How Spatial Analytics Killed The Mid-Range Jump Shot

This article is presented in partnership with Draftstars

The NBA introduced the 3 point line at the start of the 1979-80 season, borrowing the feature from the absorbed American Basketball Association. An average game that season featured 90.6 field goal attempts, but just 2.8 of those shots were 3 point attempts. It was widely regarded as a gimmick. "It may change our game at the end of the quarters," then-Phoenix Suns Coach John MacLeod told The New York Times. "But I'm not going to set up plays for guys to bomb from 23 feet. I think that's very boring basketball."

Fast forward to the 2015-16 Finals. The Golden State Warriors, who lead the league in 3 point attempts (31.6 per game), are matched up against the Cleveland Cavaliers who rank 3rd (29.6 per game) in that category. League MVP Steph Curry made 402 3 point shots in 79 games during the regular season. That's 5.1 made 3s every game. To put that into perspective, the first time a team broke 400 3s in a season was the Houston Rockets in 1994, 15 years after the 3 point line was introduced. Curry alone was making as many 3s as the average NBA team did in 2004.

Curry is an exceptional player. But he simply represents the tip of the spear when it comes to the rise of the 3 point shot in the league. What changed? The fact that 3 is greater than 2 was true in 1979. Put simply, the last few years have yielded the technology and analytics to make that simple truth undeniable. The era of big data basketball arrived.

In the NBA, SportVU has taken spatial analytics from the backrooms of MIT to the front office of every NBA franchise. The technology, originally developed for the Israeli Defence Force for real-time battlefield monitoring and missile tracking, is comprised of six cameras strategically positioned to capture every ball and player movement that takes place on the court by taking a snapshot of the court 25 times a second. At the start of 2010/11 NBA season, just four teams—San Antonio, Houston, Dallas and Boston, all teams well known for their analytics departments—had installed SportVU in their rafters. By the start of the 2013/14 season, all 29 arenas had the system in place.

Here's a visual representation of the SportVU data taken from a Knicks Raptors game in 2013. The technology not only tracks individual player and ball movements, but also keeps track of the action being run within the play (two side pick and rolls, both kicked out to the three point line), as well as the player and area each player is responsible for defensively within the play. When Jason Kidd takes the 3 point shot, the system records the time left on the shot clock, the distance from the rim, and proximity and acceleration of DeMar DeRozan attempting to close out when the shot is taken, amongst dozens of other data points.

Obviously, this is just one 15 second stretch from the start of a game. But the applications and conclusions that can be reached from a play like this, integrated into an entire season's worth of plays, are innumerable. The data creates opportunities to quantifiably measure things we could only theorise on before, like how effective Carmelo Anthony is at setting up his teammates for open shots, the value of Tyson Chandler's picks in order to create space for his teammates, and how much of an impact DeMar DeRozan has closing out on 3 point shooting.

Perhaps most impressively, the video shows blank circles around the Toronto defenders. These 'phantom' players represent what the system believes would be the ideal defensive positioning and movement for each of the players in order to create the lowest possible percentage shot for their opponent. This optimal positioning estimation is based on over 140,000 plays recorded by the Raptors alone. Today, teams are able to make such estimations with three more entire season's worth of data to rely on.

Over the past few seasons, the revelations from the data gathered by SportVU have led to some profound breakthroughs in understanding and properly assessing individual player value and optimal team strategy. The 2014/15 Houston Rockets under GM Daryl Morey are perhaps the poster child of this new era of analytical basketball, running an offence to maximise the potential of their roster that consisted almost exclusively of 3 point shots and shots at the rim; just 6.2% of their points came from midrange shots. SportVU also helped to better quantify a player like Dwight Howard's defensive value to the team as a rim protector. While someone like Tim Duncan's presence in the paint decreased the efficiency of opposition shooters from close range, the new dataset shows that Howard's presence both lowered opponent efficiency while dramatically reducing the average number of shots attempted by opponents in the paint. Remember that saying about missing 100% of the shots you don't take?

The effects haven't been limited to traditionally difficult to determine defensive stats either. The case for James Harden being worthy of MVP consideration that season could be argued through traditional stats; he was 2nd in scoring with 28.1 points a game and 8th in the league with 7.0 assists. But the insight from the advanced analytics allow us to better understand how important Harden's unique offensive game was to his team. He was able to get to the free throw line a league-best nine times a game that season. When you shoot 86.8% at the line, the value of a possession that ends in two free throws is 1.74 points. For reference, the NBA average possession was worth about 1.04 points that year. More importantly, SportVU data tells us that 84% of NBA 3 pointers are assisted. For a corner three, the easiest shot to make from deep, the rate jumps to 96%. It's one thing to want to shoot more 3s, it's another thing entirely to create more opportunities to shoot 3s. Harden's ability to create his own shots and collapsed defences allowed his teammates to space the floor and set up for the best 3 point shot opportunities. He allowed the Rockets to shoot a NBA record 2680 NBA 3s, more than 400 clear of 2nd place Cleveland's 2253 attempts that season.

In many ways, while the emergence of this dataset has led to many breakthroughs about how to optimise team strategy to get the maximum value out of every player on the floor, the data has also confirmed the importance of star players. Houston's ability to rain down 3 point shots by role-players wouldn't work without James Harden creating offence for these players. Detroit Pistons Coach and GM Stan Van Gundy pointed this essential fact out during a 2014 SSAC panel discussion. "I think that's the one thing that's sort of lost in the whole thing. Everyone's gotten into these generalisations that you need free throws, shots at the rim and 3s. That's all well and good, but if you don't have guys that can shoot the 3, that doesn't help you a lot," he said. If you watched a Philadelphia 76ers game over the last few years, you'll notice they attempt to play the same way the Rockets have; fast breaks, plenty of interior shots and as many 3s as possible. The difference is they have largely been a team of raw prospects and old vets, with no elite level players. Last season they went 10-72. Ben Simmons better start working on his outside shooting.

The other major flow-on effect from this shift is that now that everyone knows that on average a 3 point shot is better than a 2 point shot, shots in the no-mans-land outside the paint but within the arc have become a rare occurrence. The midrange shot is becoming such an endangered species that teams with high efficiency midrange shooters, like Dallas' Dirk Nowitzki, have almost have gained an advantage through the number of open shots available to them against 3 point orientated defensive structures. The market for players who can shoot the 3 pointer has also adjusted accordingly. Ryan Anderson, a 6'10 power forward who can stretch the floor with his outside shooting but offers little defensively received a 4 year, $80 million contract this offseason. Conversely, big men who can't offer elite rim protection or guard the perimeter are finding it hard to survive in the modern game.

No one is sure where the rising tide of the 3 pointer will peak. Some coaches and pundits wonder aloud if the epidemic of long range shots may be ruining basketball, and there are rumours that the NBA is considering introducing a 4 point line. One thing is clear: spatial analytics have changed the way coaches and players understand and play the game. From Steph to scrub, the 3 point shot has become the great weapon and the great equaliser in modern basketball. Elite teams expect every player on the court to be able to knock down open 3s, even those over 7 foot tall. It took the most advanced spatial tracking technology we've seen in sports to reveal a simple truth: 3 points are more than 2.

This article is presented in partnership with Draftstars