Book Cover

Chapter 16, Causes Trump Statistics, was revelatory for me the first time I read  Thinking, Fast and Slow, and it was revelatory during this read. Over my career, I have been shocked many times to see a perfectly sane leader stand up and show a single statistic or estimate which promises delivery of a product at a cost or in a timeframe that is well outside of normal performance.  This chapter provides a rationale for what often seems to be less than rational. The content in this chapter helps me understand why statistical facts aren’t perceived to generate black and white answers, even when they do. Kahneman uses a story about taxi cabs to illustrate the difference between statistical base rates and causal base rates.  Statistical base rates are facts about the population but are not specific to any individual case. Causal base rates are effective because they are specific and are easily woven into a narrative about the case.  

Here is an example of the statistical base rate. Last night firm XYZ’s nightly maintenance application experienced a defect.  There are two teams that develop code for the nightly maintenance application. Team One supports 85% of the codebase and Team Two the other 15%.  The Production Manager blamed Team Two for the defect. He is right 80% of the time. Statistically, the chances of the defect belonging to Team Two is 41% (I got my statistics book out – I highly recommend having a copy of Schaum’s Outline of Statistics close at hand). As I read and re-read  Kahnemen’s version of the example, I found it hard to accept the implication of the base statistics, the Production Manager is right 80% of the time when assigning guilt — that is not chopped liver.  If instead of being given code base ownership statistics, we were given the causal statistics that Team One delivers 85% of the defects, we jump to a different conclusion of the Production Manager’s veracity. Causal base rates deliver information fits into System 1 thinking and helps the brain create stories that fit into neat categories of norms and probable behaviors.  Kahneman defines categories of norms and behaviors as stereotypes.  

System 1’s use of stereotypes (a form of cognitive bias) has value, it speeds decision making.  The downside is that stereotypes can lead to poor behavior. Defeating stereotypes requires effort and slows decision making by forcing us to get System 2 thinking involved. The early adoption of agile methods had to overcome the stereotype of agile practitioners.  The early stereotype types as that they were all just a bunch of loose cannons. Change often requires confronting stereotypes and changing behavior which means finding the right casual statistics to support the change narrative. Hence the importance of this chapter.  

The last part of the chapter is about changing behavior (Kahneman uses the phrase ‘teaching psychology’). Quoting statistical base rates, for example, the annual VersionOne agile statistics report is interesting and useful for trivia contests but they don’t change behavior. Causal bases rate, because fit the case, are much more apt to change behavior.  Changing behavior boils down to having the right stories that the statistics fit into. The final sentence in the chapter drives the point home very pointedly. “You were more likely to learn something by finding surprises in your own behavior than by hearing surprising facts about people in general.”

Remember, if you do not have a favorite, dog-eared copy of Thinking, Fast and Slow, please buy a copy.  Using the links in this blog entry helps support the blog and its alter-ego, The Software Process and Measurement Cast. Buy a copy on Amazon,  It’s time to get reading!  

The installments:

Week 1: Logistics and Introduction

Week 2: The Characters Of The Story

Week 3: Attention and Effort

Week 4: The Lazy Controller

Week 5: The Associative Machine

Week 6: Cognitive Ease

Week 7: Norms, Surprises, and Causes

Week 8: A Machine for Jumping to Conclusions 

Week 9: How Judgement Happens and Answering An Easier Question 

Week 10:  Law of Small Numbers 

Week 11: Anchors 

Week 12: The Science of Availability 

Week 13: Availability, Emotion, and Risk 

Week 14: Tom W’s Speciality 

Week 15: Linda: Less Is More