Greetings and welcome to the Carousel Analytics blog. The goal of this blog is to evaluate head and assistant college basketball coaches and coaching staffs from an analytical perspective. In general I will look at individual coaches (how does Coach X compare to other coaches with similar profiles?), or entire staffs (do more experienced staffs tend to experience more success at the men’s division 1 basketball level?).

To be clear, Carousel Analytics isn’t designed to compete with Kenpom.com or any other similar websites. The work shown on this blog won’t give you any insight into the quality of a team, or whether Team X will beat Team Y in a given game. The focus is on finding ways to evaluate coaches and coaching staffs to determine which qualities are most important for success. You might be able to use some of the work from this blog to find an edge between two evenly matched teams on a given day, but that is not the goal of my work.

Future blogs will dig into some actual data, but for the first post I want to discuss what Carousel Analytics is, give some background on how the idea came about, and provide a broad overview of what you can expect to see moving forward.

Over the years as I have read the forum of my team of choice I have seen two ideas with regards to head coaches and coaching staffs that got me thinking about the importance of coaching experience and coaching staff composition. First, “Coach X isn’t very nice and his assistant coaches don’t stick around. The level of staff turnover is higher than normal and is having an impact on success” or something along those lines. Second, “Coach X is inexperienced; he/she should hire an experienced staff to make up for his/her inexperience.” etc…

After finishing graduate school, I started tracking coaching staffs for a few High Major programs to look into whether or not those internet forum ideas had any basis in truth. What started with mapping out who the Head Coach and Assistant/Associate Head Coaches were since the 2007-2008 season at one school, became an entire High Major conference. Then it became a second High Major Conference. Then I added a few Mid-Major schools to see how they compared. Eventually I decided to map out the entire coaching staff (or most of it) for every D1 Men’s basketball program since the 2007-2008 season so that I would have a comprehensive dataset to answer a variety of questions across the full D1 basketball spectrum.

In the end, the entire initial data collection process took… a long time… and is ongoing. Teams change their coaching staffs throughout the year, with the majority of changes happening between March and August. In order to keep up with all of the changes I update the dataset on an rolling basis.  To date, there are 3,984 unique coaches since the 2007-2008 season included in the dataset.

For each coach in the dataset I have collected their entire coaching history. This includes any coaching experience from the high school or AAU level through professional or Olympic level teams, for both men’s and women’s basketball. Once I know where each coach has been throughout their career, I am able to match that data with team performance data (ie, wins/losses, points per 100 possessions, NCAA tournament performance, etc…) to come up with ways to evaluate individual coaches and coaching staffs.

From there it becomes possible to start evaluating individual coaches and coaching staffs. For example: How does a specific coach compare to other coaches with similar experience profiles? Does a school tend to perform better or worse in the seasons before, during, and after a coach was on staff? Do teams with more experienced staffs have more success at the D1 men’s basketball level? Does a team with less staff turnover have more success than a team with more staff turnover? The list of questions is endless.

I hope you enjoy what the Carousel Analytics blog has to offer in upcoming posts. As I work through the data the posts will start with broad topics and gradually become narrower in scope. If you want to discuss a post or contact me, Twitter is probably the best/easiest way to get a hold of me (@carouselytics).

Steve