Hacker News Reader

Comment by sopooneo | Hacker News Reader

How do we know video games aren't partly to blame? To me it seems unlikely they are to blame, but I have no proof. I don't play or create games so I don't have any direct stake in this discussion.

Most arguments I hear from gamers use plausible, and even persuasive, theories on why games wouldn't contribute to mass shootings, but I don't recall hearing any empirical evidence backing them up. And the older I get, the more I find that plausible theories and reality don't match up often enough to be fully convinced.

View on HN

I'm sick of this "debate".

We can talk about more gun laws, less gun laws, less video game violence, a psychological approach, a strong-family-as-a-deterrent... or anything else we can imagine. But those solutions will never stop these crimes. We can only attempt to control this issue on the margins, and hope that we can find practical solutions that try to curb these random acts of violence. But the brutal truth is that they are a part of the nature of our species, inherent and practically unstoppable. Sometimes I'm shocked they don't happen more in a world of around 7 trillion human beings.

So this debate disgusts and annoys me to no end. How about we keep this national conversation geared towards respecting the memory of those who have passed, instead of as a vehicle to progress our personal politics. Because we're not solving this issue and we're looking at it all wrong.

Changing the way we approach video games will have no appreciable affect on these tragedies. People who are so mentally bent as to carry out heinous acts of violence against other humans will always find a way.

I don't believe in a stark concept of "Good vs Evil" in a biblical sense. However, as much as people do not like to admit it, as much as everyone wants to think that a few hours of psychotherapy and some modern pills will turn any lost soul around, there will always be those people who are outliers and are simply "broken".

> "But the brutal truth is that they are a part of the nature of our species, inherent and practically unstoppable. "

We may never stop all violence, but this debate is worth having specifically because other societies of our species (i.e., not American) have been able to achieve much lower rates of violence in their societies than we have.

Not everywhere is as violent as the USA. These mass shootings are statistical outliers no matter which way you cut it, but violent crime (involving firearms or otherwise) in the US is substantially higher than many developed nations.

The question remains as to how, and why, these other nations have been able to achieve a dramatically lower incident rate of violence.

BTW, there are ~7 billion people on Earth, not trillion.

> "People who are so mentally bent as to carry out heinous acts of violence against other humans will always find a way."

Violent crime is not in and of itself indicative of mental illness. There are many perfectly rational (at least at the local level) reasons to hurt or kill someone.

You are right to be suspicious of plausible sounding hypotheses that aren't backed by empirical evidence, but you are failing to correctly account for prior probabilities. The prior probability that any particular phenomenon is causally linked to a given phenomenon is very low (and as a corollary, the probability that one phenomenon is not linked to another is very high).

Unfortunately, humans are pretty bad at handling Bayesian inference implicitly, so we have to use some tricks to keep ourselves in line. A particular problem that we face is that once we start investigating a hypothesis, we seem to upgrade our prior for that hypothesis, and so when we find weak evidence, we become overconfident in the hypothesis. This is related to the prosecutor's fallacy[1], and is also known in rationality circles as "privileging the hypothesis"[2].

The other bad thing about privileging the hypothesis is that we only have so much time to investigate hypotheses, and we have a ton of bad hypotheses. In fact, because of Occam's razor, we actually have way, way more bad hypotheses than we have decent ones.

So investigating hypotheses with low priors has a low ROI

So in summary, we don't really need to (and shouldn't) gather evidence to see if video games can be linked to mass shootings because the hypothesis "video games aren't linked to mass shootings" has such a massive head start over "video games are linked to mass shootings". If, however, simply surveying the data available at the outset revealed significant reason to believe that video games might be linked, it would be different. That's why it is worth investigating the link to guns, even though the prior for that connection was also low.

[1] http://en.wikipedia.org/wiki/Prosecutor%27s_fallacy

[2] http://lesswrong.com/lw/19m/privileging_the_hypothesis/

I've seen data at some point that other countries such as Canada and European countries have lower gun violence rates than in the US. These countries have similar video games, movies, and other media. It's not really enough to make any strong claims, but does seem to point to video games and other media not being the main issue.

Not to mention Japan. And I hear they play games there.

There's quite a bit of empirical evidence available that have repeatedly demonstrated a lack of even correlation between playing violent videos games and being violent. It's all very easily available by praying to google, so I won't link them here.

The thing I really want to say is that this is all really obvious when you stop and think about it. If there was an actual causal link between playing video games and becoming violent, we would expect much more violence than we actually see. For some reason, no one seems to look at it that way, though.

Innocent until proven guilty? I'm not sure if there is evidence linking games to violence but generally I suppose the industry feels discrimination is at work since they're a relatively new phenomenon. Other industries such as violent tv or movies, water gun manufacturers or comic book companies aren't required to justify how their products do not contribute to mass shootings. I agree that perhaps all facets of our society should be investigated to see the relative benefit and harm but singling out one industry seems like playing the blame game.

I strongly agree about the need for good quality research. This need is strongest when we have plausible sounding reasons.

There is an interesting disconnect in public discussion about influence. Many people say that games and movies do not make people more violent, or that they only have a very small influence.

And many people say that "Thinspo"[1] or advertising is evil and should be banned because we're creating a generation of women (and also men) with body dysmorphia.

It feels odd to me that people can be so sure that games have no influence yet images of thin women in magazines has such a strong influence.

[1] thinspo style images are banned under quite a few company terms of service.

>It feels odd to me that people can be so sure that games have no influence yet images of thin women in magazines has such a strong influence.

There is clearly an invitation to mimic the presented behaviour in both cases, however whether or not people follow up on this is another issue - mainly shaped by social norms. Being thin is socially acceptable and encouraged(whether or not it is a good thing is irrelevant). Killing people - isn't.

Also the action presented has to be related to the context it is presented in. An ad shows thin people in everyday life - so kids try to imitate by starving themselves in everyday life. Wargames present murder in context of a military conflict - well since we are NOT in the middle of a military conflict, there is no point imitating.

Let's say 1 million people bought the last Black Ops game, and that our Connecticut shooter was such a player. So far, 3 months in, we've had 1 mass shooting. Correlation ~ 1/1000000 = 0.00001% at 3 months in. Let's assume we're off by a factor of 100 in our violent video-game audience. That would require 100x the number of shootings to even match parity.

Black Ops 2 has sold about 22m copies to date.

I'm swagging a factorization of a reasonable pool of "violent" video game players. In any case, if you just picked Black Ops II, for there to be a 1% correlation, there'd have to be 220 thousand mass shootings from players (without any from non players) to get such a correlation. Of course, even then, we'd probably assume some other co-factor, since the ratio of players to shooters is so low.

Along those lines you couldn't possible dismiss the wild theory that it was actually toast or breakfast cereal that made them do it.

Exactly, there is this joke in german that goes like this: "90% of violent acts are committed within 24 hours of the consumption of bread. Ban bread!" This kind of "logic" is just ridiculous.

Now if it were: "after consuming bread, 90% of the time the consumer will commit a violent act within 24 hours", we'd have to worry.

It is not "after playing a violent video game, 90% of the time the consumer will commit a violent act within 24 hours" yet people act like that and want to have video games banned...

100% of all school shootings took place at school. Clearly school is the problem, right? :)

No, you could say it's very unlikely that toast or breakfast cereal is to blame, but that we do not have any research to say either way. And until someone comes up with a credible link between breakfasting and violence we won't have money to do any research.

Exactly. sopooneo's claim is just as ridiculous, unfounded and manipulative because it puts the blame on the video game industry instead of applying "in dubio PRO reo".

Reply on news.ycombinator.com