Friday, July 3, 2009

birthday paradox

I was going throught the question that what is the minimum batch size required in a system to ensure that two people at least have the same birthdays and did a simulation of that in excel.
However, I thought that if we can have a maxima or minima funda in P(n)=1-365!/(365-n)!*365^n we can find the value of n.
The answer is 50 approx.
Let us see.

2 comments:

The Avalanche said...

shouldn't the batch size be 366 for normal years and 367 for leap years?

How come it can be 50??


Kindly enlighten me on this... :-)

rakki said...

try it for any group sixe of 50 and you will get it...
i ll try to explain the rationale very soon...
thanx for the inquisitiveness!