jump to navigation

Numbers in my life – Orkut friends edition January 4, 2007

Posted by Sharath Rao in humor, numbers-in-my-life, science.

Disclaimer: Following Achala’s comments, I decided to put this up here. This post is not intended to convert friends into a numerical representation. Although this list has different biases – more guys than girls, more Littlerockers, KRECs, Carnegie Mellonites than from other schools- it was just easy to download and analyse. Respect, everybody.

The zipf’s law suggests that few unique words take up most of the language when we speak or write. For eg : words such as “a” or “the” are much more common than say the word “letter”. Also related to the Pareto principle or the 80-20 rule.It also applies to several other things in life. See here for example .

You can have some fun by entering a word and seeing its rank among frequently used words. On the lighter side, its heartening to know that love ( ranked 384 ) is more common than hate ( ranked 3107 ). ;), though there is more war ( ranked 304 ) than peace ( ranked 1155 ).

Thats probably because we love war and hate peace ;).

Anyway, does the famed zipf’s law apply to people’s names ? Lets see. If you are my friend on Orkut, then thanks for being a part of this experiment :). Here are some statistics on my Orkut friends list – just with their names.

Number of friends in all – 690. Of these 690, only 664 names were considered – the rest having words such that “Happy New Year” etc. as their names or some other special characters that arent their original names and since I was getting this through a simple program, I didnt want to do any manual modifications other than this – with all respect to the individual’s I normalized names to make them equivalent if they sounded identical. For eg : Shwetha and Shweta or Gautam and Gautham. I hope I wont be burnt alive for this.

Turns out although I have 664 friends, in reality there are only 478 names – because there are so many friends with the same name. Here are the most common names in my list with their counts.

sandeep 11
smitha 8
gautam 8
rajesh 7
karthik 7
deepak 7
ashwin 7
nitin 6
shilpa 5
krishna 5
ganesh 5
sunil 4
shwetha 4
sachin 4
rashmi 4
prakash 4
pradeep 4
mahesh 4
divya 4
deepa 4

Not surprising, except I expected Divya to be less common than Deepa.

That apart, there are :

391 names that appear once.
43 names appear twice.
24 names appear thrice
10 names appear 4 times
3 names appear 5 times [ Shilpa Krishna Ganesh ]
1 name appears 6 times [ Nitin ]
4 names appear 7 times [ Rajesh, Deepak, Karthik ]
2 names appear 8 times [ Smitha, Gautam ]
1 name appears 11 times [ Sandeep ]

So in summary, 1.6% of all my friends bear the name Sandeep – in other words, given that you are my friend, there is a 1.6% probability that you are Sandeep :). I googled but I could not find the frequency of the name “Sandeep” in the common population ( like I have for American names ) ; if that were say 0.6% or statistically different, then that would mean there is a special affinity. [ of it were say 2% ( unlikely ), then lets maintain a distance if your name is Sandeep ;).

[ Plotting name against the probability of occurence; there are few names that occur very often and several names that occur rarely ]

What about for names or atleast my friends ? Not exactly, though to a reasonable degree. A quick check – if it were, approximately 80% of the friends [ 531] would have 20% of the most common names. [ 96 names ]. In my case, 96 most common names account for 282 people ( 42% of the friends ). Or maybe I dont have enough data to make a conclusion. [ as is usually the case with empirical methods ]

On the more naughtier side, if I were somehow obsessed with this law, I would have to stop sending requests or accepting requests from people who have names other than the ones in the top 20 most frequent names.



1. Achala - January 4, 2007

Converting friends into statistics… *gropes for words to convey disbelief at choice of time-pass*


2. Sharath Rao - January 4, 2007

Thanks for the comment 🙂

Its not about friends, its about people’s names. Orkut list was just easily downloadable available and there was something reasonably random about it.

And whenever you come across things that you think are really pathetic, remember – it can get worse :





3. achala - January 4, 2007

Yeah, I know that it was about names, not people, still…

… you’re incorrigible. 🙂 From where did you get three extra links so fast?

4. Sharath Rao - January 4, 2007

No..I knew that the levitt and dubner of the freakonomics fame had analysed the effect of names on people’s success/discrimination etc…so I just searched for
“names freakonomics” in google. ( remove the quotes and try 😉 )

On second/third thoughts, the more you condemn the denegrating digitization of people’s names, the more you may want to appreciate my honesty in revealing the source of the data 🙂

5. achala - January 4, 2007

I do appreciate the honesty. 🙂

Just noticed the disclaimer. My disbelief was more at the extent of your statistical analysis (that someone would spend so much time analysing frequency of names with plotting et al) than your choice of sampling. I should’ve made it clear earlier. Didn’t know my comments would result in you editing your post…

6. Sharath Rao - January 4, 2007

Oh, easy..easy…was just kidding about honesty n all. Your rebuke mite turn out to be mildest considering the orkut-friends I have made. 🙂

7. achala - January 4, 2007

You’re expecting more rebukes? This is fun for you, isn’t it? 🙂 Anyway, I’m logging off now. Been online for almost 4 hrs now. You’re a bad influence *frown*

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: