Sunday Thought - Should I Hope to Live Not Too Long

This Marginal Revolution post pointed me to the Atlantic piece by Ezekiel Emanuel: Why I hope to Die at 75 with this key message:
Once I've lived to 75, may approach to my health care will completely change.  I won't actively end my life. But I won't try to prolong it either.
I am 70 and obviously not neutral to this topic. BTW, when I was 12, people at 35 looked so old, and old fashioned in their behavior, that I thought: "I hope to die before becoming that old". Now, I want to live long - but, I mean long and full, rich, exciting, mobile, recognized, loved, …

Death is a loss. But is living too long also a loss?

Loss is overrated

Beyond philosophical, psychological, socioeconomic and cultural aspects, this sounds to me like a question of prediction and risk management. It's about loss aversion (in the sense of  Kahneman). Paradoxically, people, who hate to realize a loss often take more risk when losses increase. Is this what Ezekiel Emanuel wants to avoid?

Optimal Risk

It is difficult to optimize risk, if you don't have enough quantitative information. In Optimal Risk I have briefly described my try to find the optimal risk when I skate on cross country ski trails. But life is more complex than cross country skiing. It "grants" more unexpected events.

Long, but boring?

But, let me take the roulette metaphor. You know you can't win on the long run. The Kelly Criterion (the Kelly bet on "red" was -1/19) tells you not to bet. But you can use a small fraction of your current payroll to stay "long" at the casino playing (just betting on "red"). Boring, isn't it? And the longer you play the payroll will go to zero and a fraction of it may become really too small to continue …

A complete life?

I have most probably celebrated more wine and dine events, slept, exercised, trained my brain, applied  preventive medications… less than I should, to prolong life as long as possible. And maybe, I risked to become slower, less creative and less productive earlier as necessary.

My statistical life expectation is 78. But statistics also says: most probably, I will suffer from this and that "long" before. To NN Taleb "long in history" means "long in the future, but future is unknown. You can't really predict it, but build it.

Logarithmic loss (LogLoss) vs 0/1 loss?

It's not quantifiable. But maybe I was lucky having found a kind of an optimal risk for a full, rich, … life. Mayby, I have intuitively used a kind of LogLoss pay-off instead of a crisp loss (of living long) function, Ezekiel Emanuel seems to "apply"? LogLoss penalizes the extremes (confident or wrong) and "predictions" under its regime are not 0/1.

Inquisitiveness wins

However, it's too many things I haven't seen, understood, managed, … yet. Too many corners I haven't looked around. Life is never "complete". I still work. Good partners and friends will tell me, when I should stop.

But what I certainly know: I do not want to live infinitely long. There is an individual age x_1 … But 75? Really?