Netflix Prize: Number of ratings per customer

Does anyone have a good distribution for characterizing the number of ratings per customer? Graphing the histogram log-log (where the vertical axis is density, not count) it’s clearly not the straight line implied by a power law:

Histogram of the number of ratings per customer

The red line is the gamma distribution fit by the method of moments. It’s way too high on the left, and way too low on the right. Can anyone think of anything better?

This thread on the Netflix Prize forums:

This entry was posted in Random Thoughts. Bookmark the permalink.

Leave a Reply

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

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

Facebook photo

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

Connecting to %s