Search Now

Recommendations

Friday, September 02, 2005

Small-world effect


CONSIDER this. You receive private information that a certain stock is likely to move up. You share this information with your friend. Two days later, you find that the volumes have increased five-fold and the stock has hit the upper circuit! You wonder how the market knew about the stock. Such phenomenon can be explained by the small-world effect.

At a birthday party, you meet a person for the first time. After a couple of minutes, you find that you are related to that person, or that both of you come from the same hometown. You conclude that the world is indeed small!

In the mid-1960s, Stanley Milgram, then at the Harvard University, performed an experiment to test the small-world effect.

Suppose you give an unaddressed letter to your neighbour, asking him to give it to someone he thinks may know your friend in the US. Based on the experiments conducted by Milgram and others, chances are that your friend will receive the letter by the time it changes six hands. This came to be famously called as the six degrees of separation.

Academicians have applied these experiments to study networks. What if we want to connect various towns across the State?

The small-world effect shows that you need not build roads linking each town to every other town. Paul Erdos, the famous Hungarian mathematician, computed the minimum number required to link any network. Interestingly, the minimum number gets smaller, larger the network.

Now, we know that the investors' network is large. Your friend may have told his friend who, in turn, may have told some others. Soon, the message may have passed on to the entire market. That is, perhaps, why no market information remains private for a long time.

Hindu Business Line