 years ago

I would think that cartel violence is targeted enough not to qualify as a rampage shooting. Often they are perpetrated by larger groups of 10-15 armed men, and the victims are targeted because of their affiliation with another rival group. Yes, it's a mass killing, but it's a mass killing with a discrete purpose.

I would also think that a bit of the Republican hypothesis about guns holds up here: people don't perpetrate mass shootings where guns are ubiquitous because the risk of someone shooting back is much higher. Law enforcement in these places is often corrupt if they exist at all, so individuals take policing into their own hands by carrying weapons everywhere. But personally, I don't want to live in a country where I have to carry a gun to feel safe.

sudioStudio648• 11 years ago

Petzold is awesome. His blog has several articles he's written about his political views. I'm glad that someone with that kind of notoriety isn't afraid to write about politics.

kyberias• 11 years ago

I'm sorry, what notoriety?

CodeCube2• 11 years ago

He's known, amongst certain people in the industry - https://en.wikipedia.org/wiki/Charles_Petzold

geggam• 11 years ago

I would be interested in seeing the data correlating SSRI drugs to mass shootings statistically analysed so neatly

innguest2• 11 years ago

I've been panning these comments with Ctrl+F for "SSRI" and you're the first one.

You and I would be interested in seeing that data, and you and I will never see that data. ;)

Also the media never ever mentions most of those shooter kids were on those hardcore FDA drugs.

What's the point? We know the gun was responsible for the killings. /s

hga22• 11 years ago

Seriously, we do believe these shooting are correlated with just plain mental illness. Which are largely "treated" with "those hardcore FDA drugs" (which SSRIs, due to the low side effect profile, really aren't), so therefore I'm not surprised there seems to be a correlation between the two. However, after a fair degree of study, and discussion with one of my psychiatrists a dozen or so year ago, I don't think there's causation, just correlation. Or at least he didn't believe the "antidepressants give you enough energy to be violent" thesis based on all three generations of them.

grkvlt• 11 years ago

Isn't the fact that Norway appears near the top of this list only due to a single incident with a large number of victims, i.e. https://en.wikipedia.org/wiki/2011_Norway_attacks which skews the statistics. This is what he means when discussing small-N statistics being meaningless.

jamesfe• 11 years ago

You know what Twain says - there are lies, damned lies, and statistics.

(https://en.wikipedia.org/wiki/Lies,_damned_lies,_and_statist...)

phantarch• 11 years ago

Sherlock (or Doyle) says it best. "It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts."

Petzold does a really good job of debunking the original claim that the US is not statistically significant when it comes to the number of mass shootings. However, all of the effort he goes through only points to one conclusion: the number of shootings we have in the US isn't statistically random. All this means is that there's something causing the rate of shootings to be higher here than is expected by random influences. It doesn't point to what that cause is.

It's odd to see how statistics are abused by interest groups/media. Intuition is usually a pretty good yardstick for the validity of a statistic, insofar as knowing what kinds of conclusions should be able to be drawn from what kinds of data. If you've got a bunch of data about who likes what flavor of ice cream, it's not likely that data says anything about /why/ each person likes what they like. In the same way, it doesn't matter if you say that the US is normal when it comes to number of shootings or to say that it's abnormal as Petzold does, when you try to say /why/ this is the case, you better be drawing direct conclusions from your data.

avn2109• 11 years ago

Speaking of validity, it's not immediately obvious to me that each citizen of an OECD country constitutes a valid Bernoulli trial. Applying a uniform empirical prior of shooting rampage probability across the entire population strikes me as questionable. Presumably the results would be roughly convergent if he had done this with an inhomogeneous Poisson or something, but it seems a little handwavy to just fire off a binomial and call it good.

adamtj• 11 years ago

Intentionally or not, the author has cherry-picked the data in a way that completely undermines the article. The conclusion that "lives are at stake" is obviously true and supported by the data, but the implication that the US has too many guns is not. An analysis is only as accurate as the data. Here, the missing data makes it painfully obvious that, while we should whatever we can to reduce gun violence, we cannot morally accomplish it by disarming US citizens.

If you merely look back to about 1939, you would find numbers for Poland alone that dwarf any statistics for the US, even ignoring differences in population! Poland isn't the only such example. Around that time, certain groups of people suffered from numerous horrific shootings, decimating their populations, despite those groups having been recently disarmed almost entirely. Of course, I'm being sarcastic. These people suffered at the hands of their own governments because they were disarmed, not despite it. There are some still alive who remember.

An honest assesment of history, even very recent history, makes it clear that America's current level of gun violence is a mere rounding error next to the real threat: descent into tyranny. We're lucky that we live in such a relatively peaceful time, but let us not delude ourselves. No human will ever live without the threat of tyranny. No government of the people or by the people will ever, by it's own internal guidance, remain civilized indefinitely. We must never surrender the tools needed to provide the most drastic of course corrections for a government gone astray.

We are fortunate that by keeping and maintaining arms, we make it more likely that we will never have to use them. Armed violence is always the absolute last resort, but if we give up our last resort, we give up all the other ones too. Speech, for example, is of little use against a tyrant with absolute power. But speech from the mouths of well-practiced riflemen, with their weapons close at hand? Those words cannot be safely ignored by anyone.

It is for this reason that the right to keep and bear arms is not unique to just the United States. Rather, it is a fundamental and irrevokable human right. The Constitution of the United States does not and cannot grant this right -- though it does explicitly recognize it, lest we forget.

Yes, gun violence is awful, and we should work hard to reduce it. But let us not forget that is is the price we pay for a civilized society, and it is a small price compared to the alternatives. I grow tired of articles that completely miss this point.

smoyer• 11 years ago

I'm to the right of president Obama, own guns (yes plural) and yet agree with him whole-heartedly. We, as a nation, need to figure out how to stop this violence. Please don't paint all conservatives with the radical right-wing brush.

As an aside, I don't think Finland is celebrating in the streets with shouts of "we're number one".

awptimus• 11 years ago

This article demonstrates the danger of thinking statistics is logic - you can only use statistics with the analytical framework firmly set up...not as a replacement.

The United States is just that - United. States. 38 mass shootings, however defined, means there is /less than one per State/. Guess what? That makes Finland's two /much/ higher. That's...statistically significant!

Explaining how completely incapable frequentist statistics are at describing anything at a low sample size doesn't mean that something succumbs to "random fluctuation." There is nothing random about a person deciding to take arms and shoot lots of people. Not only that, this data is based on a full population sample - it's not a survey of 5,000 people extrapolating to the whole population - it IS the full population.

And he removed the idea that the original chart was designed show - lack of data that shows compelling evidence that more restrictive gun laws prevent either mass shooting incidents or mass shooting deaths.

He should step away from the statistics - it's a loaded gun and he's not trained to use the weapon properly.

De-Obfuscating the Statistics of Mass Shootings

Related Stories