Friday, March 02, 2007

Freakonomics

I am not a great book reader. I somehow "lack the patience", you might say, for reading books. I feel it is a rather slow process. It is not that I read slowly, I read pretty fast. I did quantify it too the other day :-). But then it still seems slow compared to watching visually. The data transfer I feel is much faster in case of video. Even if I read books, I don't go near fiction, movies are better :-).

But offlate I have started reading much to my own surprise. I still read only non-fiction though. One of the recent books I read was "Freakonmics" by Steven Levitt and Stephen Dubner.

As one can guess from the title, it talks about "Freak Economics". But actually it is a real eye opening kind of look at various things which we take for granted and assume the common knowledge about it. The book emphasizes on one thing, "numbers speak more than just plain words". The author(s) say how a look at the data of a certain aspect of life or society can tell a different story than the way we know it. They call what we know generally as "conventional wisdom". It brings out the fact that how this "conventional wisdom" is so wrong when we just get to the bottom of the thing with data and numbers.

The book at the outset is structured more like a blog. And as rightly claimed by the authors it does not have a central theme. I guess, the earlier paragraph in this topic is the closest to the central theme we can get of this book. This might not be liked by some purist readers but then that is how it is. I did not mind. :-0

The authors have asked completely non related questions on different topics and showed how they are actually linked, that too by hard data. The topics range like the similarity between school teachers of the public schooling system in USA and Sumo wrestlers in Japan. It talks about parenting myths etc.

Now then I guess I liked this book simply because I am myself like that. I generally fish for numbers before making decisions. I feel calculated decisions based on data or for that matter analysis of certain event/topic based on data rather than just gut feelings and "conventional wisdom" are more accurate.

I can recall one such incident long back in my career where I felt how the data makes a difference. We had tried to emphasize on our management against a certain decision in our project. We had met the managment umpteen number of times had many meetings and brainstorming sessions in trying to convince the management against the decision which had been taken. We were well into the project cycle and we were still unable to reverse the decision.

Then one fine day I hit upon an idea and made some impressive charts based on hard data from many tests and extrapolations from that. The charts immediately showed that the path we were following is not the right one for the project. That was it. The approach was changed by the end of the meeting. We had prevailed; or was it we or the "data". It was surely the data and its presentation. All our earlier meetings were only talks about feelings and common knowledge. But that was not exactly quantifiable. The moment we were able to put it in numbers it hardly took us time to convince the management.

This exactly is emphasized in Freakonmics too. The fallout of reading this book has reiterated by belief in numbers and its use.

Read the book guys. BTW there is a blog too hosted by the authors.

3 comments:

Saumya said...

Sounds interesting. Must pick it up sometime...

Parry said...

hmmm ... interesting ... I remember Vidya mentioning this one a lot as well ...

Anonymous said...

I picked up Freakonomics with Pooja just before she quit. I finished it the same day, by night fall, It was un-putdownable.

To me it was the diverse topics and the emphasis on information asymmetry when we make decisions (the one on real estate brokers) that made great sense.

Interestingly, you can also read Malcolm Gladwell's critique on Freaknomic's chapter on why crime fell in America in the 90s at Malcolm Gladwell's blog. It's a fine example of interpretting data in multiple ways.

-Vidya