Thinking, Fast and Slow - Not a book review

Daniel Kahneman (Nobel Laureate and one of the founders of behavioural economics) in Thinking, Fast and Slow, identifies a number of biases in our thinking that can lead to poor decision making.

The question is: What done do with his work?
The biases that Kahneman discusses include:
  • Anchoring
  • Availability
  • Substitution
  • Optimism and loss aversion
  • Framing
  • Sunk cost
  • Overconfidence

Unfortunately merely knowing of the existence of a bias is not sufficient to defend against these biases. Mechanisms and habits are required to defend against them.

However it is not sufficient to merely avoid bias, as on occasion a belief in personal exceptionalism can be good (at least for the community as a whole, if not for the average individual). My personal favourite from the book of this is his commentary on the overconfidence of researchers and entrepreneurs … given the statistical chances of success a certain amount of belief in personal exceptionalism is required to even start in these areas.

Some of the solutions suggested include the use of checklists and procedures are used to force the examination for bias.

The flip side of the work is that framing and anchoring are the basis of nudging people to make decisions. The presentation of an idea controls its reception. Which can be used for good or ill.

Operationalising the ideas from his book is a challenge and one that is unlikely to be perfectly managed. Some approaches include:
  • Forcing yourself to do the math when dealing with quantitive especially probabilistic problems - check your gut
  • While optimism can be essential for getting started - pay attention to the failures of others and make tangible plans to avoid the pitfalls that caused their failures
  • The perception of losses and wins can be a matter of starting point - remember the old joke "How do you make a small fortune as an X; Start with a large one"
  • Substitution is a bit vexing for me as problem transformation has served me well1. Make sure that you know that you are substituting the answer for one problem for another and what that substitution means
  • Availability bias has a number of forms:
    • The tool form - "If you have a hammer the world looks like a nail" - always search for other solutions
    • The data form - ask the question is this all the data?
    • The domain form - answers to problems don't necessarily come from within the domain they are found - look for the left field answers
  • Sunk costs. Lots of strategies - none of them always appropriate:
    • My Aunt's Vegas strategy: I put this much money into this visit and stop betting when I get to the end of it (stop loss strategy - she still enjoyed walking away with the winnings but never chased the losses deeper than her holiday budget)
    • External advisors - someone without an emotional investment to analyse your current position
    • Portfolio of opportunity - some can fail if enough succeed - lessens the pull of any individual loss




1 Numerical methods for solving equations are an excellent example. Frequently a set of equations cannot be satisfied by symbolic methods so an exact answer to all possible inputs cannot be found, however, there are numerous iterative methods that solve for a point solution that is good enough