But suppose you’re taking on a tough orchestral piece in an educational setting or working extra hours to come up with a general solution to a messy problem that’s tying your organization in knots? Suppose you’re attempting to learn a new skill that the local experts say you shouldn’t attempt because existing data say you’ll fail? Suppose you’re the only one willing to step up to the plate when everyone is shrinking back from a necessary task that threatens to be tiring or humiliating? Or if you’ve got a bright idea for a tool or process that might succeed in making society better off, or just might wind up being a waste of your time and energy?

Over the years, I’ve seen people confuse guts and irresponsibility. When you gamble on something you don’t understand and/or can’t control, that’s giddiness. When you gamble that you’ll finish cleaning up a huge mess that no one else wants to touch with a ten-foot pole, that’s pretty noble, I think. We need to recognize the difference!

It may seem like that, but it’s not. The results of this study are not showing things by random chance. A real world example of why something like this might have evolved would resemble maybe a hunter in the past who looked at a bear and thought he could kill it for lots of food. Clearly, bears have a quite distinct physical advantage over us in every department. A rationally thinking person at the time (without guns or whatever tools we may use today besides a spear) would look at that situation and see it as a poor idea and not even try (even though they may have been able to). An underconfident person, as well. Thus, neither have food. An overconfident person that looked at that scenario and actually was able to kill the animal (the important part) now has food for a long time and can devote time to increasing his tools or whatever he may wish to do. That would be quite attractive to primitive mates (the ability to care for a mate/family).

http://www.nytimes.com/2011/09/16/opinion/brooks-the-planning-fallacy.html?_r=1&ref=davidbrooks

I think his overarching point was that the time needed to change complex systems, where multiple players must agree and coordinate action, is consistently underestimated by the players.

Related to the above, I think, by the fact that the players perhaps also don’t have complete information about their situation. Of course if the problem is truly one of overconfidence, then complete information would not really help, would it? At best it might be part of a solution, but it could not be a full solution by itself.