Finding Clues In The Counter-Intuitive

I was training a client recently and overheard a conversation in which one analyst was helping a new team member learn to use a recently deployed strategy simulator.  The new analyst had been experimenting with the simulator and the more experienced one was attempting to verify with her that the outputs were “as expected” and that “nothing counter-intuitive happened.”

I suggest taking a different perspective.  Counter-intuitive results are precisely what we should be looking for because those indicate one of two things: error or opportunity.

First, counter-intuitive results can be a sign that some critical piece of logic or data was left out of the model, that a formula uses an incorrect equation, or even something as simple as an output displays the wrong variable.  Here the counter-intuitive result indicates a bug that needs resolution.

More importantly, counter-intuitive results can flag an error in our own thinking (literally “counter to intuition”).  If we expect a certain result and something surprising happens, that sets us on the path to learning.  Why does that result differ from what we expected?  Which factors contributed?  Are non-linearities pushing us into a new behavioral realm?  Are interactions occurring between variables that have reached values we didn’t expect to appear together in reality?  Understanding the causal logic behind the counter-intuitive opens the door to better solutions.

Once we uncover the reasons for counter-intuitive results (assuming model error isn’t the culprit), we can take advantage of our new understanding by capitalizing on some effect that runs counter to the conventional wisdom.  Perhaps there’s a potentially harmful blind spot in our strategy that can be repaired before competitors take notice?  Better yet, maybe there’s a flaw in a competitor’s strategy that we can seize upon or a market opportunity we can exploit?  This is one of the big advantages to using a simulator, by the way: you can freely experiment and explore the virtual universe in search of counter-intuitive outcomes that might be beneficial.

The key is to recognize the counter-intuitive as a harbinger of new revelations about the system we’re investigating.