Dave Radcliffe’s neat “linearity of expectation” tweet

Learned something from a Dave Radcliffe last week:

The link in his tweet goes here:


I had to dig a little deeper to see what Dave meant about linearity of expectation, but when I did I found an incredible solution to the problem in just a couple of tweets:

Screen Shot 2015-12-15 at 7.04.05 PM

Prior to this series of tweets from Dave, probably the most interesting “linearity of expectation” example that I’d seen was a new-to-me proof of the Buffon Needle problem in Jordan Ellenberg’s How not to be Wrong.

Although I don’t have the exact reference in Ellenberg’s book (I have the audio book version),  Lior Pachter has more or less the same neat explanation on his blog. It is amazing to me that changing the needle to a circle solves the problem in a snap!

Buffon’s Needle Problem on Lior Pachter’s blog

Anyway, a couple of points about problem in Dave’s tweets.  First, it wouldn’t have occurred to me to use linearity of expectation to attack this problem, and that is definitely my bad.  Dave’s example really showed me the power of the approach.  Second, my intuition for an approximate answer to the question was off by miles!  It actually took a while for me to understand how the number could be as high as 488  (I mean, you’ve got a 50% chance to win after about 500,000 tries, and, really, how many different people could have won 10 times by then . . . .), but I’m glad that Dave’s tweet made me think about this problem – I definitely needed the intuition adjustment!

Pretty incredible what you can learn on Twitter 🙂


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: