Hacker News new | ask | show | jobs
by crwalker 2779 days ago
It's a simple way to express the difference between probabilities for an ensemble of game players and probabilities for a particular player over time.

If the player is ruined, the game stops for them.

More from Taleb: https://medium.com/incerto/the-logic-of-risk-taking-107bf410...

--edit: added "for them"

1 comments

Optimal strategy is probably a curve: start compliant, follow the rules, be conservative, then take bigger and bigger risks until you can go “all in” on a series of risky bets, and if you aren’t obliterated dial back your risks to favor longevity.

If you are obliterated, well, you served the population.

This is probably more psychologically palettable for men than women. Although some women can do it for sure. Just... well at the population level it doesn’t make nearly as much sense evolutionarily for folks with wombs to take these kinds of risks as folks producing sperm.

Optimal strategy for the population, perhaps, but as an individual avoiding 'all-in' is very nice, if not always possible.

In the context of SoftBank, though, I'd expect all-ins and near all-ins to be avoidable...

Death only hurts once, but success feels good forever. So, in that sense it’s worth it to take the bet.

Optimality is just maximizing your upside and minimizing the window of exposure to oblivion. I don’t think there’s any “avoid oblivion at all costs” strategy that comes close to “court oblivion once” in terms of the total expected utility.