5 Comments

Teach them poker.

Expand full comment

So if it's confirmed repeatedly that people suck at estimating risk, what exercises can people do to better assess risk?

E.g., I met a guy in Australia who was launching a startup. He'd been a grad student in the mathy part of computer science known as "machine learning" (Artificial Intelligence without the magic robot promises) and he'd realized he had a talent for poker. He attributed his balls to start the startup and his success to date purely to the poker--it taught him, he said, how to balance risk and reward.

I have kids and would love to know what I can do to help them get a more accurate intuition for risk and reward.

Expand full comment

Stuart, if your, my and Robin's goal is to maximize persistence odds for the three of us, it could be wasted effort to try to convince 6 billion people (or 300 million) to estimate risks better in a diffuse way like that.

Expand full comment

It's funny the line you end on, Robin, because that would be a very enjoyable blog, at least to me. What would supplement such asbtracts nicely would be commentary on them by the experts here.

Expand full comment

How can we get the message to everyone, though? Giant add campaigns that scream "You are bad at estimating risks!" probably won't work. Though there are a few sucessful campaigns along these lignes - especially for drunk driving, convincing people they can't estimate how much they've drunk.

But that has two advantages - people are far more willing to believe that alchool impares their judgement, than that their judgement is naturaly impared. Secondly there is a paternalism behind it all that "knows" and enforces the correct behaviour. Neither of these are present in general.

Expand full comment