I'm taking a quick break from my KenPom data visualization work this week (though I have made some updates; work in progress on the projects page) to talk about another project I've been working on. I'm an avid runner and have an upcoming race where I'm a little worried about the elevation changes so I thought bringing in some programming could help.
In the process of writing an article about March Madness ticket prices (coming soon), I ended up on kenpom.com, an absolutely fantastic site for college basketball analysis. The rankings and associated metrics produced by the site's namesake, Ken Pomeroy, have become nearly industry standard for coaches and fans alike. Note that parts of the site are behind a paywall but if you have even the slightest overlaping interesting in college basketball and statistics, it's definitely worth the subscription fee of $19.99.
The whole site is worth a look but I'd like to focus on one particular statistic here: win probability.
The last thing the internet needs is one more demonstration of Benford's Law but I'm going to do it anyway for three reasons: