The Birthday Problem

By Chris Orwa
Data Science Lab
  Published 07 Jan 2016
Share this Article

If you keep tabs on birthday notifications on Facebook, you’d realize some of your friends do share a birthday. If so, what are the chances of two randomly chosen friends having the same birthday? This is a well-studied problem in statistics referred to as the ‘Birthday Problem’, which states – If N people are in a room, what is the probability of two people sharing a birthday? I decided to investigate. Straight to my Facebook account, I collected 628 friends’ birthdays, wrote a scriptand bingo I had a graph.

A shown on the graph above, there is a 0.5 probability of sharing a birthday in a group on 24 people, and a 1.0 probability in a group of 70. The interesting bit of this problem is the fact that most people expect a group larger than 365 people to have a chance of sharing a birthday. Go to Facebook > select 70 random friends >  you’ll be guaranteed at least two of them share a birthday. This is not just fancy recreational mathematics, a lot of computing, engineering and scientific processes rely on this concept.

In medicine, the concept is used to calculateClass Phenotype Probability – the chance of two people sharing the same blood-type? This particularly important in calculating the likelihood of finding matches between donors and recipients in blood transfusion. In crypt-analysis, a code breaking technique known as the birthday attackutilizes the same concept to find chances of a pattern repeating itself in an encryption scheme.


 Since the US has had 44 presidents, according to my graph there’s a 0.89 chances two of them will share a birthday. Bingo! Yes! Warren Harding(29th presidents) andJames K. Polk(11th president) both have a birthday on November 2nd.


 

comments powered by Disqus