19 Comments

One theory for this pattern which I like, possibly since I gin'd it up myself, is that this easy to comprehended using a frame work from ecology. View it thru dialect of r/K selection theory. That theory is usually used to describe how many offspring a species has. Which strategy gives rise to more arrogant, overconfident offspring? Turning that around and arrogant overconfidence is a sign. Throw in a little survivor bias and your done. Says something about brightsiding doesn't it?

Expand full comment

Federal Reserve chairmen come to mind.

Expand full comment

Interesting. How accurate are the predictions of professors of economics?

Expand full comment

Interesting post!

Gruen Tenders turn out to be a simple way to induce an unbiased prognosis.

It requires two pieces of information:1. profit projections (for example);2. the degree to which different financial controllers tend overestimate.

Then simply adjust the profit projection by the degree of overconfidence.

Expand full comment

Most likely the disadvantage of making inaccurate predictions is outweighed by the advantage of higher social status.

Leading your nomad tribe into bad weather or poor hunting grounds may les of a disadvantage as not being its leader.

Expand full comment

"unjustified arrogance" may be a universal trait.

This was brought home to me when Burton Malkiel got us all to stand in a room and close our eyes. He then asked all those who knew they were an above average driver to raise their hands. Then he asked us to open our eyes and look around 95% of us had our hands up.

I think he suggested [but I am rusty on my memory here] that if he had asked us the same question about our "investment savvy" the answer would be the same.

And that in a nutshell is why Vanguard is an Index Fund manager not an Active Fund manager.

What evolutionary advantage could this "unjustified arrogance" give us, when it is clearly not accuracy in predictions?

Expand full comment

This reminds me of the "women see modesty in men as a sign of weakness" story. Perhaps there is a link between "successful" dating and arrogance?

Expand full comment

With people of limited ability modesty is merely honesty. But with those who possess great talent it is hypocrisy.

Arthur Schopenhauer

Expand full comment

There is a thread just starting on LW that connects the irrational bias when economic choices are framed as a potential gain or as a potential loss in the context of end of life decisions.

http://lesswrong.com/lw/2jd...

I think these are both caused by the same type of bias.

Expand full comment

If we are used to this level of arrogance, and in my experience we are, then we calibrate the predictions of such people for them based on these high levels of arrogance. I would have assumed that in all the situations described above the professional is probably either overconfident, lying about the odds or both. In order to communicate accurate predictions they then need to keep making the same mistake as everyone else. In my experience these people usually know they are not well calibrated and when they need to be otherwise their answers change but much prefer the miscalibration most of the time.

Expand full comment

That doesn't account for it being 5x too long. If it was just ignorance you would expect a fairly broad distribution around the correct average, or a little high do to normal optimism; there is something else going on here, too.

Expand full comment

Re: "If you are a successful professional, that is probably in part because of your unjustified arrogance."

If there is a causal association between arrogance and success, that would probably act as satisfactory justification for many.

Expand full comment

Robin - regarding the decisions of docs in hospice care, I think there are some very good reasons. First, if you or a loved one are terminally ill, and the doc discloses that treatment X gives you only a 5% chance of extending your life 6 months or more, do you still want it? Almost certainly, and the doc will still want to give it, on the chance that it can help you. Most patients will be happier if they're receiving any treatment that has a remote chance of doing something, rather than accepting fate, and if there is no complicating question of a diminished quality of life as a result of that treatment, then it's a no-brainer. (There often is that complication.)

Second, my experience as a former drug developer and current medical student accords with the article's claim that most oncologists are very reluctant to provide specific life expectancies - for exactly the reason that most of them are well aware that given our current state of knowledge, it's a guessing game. So it seems strange that the article then goes into detail about how badly the oncologists guess. Yes! They know that! That's why they're so reluctant, you dummies! But the story seems to overlook this glaring inconsistency to continue playing into the old stereotype of physicians who think they're God and are unable to gauge their own biases and shortcomings, and I think it's about time we put that to bed. So far in my education I haven't met any of those. Irritable and sleep-deprived docs, yes, a few. But suffering from wild hubris about their own knowledge? Not so far.

Expand full comment

It was posted on OB here.

Expand full comment

Whenever I see bias, I think: is there a way to make money from this? Unfortunately, most people seem averse to "gambling", so betting against their mistaken beliefs doesn't often work in practice. A risk calibration consultant could help bring their beliefs back into line with reality, but the problem is getting them to realise they have a problem in the first place.

Expand full comment

I'm not sure that that's a great test. If there's a very good reason to think a case will go one way or the other then I'd expect a settlement to be much more likely. The cases that go all the way should be the cases which are harder to predict the outcomes. That makes this sort of result still appalling but not quite as appalling.

Expand full comment