clock menu more-arrow no yes mobile

Filed under:

A Word On Statistical Analysis In College Basketball

This was going to be an asterisk leading to a footnote. Then things got out of hand.

In a perfect world gym's around the country would be wired from top to bottom transmitting and sorting data of what transpires on the floor in near real time thereby enabling a more perfect understanding the sport of basketball. We do not live in a perfect world. The NBA is getting closer, but college is far from it.

Every college program has some degree to statistical rigor embedded within that. The Bearcats are no exception to the rule, but the data they gather is rightfully proprietary. What the fans get is advanced statistics from Ken Pomeroy and the good folks of statsheet. All of which is invaluable if the goal is to reduce a facet of a player or a team's game to a single characteristic.

Is team X good at shooting? Player Y looks like a turnover machine, is he really? Is player Z really a good rebounder? There are ways to measure all of those by simply looking at stat sheet and applying a little math. A teams true shooting percentage is simply a weighted average of attempts from three, two and the foul line. Turnover rate measures how many possessions consumed by a player end in turnovers. Rebound percentage is the frequency with which a player rebounds a missed shot when they are on the floor.

All of those stats give meaning and shed some light on what is happening out on the floor, but they provide little in the way of context. The context is the key. There are simply too many games for a sportVU like system to take hold in the college context. But there is a way to get a rough approximation of what is happening on the floor over the course of a season by simply combing through the play by play in every box score.

That is what Jeff Haley of does. He sorts the many thousands of shots taken in the course of a college basketball season into three categories; at the rim, two point jumpers and three pointers. He also provides a percentage of baskets that are assisted in all three categories on a player by player basis.

So we can find out roughly where a given player or team is shooting, and how often a given player gets a shot created for them. Or conversely when looking at a team how often a team goes it alone in creating offense. It's not perfect, but it is the best thing going right now in this field.