top of page

Statistics and Coaching – why a coaching session is like a Bayesian experiment.

I previously wrote about the link between George E.P. Box’s quote that “All Models are Wrong, but Some are Useful.” I am now going to expand on my thinking about a coaching session being like a Bayesian experiment (with reference to another great figure in statistics—the Reverand Thomas Bayes).



In Bayesian inference, we are updating beliefs when there is uncertainty. We may or may not know anything about the belief before any experiment, and we would quantify our belief in a prior distribution. We conduct an experiment to collect data to update beliefs. For the non-statisticians out there, these data form the likelihood. Then, Bayes’ theorem is used to update the initial beliefs based on new information. Then after the experiment, we have an updated belief represented by the posterior distribution. We can then use this to make predictions and make decisions. In all cases these beliefs are represented by probabilities, which have distributions to represent the uncertainty.


So why is a coaching session like a Bayesian experiment? Well, firstly, I am going to say this is less for the coach and more for the Thinker. Secondly, I will say that in every day life we often think like a Bayesian. In coaching, the Thinker will come to a coaching session with a topic and have some degree of belief around that topic. They may not have any belief but I doubt that is really the case (this is their prior). Then the coach works in partnership with them to help them get to new thinking. This new thinking is their new data or their likelihood.


So, at the end of the session, the Thinker will know something about their situation that they didn’t know before. They now have their posterior. We can ask them about their new thinking: “What do you know now that you didn’t know at the start?” They can use this posterior to think about new things. We can prompt this by asking the question, “What are you going to do with the new thinking?”


So now you can see how a coaching conversation is like a Bayesian experiment. I would be interested to hear other people’s thoughts and views.


As a statistician and a coach, I am inspired to look for links between the two disciplines. 


If you would like to explore having a safe space in which to think, please use the link on my contact page to book a discovery session with me.




TO MAKE AN ENQUIRY PLEASE CALL +44 (0)7824 872561

  • LinkedIn
bottom of page