A graph for the gun violence / video games conflaters

Via The Atlantic Wire, this graph shows a clear correlation between gun violence and video game spending per capita…

26ae3fa8e130bd1a591e5bad36638189_623x330

…oh wait, no, it doesn’t.

Like, at all.

Your move, gun apologists.

{advertisement}
A graph for the gun violence / video games conflaters
{advertisement}

35 thoughts on “A graph for the gun violence / video games conflaters

  1. 2

    and South Korea has 0 gun violence, while being #2 in videogame spending and probably #1 in videogame culture (they make lots of them, and it’s Serious Business when they play tournaments, like Golf in the US).

  2. 3

    It sort of does show a correlation, just that more games imply less murder

    Actually it kinda looks like the Netherlands is a fat guy jumping on the end of the seesaw that America was sitting on and now its being thrown through the air in a comical fashion

  3. 4

    Germany,
    Japan,
    Australia,
    the United Kingdom,
    Canada,
    France,
    and of course South Korea and the Nederlands spend more per capita than the US, per the graph. Don’t just look at Canadians.

    In fact only China spends less. Probably because they are too busy being taught how to hack into US Federal Agencies and Fortune 500 companies.

  4. 6

    While we clearly have a great lead in the gun murder part of the graph, we’re in the middle of the pack on video game spending. There’s a video game gap!

    Clearly, the appropriate move for gun apologists is to use the resources of the NRA to buy more video games, pushing our little blue dot far to the right and creating a correlation. Then every time people complain about guns, they can pull a switcheroo and say “Look, it’s video games!” and then….VICTORY.

    And wait…Canadians spend more on video games than we do? And just what are they doing in the Netherlands?

  5. 7

    At first I was thinking “why is all the data down at the bottom of the graph?”, then I saw the USA dot.

    To be fair if a South Korean did want to go berserk in the mall, he’d have a hard time finding a gun to do it with. So, hey, maybe that works.

  6. 8

    I don’t understand something: where is the US? Is that the chart with the US factored out?

    If this is a comparative statistic with the US omitted, then a label like “US omitted for *” would help. Even better is a side by side comparison (i.e: “global gun violence per capita by video game expenditures, US excluded versus global gun violence per capita by video game expenditures, US included” which would serve to illustrate the disparity observed in the US, or whatever.

    (* “because it’s monday” is fine with me! But put it in your chart)

  7. 9

    Look top middle, Marcus. Right under the “m” in “murder”.

    Also, what’s the r^2 on that? It looks like it sucks. Technically we’re showing negative correlation, but so weakly that no one should care.

    Chalk up another for the Null Hypothesis.

  8. 10

    Hate to say it, but there isn’t really much correlation at all. This is kind of just showing that curve fits will give you a result whether or not the data really shows correlation. If you removed the US you would get something about constant.

  9. 12

    If they needed to make one axis six times bigger to include the USA in the graph, you can safely omit the “near”.

    And no, Aliasalpha, it doesn’t show any sensible correlation, positive or negative. It shows just a swarm of points one of which is so off it prevents any good analysis. If an equation was written by the regression line, including interval of reliability of the parametres, you’d see that. An honest conclusion of this is “no dependance observable” (unless you want to claim there’s a strong resonance peak).

  10. 13

    I don’t see how this chart shows the thing it claims to show(a lack of correlation between Gun violence and video game consumption). To use a less loaded and entirely imaginary example to see why, what if one made a chart to test weather video games cause Lamborghini violence(that’s violence caused with Lamborghini’s, not violence towards them). It would be likely that countries with absurdly high ownership of Lamborghini’s(say Italy, though I’m only guessing) compared to the other countries on the list would have far and away the highest Lamborghini related violence rate. This would obviously not prove that there is no correlation between video game consumption and Lamborghini violence, only that Italians(again, a guess)have so many more lamborghini’s that the other countries Lamborghini violence doesn’t register.
    |
    A better chart would either correct for Lamborghini ownership rates, or switch to a more general metric like automobile violence in general vs. Video game consumption. It seems quite odd not to do one of those things, or at least include such a chart in addition to the uncorrected one.
    |
    To be clear, I think the United states has way too many idiots with guns, and I am damn near positive that video games don’t cause violence. I Just don’t think this chart shows that. it only shows that the United States has WAY more guns than the other countries in the top 10 video game consumption set, and if you want to chart that, the video game axis is entirely superfluous.

  11. 14

    Hate to say it, but there isn’t really much correlation at all. This is kind of just showing that curve fits will give you a result whether or not the data really shows correlation. If you removed the US you would get something about constant.

    This. The US is clearly so much of an outlier that it should have been removed before drawing the trend line; it’s distorting the analysis. I don’t think this graph should be used for the trend line. It’s just going to be, quite legitimately, ripped to shreds.

    What it should be used for is to show how far outside the norm the US really is with regard to gun violence and how that cannot be accounted for by video game consumption. It makes that point very nicely.

  12. 16

    The UK is a bit more stabby when it comes to murders, and there has definitely been various uproars over video game violence here. I’m wondering what a murder, as opposed to gun murder, comparison would show?

    Games that have outright violence with humans (not spaceships, or non-human whatevers that use weird stuff that has no Earth counterpart – think Flipull, Qarth, Balloon Fight) use either guns, or swords, or magic. And the pvp-focused games tend to use more guns (because range). The pve-focused games tend to favor monsters and non-human enemies, except for the final bosses and the likes (though they like to veer off into Eldricht Abomination territory at that point) and The Empire that has soldiers and such.

    I’ve played tons of violent videogames since I was 3. My brothers also played tons of videogames since they were just as young. I’m 30, they’re 28, 22, 19. None of us have killed anyone, own a gun, own a knife, or have tried using them against people. One of my brothers likes brawls, but he did before and independantly of videogames – and he doesn’t “finish off” his enemies, even if he plaued Mortal Kombat.

  13. 17

    Also, what’s the r^2 on that? It looks like it sucks. Technically we’re showing negative correlation, but so weakly that no one should care.

    Chalk up another for the Null Hypothesis.

    That’s my opinion too, on cursory examination. The math holds the answer, and I lack both the raw data and the statistical expertise to do it… but eyeballing, this sure looks like “no correlation”.

    FWIW, there’s not a very good correlation between gun violence and how strict gun control laws either. America’s outlandish gun problem has myriad causes, there’s no simple answer. I support stricter gun control nevertheless, but even if we instituted the strictest gun control policy in the world tomorrow, in a decade we’d most likely still be the weird outlier when it comes to per capita gun violence.

  14. 18

    LykeX says it better than I did:

    What it should be used for is to show how far outside the norm the US really is with regard to gun violence and how that cannot be accounted for by video game consumption. It makes that point very nicely.

    Yep. Do violent video games have an effect on gun violence? The data seems to suggest not, but “hard to prove a negative” and all that…. Is America’s fucked-up gun problem explained by video games? Not. Even. Close.

  15. 19

    Jesus Christ, what the fuck is wrong with you Americans?

    I’m thinking it’s a combination of crappy low-alcohol, no-flavour beer and lack of fighting in your favorite national sports.

    Let’s see how long it takes Jason to run out and shoot somebody now that he’s a Yank.

  16. 21

    You also believe in millions of years and don’t have your morality grounded in God’s Word, so it’s really just a matter of time until you decide that it’s okay to murder all your fellow chemical meat bags with a Bushmaster rifle.

    Of course, note that I’m not saying you shouldn’t be able to purchase a personal armory that rivals that of a 3rd world despot, thousands of rounds of ammunition, and a Marauder outfitted with a pair of PPCs and a AC/5 cannon (although I prefer the Clan variant with two large pulse lasers instead of PPCs). It’s definitely video games like Mariokart that will push you over the edge. Ban Yoshi!

  17. 22

    in Demolition Man, criminals in Los Angeles own anti-air weaponry. This is where you’re headed with the 2nd amendment “that means I can own missiles” interpretation, when it originally meant you could form a militia (and be armed in that context).

  18. 23

    It nearly caught me out until I saw the US lurking there at the top.

    Based on a sample of one, I have never bought a video game, never bought a gun, and never murdered anyone.

    Surely the missing information is not the correlation (or lack thereof) between video game spending and gun-related murders but the correlation between: access to video games AND guns, and gun-related murders.

    Much more significant would be to determine whether those committing gun-related murders play violent video games more than an innocent control group. But how do you choose the control group?

    In the UK one would probably get a positive correlation between gun-related murders and the consumption of porridge.

    Hint:

    por‧ridge [uncountable]

    1) oats that are cooked with milk or water and served hot for breakfast

    2) British English informal a period of time spent in prison

  19. 24

    According to the graph it appears to be zero gun related murders in China, UK and South Korea. I looked at wikipedia, and they say there were 0.04 gun related homicides pr. 100.000 in the UK and South Korea. Maybe they have used numbers for different periods? I also think that it would be interesting to see what the graph would look like if more than 10 countries were included.

  20. 25

    @LykeX, #14

    The US is clearly so much of an outlier that it should have been removed before drawing the trend line; it’s distorting the analysis. I don’t think this graph should be used for the trend line. It’s just going to be, quite legitimately, ripped to shreds.

    What it should be used for is to show how far outside the norm the US really is with regard to gun violence and how that cannot be accounted for by video game consumption. It makes that point very nicely.

    That is the point being made here: Gun violence is pretty much independent of the amount spent on video games, except in the USA.

    Ignore the slope of the regression line (which is clearly distorted by the outlying values), and just pay attention to the distance from the line to the most egregious outlying value — the one that ordinarily would makes you suspect a measurement error, as though one respondent had recorded their weight in stones instead of kilos.

  21. RH
    26

    While I am all for trying to correct myths, including gun myths, that influence our policy….

    From a serious scientific analysis point of view this graphic really doesn’t show anything meaningful.
    How was spending per capita adjusted for local costs?
    Was piracy taken into account?
    Why pick these countries?
    Are the gun death rates reliable (China I am looking at you)?
    Are there substantial factors not taken into account such as availability of guns?
    What does your trend line look like when you remove the outliers? Say top and bottom gun violence rates?

    Video games don’t cause gun crime.
    This graph doesn’t demonstrate that… or anything else as far as I can tell.

  22. 27

    I feel silly. I was sitting there wondering, “Where is the USA? Surely it has to be there. It has to be above Canada because we have more gun deaths, so where the flip is it?”

    I was focusing too much on the line and as such wasn’t looking anywhere near high enough.

  23. 28

    Apparently Not Erin (#6):

    Obviously Canada just has more geek per capita than the US. Or maybe it has something to do with the higher price per game regardless of where the dollar sits.

    Just as likely, it has to do with crappy winters and nothing to do. I have no link to cite and prove it, but my old town in BC once had the highest per capita rate of movie rentals in the country. It also had Januaries with -40C and a metre of snow.

    Cabin fever is an actual mental condition, sometimes linked to violence and murder. One would actually expect a higher rate of such violence in Canada than the US because of the weather. If video games did have an actual effect, it’s more likely as a catharsis that prevents violene rather than a trigger that causes it.

  24. 29

    It also had Januaries with -40C and a metre of snow.

    Here I thought I was the only person reading FtB who grew up in shithole Mackenzie.

    There were only two seasons there: winter and mosquitos, both of which made staying indoors to watch movies or play video games preferable to running around shooting fellow Canadians (even if your victims are Canucks fans).

  25. 30

    Are there substantial factors not taken into account such as availability of guns?

    Availability of gun, ease of getting them and a culture that celebrates gun ownership as if it was like owning a car surely heighten the risk of gun violence, and because guns kill way easier (you can die from a single shot, and it doesn’t require strength, and given the shooter feels they are at near-zero risk of retaliation while wielding the gun, it promotes the use of them by people who would otherwise not have the courage).

    People who might consider theft or grand theft, but have no guns available, have to plan way more about it, and lots of them probably drop the entire idea because it’s harder. They might be just as motivated as a US grand thief (poverty can be way worse in China, for example), but the threshold where their plan starts making sense is higher because of higher difficulty in accomplishing the deed.

    If everyone and their mom has a gun, using one sounds mundane.

  26. 32

    The US have also the largest prison population in the world per capita (by far) and a high murder rate for a first world country (12 times higher than Japan and 3 times higher than Canada).

  27. 33

    sinned34 (#29)

    Here I thought I was the only person reading FtB who grew up in shithole Mackenzie.

    There were only two seasons there: winter and mosquitos,

    I lived in the interior in the Rockies, not on the plains. But I once lived northeast of the Rockies for six months and it was horrible. It was either a hundred mosquitos per cubic metre in summer, or intolerably cold and windy in winter (and less than five hours of daylight). Many people made second homes of the hockey rink, curling rink, or The Northern, just to have somewhere to walk around inside for hours without a jacket and winter pants. The “northern living allowance” doesn’t make up for how bad it is – no amount of money could.

    Nowadays, though, I live and work in Taiwan and have for seven years. I haven’t seen snow in eight years, and the coldest day here is about 5C. I say coldest day because it is only one day per year in January. Februaries here are 20C for several weeks. Not making you jealous, am I?

    both of which made staying indoors to watch movies or play video games preferable to running around shooting fellow Canadians (even if your victims are Canucks fans).

    That’s quite alright. Just don’t say that about BC Lions fans.

  28. 35

    Your stories of unlivable areas make me miss the Christmas white-out conditions and real springs that came from living in Nova Scotia. Granted, I hate winter and they’ve had more storms this year than our relatively mild winter here so I’m not inclined to move back just yet.

Comments are closed.