So, “how many hairs do you have on your head”?… is it 10,000 or is it one million?
However, it is important to understand that the exact value may not be required; rather it is the reasoning that allows to come up with a value and constantly improving the reasoning to better the value. The simplest answer to the “hair question” is to compute this by multiplying the density of hair and the area of head approximated as square. The answer can be refined by improving the reasoning that the hair density varies across the head and that we may need to consider the curvature factor of the head, hair shed factor etc.
So what are we doing? Rather than just guess, we have tried to construct a simple formula that is continually refined to better the outcome.
The act of approximation is very natural, it is in fact part of our instinct. Think about this – we do not take measured steps when we walk, we do not calculate the exact distance when reversing the car and so on. Our natural learning system continuously learns in the background and constantly adjusts the variables to refine the approximation.
Read the article by T Ashok, Founder & CEO, that was published in the Tea Time with Testers ezine August 2011 issue. The article illustrates how ‘scientific approximations‘ is very useful in testing.
This topic was also presented at the Bug deBug Conference (Chennai) on March 24, 2012
Below is the extract from a blogpost by Prasanna Kumar Sabat, on his experience about the event on Day 1.
“The First speaker was T Ashok (Founder and CEO of STAG) on a quite interesting and catchy titled subject “How many hairs do you have on your Head? Scientific approximation“. He puzzled the participants with his witty questions around the title and tactfully made his point on how to make scientific approximation. When a baffling question like that hunts you, instead of making a wild guess we should use some rational calculations and approximations to reach a reasonable number. The exact number may not be absolutely correct but the way how you reach the numbers matters, with multiple iterations the number can be refined. The idea seems to be useful for people at a higher management hierarchy who have to deal with lots of numbers and unexpected questions and rationalization. Even though as a human being we make use of approximations in our day to day life(while driving your car in a parking slot or throwing or catching a ball), we have to make some conscious effort to extend that to our work.”