View Single Post
  #50   Report Post  
Posted to uk.d-i-y,
NY[_2_] NY[_2_] is offline
external usenet poster
Posts: 1,062
Default OT: Latering thinking puzzle "Why do more peoplre die on their bithday than any other day?"

"Rod Speed" wrote in message

It will be interesting to see how many others never get it.

As far as I remember of the postings so far to this thread, you and Mike
Humphries are the only ones who have asserted that it is true. Mike has
given a fairly rigorous explanation of probability for each successive day
after today; you have said "the variation in the birth rate over the year is
all that is needed to explain it", without explaining *why* a varying birth
rate means that there is a spike in the probability of dying exactly n
calendar years after the date when a person was born.

I can see that more people will be born on some days than others. So if you
took a sample of people, you'd expect to find more people with some
birthdays than others. But even assuming a constant death rate (ie not
varying seasonally), I don't see how that explains why person A, born at a
time of high birth rate, is more likely to die on their birthday, whereas
person B, born at a time of low birth rate, is more likely to die on their
birthday. What is so special about a period of exactly one calendar year
that makes a person more likely to die then than any other day.

What is interesting is that this statistical analysis of probability of
dying on any day, and the sudden increase on the person's birthday, isn't
mentioned in the Wikipedia article, which constrains itself to variations
due to factors that are (loosely!) within the control of the person:
accidents while celebrating, suicide as people think "I'm a year older - is
life still worth living", terminally ill people striving to stay alive until
their next birthday (or the wedding of a family member or any other
significant date).

I think I'm just going to have to accept that there *is* a reason for it
which I don't *really* understand - either because you have given an
"obviously" jump of logic from "more people are born on some days than
others" to "therefore a person is most likely to die on their birthday", or
because Mike's explanation involves summing lots of infinitesimally small
probabilities ("if I'm alive today, what's the chance I'll be alive
tomorrow; if I'm alive tomorrow, what's the chance I'll be alive the next
day" ad infinitum).