Entrepreneurship is all about drawing conclusions under conditions of uncertainty. And most often your Bayesian statistics models will not fit.

For how do you assign a probability to an event to which you have no means of assigning a probability? Is this even worth the exercise?

B-school teaches you to arrive at conclusions with all prior data on hand. This is the correct way. Only, in real life there is always a paucity of data and, of course, time.

One skill I have developed is to keep trying to solve problems whenever I see day-to-day mismanagement around me. Skills develop with practice and the easiest way to develop skills is to practice.

How would you solve the traffic jams in the city? You know how to squeeze ahead of the car in front of you, but how about applying yourself to solve the city's problem?

The first thought that comes to my mind is to create a toll for every car that enters the city beyond a point. That creates a new problem! How would you collect?

The toll booth would cause yet another traffic jam. You could create an e-wallet in every cell phone within the city, and an automatic scan and debit could take place. But how does one enable all these cell phones in existence and so on.

As you pose solutions to yourself, you will find the next problem arises as a result of the solution at which you just arrived. Mind you, we are still in the realm of "make believe," so no damage has been done. As yet.

Drop that problem for now, and go on the next problem that you see around you. How can we get the entire two-wheeler driving population to wear helmets?

No amount of market research is going to help solve some of these problems and yet there is a solution out there somewhere. Develop this skill through practice on everyday issues around you, and soon you will find that you are solving the much more mundane business problems with great creative ease.

Just get out there and try it. It's a pretty humbling experience, I promise you.

R Narayanan graduated from IIM-Calcutta in 1969.