gwern comments on Open Thread: July 2010 - Less Wrong

6 Post author: komponisto 01 July 2010 09:20PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (653)

You are viewing a single comment's thread. Show more comments above.

Comment author: pjeby 06 July 2010 08:55:41PM 2 points [-]

Reflective people are LESS risk averse.

That's even more confusing. I would expect a reflective person to be more self-doubtful and more risk-averse than a non-reflective person, all else being equal. But perhaps a different definition of "reflective" is involved here.

Comment author: gwern 07 July 2010 02:09:55AM 2 points [-]

But perhaps a different definition of "reflective" is involved here.

Possibly. A reflective person can use expected-utility to make choices that regular people would simply categorically avoid. (One might say in game-theoretic terms that a rational player can use mixed strategies, but irrational ones cannot and so can do worse. But that's probably pushing it too far.)

I recall reading one anecdote on an economics blog. The economist lived in an apartment and the nearest parking for his car was quite a ways away. There were tickets for parking on the street. He figured out the likelihood of being ticketed & the fine, and compared its expected disutility against the expected disutility of walking all the way to safe parking and back. It came out in favor of just eating the occasional ticket. His wife was horrified at him deliberately risking the fines.

Isn't this a case of rational reflection leading to an acceptance of risk which his less-reflective wife was averse to?

Comment author: gwern 09 July 2010 05:16:07AM 2 points [-]

In a serendipitous and quite germane piece of research, Marginal Revolution links to a study on IQ and risk-aversion:

"Our main finding is that risk aversion and impatience both vary systematically with cognitive ability. Individuals with higher cognitive ability are significantly more willing to take risks in the lottery experiments and are significantly more patient over the year-long time horizon studied in the intertemporal choice experiment."