Daniel Kahneman -- Thinking, fast and slow
============================================

The parts I found to be most valuable were those that warned of
common decision making mistakes or fallacies.  Many of these are the
mistakes we make when we use what Kahneman calls System 1 thinking
where we should be using System 2 think.  Roughly this means that we
use quick, intuitive, evaluative judgments where we should use slow,
calculating, judgments that take into account lists of positives and
negatives, possible problems as well as possible benefits,
experiences in other similar situations and projects, and so on.

Kahneman is a bit of a pessimist, or depending on your point of
view, a realist.  He is not the person you want present when you
are trying to convince a possible funder that the project will be
finished quickly, that it will cost very little, that it will pay
for itself, that it will be completed as described and without
incident, etc.

He is also a proud theory buster, or at least he seems to pride
himself on criticizing decisions and proposals that are made by
someone who is blinded by a theory, especially a cherished and
tightly held theory.

Here are some of the decision making fallicies that Kahneman warns
us about:

- The optimistic fallacy -- It's all too easy to become convinced
  that a proposed project can be finished quickly, that it will be
  finished below cost, and that the finished project will do
  everything that the user wants it to do.  As a computer
  programmer, I can say: "Yes, I've seen that before."

- The sunk-cost fallacy -- Once you've put time, effort, money, or
  other resources into a project, it is much harder to back out or
  drop the project, even when it is failing.  Dropping the project
  means admitting failure; it means acknowledging or accepting your
  losses; and it means giving up something in which you have serious
  ego involvement.

- The planning fallacy -- This is one where the decision makers
  consider only unrealistic, best case scenarios and where the
  decision could be much improved by considering other similar
  situations and projects (i.e. by taking a outside view). A variety
  of other considerations should be made in an attempt to avoid the
  planning fallacy, for example, what might go wrong in the future,
  and how the world (the competition, the user's needs, etc) might
  change in the future.

One piece of important advice from Kahneman: Learn to use (short)
checklists and simple algorithm in your daily decisions.  If nothing
else, doing so will encourage you to slow down, to think more
methodically, to use objective criteria rather that intuitions and
likes or dislikes.  Try to set your cherished views aside and to
avoid group think by considering ideas from other sources.  Try to
use a check list or set of questions that reveal a small set of
objective attributes, then use a simple algorithm (often a straight
forward sum is good enough), and finally factor in considerations
that are specially to the current project or decision.

Over and over Kahneman emphasizes *his* opinion that *expert*
opinion and expert judgment is no where near as good as we often
believe it is.  We all, experts and non-experts alike, think that
our judgment is better than it really is.

Kahneman also has a reasonably good discussion about predictability,
under what conditions the future *is* predictable, and under what
conditions "experts" can learn enough to be able to make valid
predictions.  A rough summary is that (1) the future must not be
random; (2) there must be cues and indications from which valid
predictions can be made; and (3) the expert must be given the
opportunity to learn those cues and what follows from them.

03/06/2013

.. vim:ft=rst:fo+=a: