I don't think that's the right way to look at the graph. It says that out of 100k 15-24 yr olds 74 died, and also out of 100k 85-1xx yr olds 13k died. This is purely a measure of death rates and doesn't take the age distribution into account.

to get the probability of dying before 50 you have to look at the total number of deaths/pear, distributed by age: https://www.statcan.gc.ca/pub/91-209-x/2013001/article/11867...

the area under the curve for 0->50 looks about 1/10 of the overall, so I'd say the op is about right.

>It says that out of 100k 15-24 yr olds 74 died, and also out of 100k 85-1xx yr olds 13k died.

Yes, which is why I took the sum of those numbers for ages 25 to 50.

>the area under the curve for 0->50 looks about 1/10 of the overall, so I'd say the op is about right

You're describing the conditional probability that, given you have died, you are under the age of 50.

what you're doing is (129+192+405)/100000 = 0.7%

this means that out of 100k 25-50 yr olds 726 died during 2016. This is the probability that you will die during this year, given that you are a 25-50 year old, which is different from the probability that you will die before the age of 50. (the prior number ignores the cumulative probability that you have died as a <50 yr old during all years prior to 2016)

Given that all of us will die, the only question is the age at which we die. The conditional probability that, given you have died, you are under the age of 50 is exactly what we should be looking at.

>the prior number ignores the cumulative probability that you have died as a <50 yr old during all years prior to 2016

Ah you're right about that. Looks like I'm probably way off. This random thing google turned up [0] says that a 25 y/o male has like a ~70% chance to make it to 50!

>Given that all of us will die

Something doesn't feel right about this, but I can't figure out what. I'll let it simmer and assume you're correct in the meantime.

[0] - http://flowingdata.com/2015/09/23/years-you-have-left-to-liv...

yeah, the probability of death depends on a ton of priors so assuming the distribution is uniform is not quite right either.

just intuitively though, a probability of .7% for dying before 50 seems way too low