Over the past several years, Goldsberry has applied his mapmaking skills to basketball. It is not hyperbole to suggest that he has changed the way the game is played by figuring out from where, how often and how well players shoot on the court — and then visualizing, or rather mapping, the data.
If early analytics, which were most famously used in baseball, focused on the outcomes of static one-on-one encounters, Goldsberry expanded the genre, hoping to better understand spacing on the basketball court.
What does that mean in practical terms?
When Portland Trail Blazers guard Damian Lillard drained a 37-foot, series-clinching three-pointer last week, Goldsberry could confidently tell you (above the protestations of the Oklahoma City Thunder’s Paul George) that the shot, despite the distance, was in fact a good one.
“SprawlBall,” illustrated by Aaron Dana, is a visual tour through Goldsberry’s work and the modern NBA game. It has meditations on players such as LeBron James and Stephen Curry but also detailed data and charts on how, for instance, the three-point line has changed the game. (The average three-point shot yields 1.07 points vs. 0.8 points for the average midrange shot.)
At its heart, “SprawlBall” is a book of maps. It’s a geography book.
During his junior year at Penn State, Goldsberry took an introduction to cartography class on little more than a whim. “I remember the census data and this software [Graphic Information Systems] that basically links databases to maps,” he said. “It was this perfect balance of art and science, and I devoted the next 15 years of my life to it.”
He switched his major, got a cartography degree and then moved to Washington to make flood maps for the Federal Emergency Management Agency. After a stint working for a software mapping company in Maine, Goldsberry got his master’s and PhD at UC Santa Barbara, focused on the intersection of computer graphics data visualization and cartography.
Set on the path of academia, he took a job at Michigan State, where his research focused on mapping food deserts and public health facilities. In 2011, Goldsberry accepted a visiting professorship at Harvard, and with a lighter teaching load he suddenly had time to think about another love: basketball.
Goldsberry had played recreationally his entire life, and the spark of an idea had long flickered somewhere in his brain. “I knew that some players hung out on one side of the court — and in my own game, I’m 13 years old and I’m better in the right post than the left post," he said. “And no one had ever been able to quantify that.”
The issue, he realized, was a mapping question. Each shot had an X and a Y coordinate.
He scraped the Internet for shot chart information from ESPN box scores and a few other places and was able to cobble together data for more than a million shots taken in the NBA from 2006 to 2011. Then he mapped it, the first effort at such research.
Goldsberry presented a paper at the MIT Sloan Sports Analytics Conference in 2012 titled “CourtVision: New Visual and Spatial Analytics for the NBA.” For the first time, you could see how NBA players — and the league at large — performed from different locations on the court. The reaction from the league community was overwhelming, and it began an odyssey that took Goldsberry from writing for Grantland to working for the Spurs to USA Basketball and now ESPN.
Still, to his former geography colleagues, the origins are clear: “It’s spatial data; more shots from one location or a distribution across the court change with the color scale. It’s a different scale from mapping a county or a country or the world, but the principles are the same,” said Cindy Brewer, one of Goldsberry’s professors at Penn State.
Added Morris Thomas, a former colleague at Michigan State: “We’d go to basketball games in Detroit, and he’d say things to me like, ‘Don’t you notice this guy is always shooting from that same spot?’ I think the wheels were turning then.”
A quick aside here: Goldsberry will never be the most famous geographer in basketball because a guy named Michael Jordan majored in geography at North Carolina. Early in Jordan’s career, Thomas found himself in the Chicago Bulls’ visiting locker room in Detroit and struck up a conversation with Jordan by telling him that he, too, studied geography.
“Jordan kind of looks at me, surprised — I’m a black guy; not too many black guys major in geography — and he lights up a little bit,” Thomas said. “And he asks me, ‘How many counties you been to?’ Turned out he had a geography professor at North Carolina who was trying to visit all 3,141 counties in the United States.”
The last chapter of “SprawlBall” is about the aesthetic of the game, and in it Goldsberry considers a different lens for the utility of analytics. Since the 2004 publication of “Moneyball,” the seminal book about the Oakland Athletics by Michael Lewis, analytics has been thought of as a strategy for how teams can gain an advantage against each other.
And that evolution has created an analytics backlash among fans who don’t favor the defensive shift in baseball, for example, or the isolation-themed basketball of the Houston Rockets. Goldsberry is sympathetic, which he credits to his time working with the Spurs, Coach Gregg Popovich and General Manager R.C. Buford.
“I would say the analytics aesthetic right now isn’t optimal,” Goldsberry said. “The aesthetics that it’s helped create are threatening some of the fabric of the game, and the product isn’t as good.”
If basketball’s inventor, James Naismith, could engineer social values into basketball — the game was created to help Christian boys blow off steam in the winter — then analytics ought to be able to engineer a better game, and not just better teams.
“In our post-‘Moneyball’ landscape, I think analysts tend to look at the game and try to find market inefficiency and competitive advantage, but folks like Coach Pop and R.C. see the sport as something much less trivial,” Goldsberry said. “It’s a living, breathing masterpiece to be celebrated, stewarded and improved along the way.”
In the book, Goldsberry throws out a few thought experiments, including adjusting the three-point line, eliminating the shorter three-point shot from the corner or even allowing goaltending on three-point shots. The specifics, though, are almost less important than the idea of what analytics should be.
“If the basketball ideal is to have guys shoot 33 percent from three-point range, we can do that,” Goldsberry said. “That’s the frame we should be thinking about — how do we engineer a better basketball game?”
The way to do it is by manipulating the court, which is just another mapping problem.
Read more on the NBA: