Tuesday, October 6, 2009

Randomness and How to beat it

Geek-speak:

I recently interviewed the Nobel Laureate in Statistical Sciences (play along guys, that's a hypothetical award made up exclusively to cater to this article) Lord Codesmith Cipher, and asked him his views on information entropy. The senior scientist got beautifully philosophical in course of elucidating his answer. Find below the precise transcript of what he spoke:

"Randomness is what we use to describe lack of predictability. It essentially signifies the uncertainty associated with precisely identifying consequences of fluctuations in inputs to an other-wise understood system. Information theorists with a background in communication systems employ entropy considerations to quantify randomness in a system. Statisticians call it 'error', and minimize it using least squares. Signal processing engineers usually model randomness against a 'baseline' and try to 'fit it' with a gaussian or 'colored noise' frequency response. The credit-worthiness analyst calls it 'risk', and uses a host of tools and tactics to bridle it.

What is common to all of the above is the intent to quantify randomness. When Man was faced with uncertainty He did not give in to it. True, you cannot predict the precise consequences of an input fluctuation until you know all the driving forces of the system. You can only do so well as to minimize the uncertainty associated with the expected output or response. But does that mean we do not give it a fight at all?

Bowing to uncertainty isn't what brought Man to where he is today. It's the intent to fight it out, which got him here. Most of us who make it across those notoriously competitive tests held for admission into prestigious institutions of higher learning do so because they choose to fight the uncertainty. To take randomness by its horns. To evaluate the merits of the System, to identify beatable elements, to work around unbeatable ones. To quantify risk. To do a SWOT analysis and lock onto opportunities. To break set patterns. To walk that one extra mile to narrow down that uncertainty window by one micron.

From experience of having interacted with several high-performers of varying capacities, I have come to understand something interesting about fighting randomness: "The incremental reduction in uncertainty increases as a function of applied effort." In English, what it means it means is this: to reduce the uncertainty from 15% to 10%, the effort required will always be more than that required to reduce uncertainty from 20% to 15%. And so on as we reach zero.

The corollary to the aforementioned hypothesis is even more interesting: "Incremental reduction in uncertainty so as to reach zero uncertainty will require infinite effort, and consequently is unattainable in the real world." If you've ever fought for a cause, you'll know that the aforementioned hypothesis-and-corrollary are eerily correct. So, however hard you try for something, you can never be 100% sure about the outcome. But that's how the world is.

We all have causes that are close to our hearts. Each of which merits an effort that'll make eyeballs turn. How about you pick up your cause and take it through? Why sit back and give up in the face of uncertainty? I respect my uncertainties. They give me a reason to exist. To fend against. They're my food for thought. My intellectual high. Where's yours?"