Hacker News new | past | comments | ask | show | jobs | submit login

Sorry if this is me being a "statistical grammar nazi," but he started his post with it... and it's not like misspelling a famous person's name, it's spreading innumeracy.

"There are approximately 7 billion people in the world. The average life expectancy is somewhere around 67.5 years. So in the next 67.5 years 7 billion people will die"

He's confusing average with actual.

Here's a simple example: imagine we have two people. 1 person who will die in 1 year, and 1 person who will die in 9. The average here is 5 years, and after 5 years only 1 person will be dead, not 2.

While It's not always true that selecting only individuals who have a value less than or equal to the average gets you half the population, this correct for normally distributed value when there are lots of samples, so the authors numbers are basically off by a factor of two.... which he almost notices here:

"So in the next 67.5 years 7 billion people will die. 67.5 years * 365 days is about 25,000 days. That’s about 280,000 deaths per day average. Right now that number is lower at about 155,000, but as the population ages that will catch up with us."

</rant>




Everybody that will be born between now and 67.5 years from now will have exactly that same life expectancy, half of those will die as well.


If everybody was 0 years old now, and the life expectancy were 67 years I guess you'd have the point?


No. Even if everybody was 70 years old I'd still have a point.


The life expectancy of 80 means that out of 100 people born 50 will die before they're 80 years old and 50 will die after their 80th birthday.

But: zero of these 100 people live more than 103 years (if we started with 1000 only 3 would live to be 103 years old):

http://www.ssa.gov/oact/STATS/table4c6.html

That means that everybody who is already living now will die in the next 103 years.

If at the moment 7 billion people live on earth practically all will die in the next 103 years. That their individual life expectancy is 80 years doesn't change the fact that practically nobody (that is, to the error of 0.5%) survives his 103rd year.

I started with the US data. If we'd adjust it for the world, we'd see that even for life expectancy of 67 less than for example 10% of those living now will survive their 80th birthday or something like that. So the rougher approximation is that you have to increase the "67 years" some years more but anyway in 100 years practically everybody living now is dead. Yes he was not technically right with using the life expectancy of a single person to give the exact year, but the "cutoff" year is still not much farther away. And how old are people now actually matters to know how many will die in which period.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: