Hacker News Reader

Comment by sandworm101 | Hacker News Reader

30 million people. How many suicides would be normal on a given day anyway? A few have probably dropped dead of heart attacks and strokes too.

View on HN

Globally, 16 suicides/100000 people/year. So the 30 million people should have 4800 suicides in a year, or 13 per day. But these two deaths are restricted to Canada which has a lower suicide rate than the global average, and we don't know what portion of the 30 million are Canadian, so we don't have enough data to make a guess.

Probably a lot less since whole Canada is inhabited by only 35 million.

But it was a Canadian based company so there may be quite a few Canadians in there. Also see: http://www.cbc.ca/news/business/ashley-madison-says-1-in-5-o...

http://www.huffingtonpost.ca/2015/08/19/ashley-madison-hack-... mentions an estimate of 273,000.

That's about 1% of 30 million, or about 0.1/day, or 3 weeks for those 2 suicides. I bet any deviation from the average isn't statistically significant (in fact, for 2 suicides, ¿almost? no 'background' suicide level will be statistically significant)

On the other hand, I do expect that Canadian police didn't base their claim on nothing. Chances are they found suicide notes, or something similar.

”didn’t base their claim on nothing”

Actually, they didn’t claim anything, and only talked about unconfirmed reports (https://twitter.com/TorontoPolice/status/635816932474814464, http://www.cbc.ca/news/canada/toronto/ashley-madison-hack-2-...)

The current suicide rate in the US is about 12 per 100,000 per year, so that translates to about 10 per day among a population of 30 million. Assuming, of course, that the risk factors among that population match up with that of the entire US population, which they probably don't.

They are probably worse. AM users were predominantly male, which I believe increases the likelihood of suicide.

Yes, the suicide rate for men in the US is over 20 per 100,000, while for women it's about 5.5.

Is that due to the much higher success(?) rate among males vs females? What are the "attempts" per 100,000?

Statistically, females are more likely to self-harm in non-lethal ways. This acts as suicide prevention, in a way, since it triggers responses from the environment and reduces the likelihood of worst outcomes.

Males don't self-harm and overall don't signal their discomfort until it's too late.

According to this web site (accuracy unknown), women attempt suicide at a rate 3x that of men:

http://www.suicide.org/suicide-statistics.html

This PDF says that there are 25 attempts for every suicide:

http://www.cdc.gov/ViolencePrevention/pdf/Suicide_DataSheet-...

I had no idea it was that high!

In England (or maybe the UK) the rate of death by suicide for men is 17.2 per 100,000 population. For women it's 4.6 per 100,000 population.

(This is from office for national statistics data)

We tell little boys that "Big boys don't cry". Then wonder why so many men in their 20s kill themselves instead of seeking professional help. :(

There is a huge stigma around seeking professional help. It can and does cost people there job and sometimes career.

Not to mention it tends to be rather expensive and of limited value.

I think it has more to do with the method than the act. Men are more likely to have guns, which work very well. Women are less likely to have access to guns and so survive more suicide attempts (pills, jumping off things etc).

Same for the military. The 'success rate' for military suicides is higher partially because of ready access to the means.

Which does nothing to explain the similar ratio in say England where access to guns is very restricted.

"Jumping off things" would be "falls and fracture", and it's not a common method for completed suicide. About 4% of registered deaths are "jumping off things". Drowning is a bit more, about 4%. Poisoning is about 20% and hanging, strangulation and suffocation is about 56%. (The rest has the unhelpful title of "other". I'm not sure what that means.)

> Which does nothing to explain the similar ratio in say England where access to guns is very restricted.

Drive a car fast. Just a lethal. Sometimes you hear about car accidents with only one car involved and the young male driver was the only fatality. Sometimes it's an accident. Sometimes it's someone ending it all.

Help seeking is definitely part of the problem. UK charity "The Campaign Against Living Miserably"[1] and Australian charity(?) "Soften The Fck Up" both campaign to make it easier for men to talk about mental health and to seek help. UK charity "Time to Change" has a number of campaigns to destigmatise mental health, some of which are targeted at men.

But in the UK there has been a rise in rate of suicide amoung people already known to MH services, so we need to be a bit careful not to put all the "blame" on help seeking behaviours.

(I'm interested in other anti-stigma organisations, where ever they're based and whatever demographic they target.)

[1] Calmzone is currently banned from Facebook. I don't know why. https://www.thecalmzone.net/

http://softenthefckup.com.au/ -- they don't censor the word on the website.

http://www.time-to-change.org.uk/

I think the problem of testing for randomness of these suicides is harder to solve than it seems. Randomness only occurs in the absence of suicide letters which state that the leak caused them to do it. At least two scenarios:

1. They leave suicide notes where they explicitly attribute it to the hack. Then, the probability that it is due to the Ashley Madison hack becomes = 1. The question then becomes, did the 'outing' push them over the edge. Especially so, if they attribute taking their life to multiple reasons.

2. If it is not attributed to the 'outing', then one has to carefully see what the probabilities of suicide are in their particular cross-section of population and also consider other risk factors. The numbers someone else quoted here, 16/100,000 is an average. It is fallacious to assume that this number uniformly applicable to the population as a whole. You have to consider prior probabilities for each individual. These priors can possibly be computed based on the individual's medical history, socio-economic background, geographical location, etc. with importance assigned in that order. I would imagine medical conditions would play a big role. For example, if an 'outed' person with no history of depression, etc. from a hyper-religious family commits suicide, it is quite likely that it was because of the 'outing' in the absence of explicit attribution in a suicide letter, etc. If a person has on his/her history previous suicide attempts, depression, etc. the marginal probability that it was solely due to the Ashley Madison outing would probably be lower.

Yeah https://en.wikipedia.org/wiki/Correlation_does_not_imply_cau...

Yeah, except in this case I believe this is a number released by the police and is likely based on evidence learned at each suicide. Notes left behind or information from people close to the now deceased, it's safe to say these two deaths have been attributed to the impact the leak had on their lives.

Also keep in mind we're talking about 30mil people in a particular population group, age range, that probably lives a stressful life (and possibly guilt)

Reply on news.ycombinator.com