Though I've put up a few smaller exercises before, I finally have a draft version of the first full data visualization that I've done the majority of the work on .
It's based on a metric I created last week using kenpom data.
It's not especially pretty or user friendly yet but, in the broadest sense of the word, it works. Check it out at the direct link or on the project page and let me know what you think.
Here's an short explanation of the axes since the project page is still light on info:
x-axis: perDrama, a measure of game excitement described in my earlier post
y-axis: combined eFG%, based on Dean Oliver’s four factors concept described here.
Circle radius is based on "Hype", the scaled harmonic mean of kenpom ratings entering the game with an added "penalty" for the lower ranked team. The exact formula (where ranking equals the 351 - the kenpom ranking) looks like:
I'll also give a lot of credit to my co-workers Ross and Ilan for talking the idea over with me, as well as Mike Bostock, Scott Murray, Steve Hall, and Justin Palmer for their libraries, tutorials, and code examples.