(urth) Memento Mori, Re: This Week in Google Alerts: conference, sketch

[Gwern Branwen]

On Mon, Jun 30, 2014 at 12:47 PM, Gwern Branwen <gwern at gwern.net> wrote:
> - Via http://www.patheos.com/blogs/jenniferfitz/2014/06/should-you-be-writing-right-now-and-what-about-the-kids-and-what-about-a-conference-this-summer/
> , http://www.christiannewswire.com/news/9005474225.html "Prominent
> Catholic Writers to Speak at Catholic Writers Conference in Chicago
> Area":

By the way, I know how easy it is to let time slip by and never get
around to doing something you always intended to do but never quite
got around to. If you want to write fanmail, see Wolfe in the flesh,
get an autograph for a book, perhaps buy an autographed special
edition etc, you should do so soon. As soon as possible.

It will sound morbid to point this out, but the odds are not good for
Wolfe surviving too many more years, and if you pass up opportunities
you may well discover you passed up your last such opportunity. List
members may not have much experience with how elderly people can
abruptly decline and die, or how mortality exponential increases over
time in the Gompertz curve, so here's some numerical calculations from
SSA actuarial tables.

https://en.wikipedia.org/wiki/Gene_Wolfe tell us that Wolfe is now age
83. http://www.ssa.gov/oact/STATS/table4c6.html 83: instantaneous risk
of death that year, 0.083230 or 8.3%; remaining life expectancy: 6.66
years.

Over the next 7 years, each year an American male's risk of death is:

- 83: 0.083230 (8%)
- 84: 0.091933 (9%)
- 85: 0.101625 (10%)
- 86: 0.112448 (11%)
- 87: 0.124502 (12%)
- 88: 0.137837 (14%)
- 89: 0.152458 (15%)
- 90: 0.168352 (17%)
- 91: 0.185486 (19%)
- 92: 0.203817 (20%)
- 93: 0.223298 (22%)

Risk increases dramatically each year, and since that's just the risk
each year, the cumulative risk increases even more dramatically. This
is why despite billions of people on earth, only one person (Jeanne
Calment) has survived to 122 years (and the record age has actually
fallen steadily ever since).

The probability of surviving to a particular age is the conjunction of
each negated risk (since you have to not die in each previous year);
so to survive to 84, one has to beat 1 * (1-0.083230) = 0.9168, and to
survive to 85, one has to beat 1 * (1-0.083230) * (1-0.091933) =
0.8325, and so on. It's easy to automate the calculation up to age 93
in Haskell:

    ghci> scanl (*) 1 $ map (1-) [0.083230, 0.091933, 0.101625,
0.112448, 0.124502, 0.137837, 0.152458, 0.168352, 0.185486, 0.203817,
0.223298]
          ~>
             [1.0,0.91677,0.8324885835899999,0.7478869312826661,0.6637885416337929,0.5811455406233024,0.5010421827404082,0.42465429364417107,
              0.35316289400058754,0.28765612144399455,0.2290269137396439,0.1778856619554089]

Or, the probability an American male (like Gene Wolfe) will survive to
a specified age, having survived to age 83, is:

- 1y, 84: 92%
- 2y, 85: 83% (4/5)
- 3y, 86: 75% (3/4)
- 4y, 87: 66% (2/3)
- 5y, 88: 58%
- 6y, 89: 50% (coin flip)
- 7y, 90: 42%
- 8y, 91: 35% (1/3)
- 9y, 92: 29%
- 10y, 93: 23% (1/5)
- 11y, 94: 18%

These figures are population-wide averages, which don't take into
account individual circumstances. On the positive side, Wolfe is
high-status and has received many honors and can look back on a
sterling writing career with great satisfaction; on the negative,
Wolfe has lost his wife recently, which is a well-known risk factor
for men, is not in perfect health, and may be overweight. On net, I'd
say these figures are underestimates.

So, this implies that if you pass up an opportunity this year in favor
of next year, there's a >10% chance Wolfe will be permanently delayed;
if you wait 2 years, ~20%, 3 years, 25%, and so on. These are not
negligible risks.

-- 
gwern
http://www.gwern.net