Using Big Data to analyze NBA shooting
University professor uses five years of shooting data to create color-coded data map
Advocates of Big Data often tout the benefits of crunching huge data sets for businesses, scientists and public-policy researchers.
Now an assistant professor at Michigan State University has shown that Big Data can be applied to hoops.
Kirk Goldsberry, a geography professor at MSU, has analyzed five years' worth of scoring in the National Basketball Association in an effort to ... well, Kirk explains it in his blog called "CourtVision":
In the quest to better understand the "average" NBA shooter I have begun making composite shooting charts for each position in the league. My eventual goal is to establish a spatially informed baseline and to map every shooter in the league against an average shooter.
Hey, why not? Baseball for years has analyzed micro-data, a practice pioneered by statistician Bill James back in the late '70s. And professional and college football both analyze metrics that hadn't even existed (or couldn't be leveraged by technology) a couple of decades ago.
Granted, we're not talking about finding a cure for cancer, but here's an example of Big Data being used in an attempt to analyze a competitive situation and draw useful conclusions. Sounds like a business challenge to me.
Specifically, what Goldsberry has created is a color-coded map of a basketball court that shows points-per-shot-attempt as well as number of attempts from any conceivable position on the court in shooting range. His data includes more than 700,000 attempted field goals during seasons 2006 through 2011.
Here's Goldsberry's explanation from an academic paper he wrote on "New Visual and Spatial Analytics for the NBA":
CourtVision integrates database science, spatial analysis, and visualization to reveal players’ and teams’ unique spatial signatures; in turn, these signatures expose important patterns and anomalies in performance that are considerably less evident using conventional evaluative approaches.
I know, sounds sort of geeky. But don't worry, Goldsberry provides his own layman's translation: "The results facilitate efficient answers to important questions about the NBA such as: Where are players’ most common shot locations and how successful are they at these locations? Which point guard has the highest points per attempt at the top of the key? Which court locations are most or least effectively defended by the Orlando Magic?"
No doubt any number of NBA coaches and players might find that kind of data interesting. They also might find some of Goldsberry's writing somewhat entertaining. Here he is on the color-coded chart he created for Boston Celtics point guard Rajon Rondo, who has accuracy problems in many areas of the court: "There is a lot of blue and green on it; that's not good, those are the morose hues of a bricklayer."
As a color commentator, Goldsberry sounds like a natural. I fear he went into the wrong field.
(H/T to Gizmodo)