When Expertise Works And When It Doesn’t

25 Jul

At this link is a Google talk delivered by psychologist and Nobel Laurette, Daniel Kahneman. The topic of the talk is expert judgement in decision making, and Kahneman discusses the collaborative work he carried out with Gary Klein.

On the surface, Klein and Kahneman seem to fundamentally disagree on decision making, with Kahneman believing that human beings are poor intuitive judges of situations and assessing choices, and Klein believing that human intuition can deliver outstanding results. The reality is that Kahneman investigates mistakes (Kahneman, 2011), and Klein investigates successes (2007). So, if you read Kahneman you will read research into bias and error, and if you read Klein you will read research into insight and successful intuitive choices. This creates the impression that the two researchers are operating in opposition. However, they both agree on the subject of decision making, far more than they disagree (Kahneman and Klein, 2009).

Kahneman and Klein agree that intuition only works well when the environment an expert operates in has visible cues and patterns, which lead to outcomes on which the decision maker gets fast and frequent feedback. To borrow a Kahneman example, an anaesthetist operates in an environment where there are visible cues and patterns (patient blood pressure, heart rate for example) and the decisions of the anaesthetist produce regular feedback (the drugs produce visible responses in the patient).

Opposite environments include stock picking (statistically no better than chance) and long term forecasting (the same). The cues and patterns are noisy, complex and highly uncertain. Any visible cue and pattern is very difficult to confidently link to an outcome making feedback problematic.

However, experience in a noisy, complex and uncertain area can still give a person superior knowledge of risk, and what could go wrong with decisions. This knowledge does not have predictive value, but it does allow the owner of this knowledge to avoid loss, and so retain options to adapt.

A good example, is the story of Jim Paul in the book, What I learned Losing a Million Dollars (Paul, Moynihan, 1994, revised 2014). Paul loses a fortune due to his over confidence in picking stocks, mistaking his luck for skill. As Paul got back on his feet and reviewed the advice of expert stock pickers, he realised the advice contradicted each other to the point of being unusable. However, all the experts agreed on how not to lose, how to stay in the game, retain options, and pivot.

Nassim Taleb (2012) calls this via negativa, a goal can frequently be better pursued by AVOIDING outcomes. For example, if your goal is to have a long life, achieve it by avoiding death (ibid). My favourite Taleb example is, if you can’t find a role model, find a fear model, someone you don’t want to end up like.

In short, genuine expertise is confined to environmental conditions. However, experienced sense making on risk and what could go wrong with decisions, allows a person to adapt to plan B (and C, D etc.) in even noisy, complex and uncertain situations.


Kahneman,D. (2011) Thinking Fast and Slow. Penguin

Paul, J. Moynihan, B. (1994) What I Learned Losing a Million Dollars. Columbia Business School Publishing

Klein, G. (2007) The Power of Intuition: How to Use Your Gut Feelings to Make Better Decisions at Work. Currency

Kahneman, D. Klein, G. (2009). Conditions for intuitive expertise: A failure to disagree. American Psychologist, 80: 237–251

Taleb, N. N. (2012) Antifragile: Things That Gain from Disorder. New York: Random House



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: