You're Likely to Live!

You’re older than you’ve ever been, as the song goes, and now you’re even older. And for each eight years you age, you double your chances of dying. That’s the Gompertz Law of human mortality, which is actually sunnier than you think. If you’re 24 today, you stand a 0.3 percent chance of dying in the next year. That chance doubles, to just 0.6 percent, by the time you’re 32. But the compound interest of your corporeal decay mounts quickly. By the time you’re 89, your chances of expiring before the end of the year rise to 86.4 percent. The Gravity and Levity blog offers a wonderfully clear explanation of the Gompertz curve, with a handy visual calculator of your chances of living out the day. And now you’re older still. [%comments]


I think all the probabilities in this little summary need to be divided by 10. For a 25-year-old, the probability of dying during the year is only 0.03%. For an 89-year-old it should be about 8%. It doesn't get as high as 50% until age 100 or so. But the CUMULATIVE probability of dying before a given age can be pretty high (about 50% at age 80).

Allan Crossman

"By the time you're 89, your chances of expiring before the end of the year rise to 86.4 percent."

That implies that there should be about 7 times as many 89 year olds as people over 90. Which is not the case:


Today is my 40-somethingth birthday. Way to make me feel better. Thanks a lot.


Actually, the Gompertz and Makeham models of mortality do not seem to model accurately the mortality rate at ages 80 and above, where the increase in mortality rates slows down.

I think the probability of dying within the year at 89 was really around 15-20% or so.


"...for each eight years you age, you double your chances of dying. That's the Gompertz Law of human mortality..."

To check the math on this you would not compare the number of 89 year olds to 90 year olds, but would compare the number of people 82-89 to the number of 90-97 year olds. He groups the ages in groups of 8 years, not in groups of one year. Make sense? I will say that I did not waste my time checking his math though because I don't care that much! LOL


Uhh ... hold on! The "visual calculator" absolutely, positively does not show your chance of dying in any year. It tells you only the percentage of folks who live to any given age. That's it. It says that 86.4% of people die by age 89, not that 86.4% of 89 years olds will die within the year.

(Of course, this means all the other commenters are right. An 89 year old's chance of dying in a year is nothing like 86%.)


Everyone knows that times matter when it comes to economy. "Time is gold" if you want to buy stock start at an young agebecause you do not want to be old and start investing on stock, it is already too late fore you. Other than you chances of die is high, it will take time for stock to rise or even go down. When your young you buy your stuff and yo do not have to care if it rises or goes down because you still have time to wait until it rises but when your old you know time is running short and in such short amount of time you cannot determine whether your stock's will go down or up. Therefore, it better to try to do things as soon as possible because every time you waste is every money that you can lose. You know that if you do not work when your on your twenties and starts at your thirties you know that your going to earn less money now that you have wasted ten years of your life. Not only in stock or jobs but any investment requires times to pay off therefore do not try to invest when it is too late.



Interesting, but covariates and confidence intervals would be nice. My guess is that the XB could belong in the exponential AND superexponential.


If 3.7% of people make it to 95 and 0.4% last to 100, then there is a ~11% probability of surviving from 95 to 100, so each year would have a survival probability of 64%.

Except that is declining, so 95-96 is probably more, and 99-100 is less, still maybe about 50%.

In Australia every used to know that "you get a telegram from the Queen if you turn 100" even if we didn't know what a telegram was.

There are variations by sex as well. Actually looking at the M:F ratios by age in the Australian census, some interesting findings:
Children are 51.2% M
Crosses over at age 27
At age 47, 48.7% M
But much to my surprise, it turns around!
Crosses back again at age 58. Age 60 is 50.5% M.
Hovers around 50% and after 65 it's always under.
Age 78 is 45% M
Age 85-89, 35% M
Age 100+ 25% M.

These are more balanced than I thought; perhaps women stay longer in aged care, and their majority at aged care homes actually overstates their majority in the elderly years.



And also, the older you get the less you age. Look at it this way: between 1 year old and 2 year old, you age 100 % ; but between 100 and 101, you only age 1 %.


This curve reminded me of the book "Black Swan" by Nassim Nicholas Taleb

Tom L

I know it's been said already, but this post is in desperate need of revision. The formula in the linked blog gives the cumulative odds of dying, not the odds of dying within the next year: a newborn has a 0.3% chance of dying before age 25, a 0.6% chance of dying before age 33, and a 86.4% chance of dying before age 90.


These numbers are bogus. Even the 86.4% cumulative chance is highly variable depending on the mortality table used. For example, the current mortality table for valuing pension plans gives the cumulative chance of dying prior to age 90 at only ~73%.

science minded

If there is one thing that I have learned in the last 24 hours, this is not the way to live.


"By the time you're 89, your chances of expiring before the end of the year rise to 86.4 percent."

And by the time you're 97, your chances of expiring before the end of the year rise to 172.8%. In fact, they are so high, that small farm animals may die in your presence, just to maintain the cosmic balance.


The probability of us dying is 100%. You can't beat the reaper. The rest is a parlor game.