Open BrandonY opened 7 years ago
Awesome! I think that's a great feature. I'd love to see a PR that recognizes the expected end time increasing as time passed increases, using actuarial tables.
I think that's part of some sort of Bayesian thing along the lines of "if you got this far you're more likely to chug along a bit longer". I imagine actuarial tables would have this covered. This whole personalised deathstimate thing is a bit of a minefield when you start to unpack it.
Assuming that the user is looking at the clock at any given moment, we know that the user is still alive. Thus, we must update the expected remaining days of life upwards, given that additional information. For example, if the user is an infant, the output might read "76 years". However, once the user is 75 years old, the output should now read around "11 years", as the average 76 year old still has another 10 years or so of life left. Similarly, once the user reaches 110, the clock should still be counting down (and display around half a year).
For U.S. humans, here's a reasonable basis for calculating remaining days: https://www.ssa.gov/oact/STATS/table4c6.html