Evolution of the NBA offence by geography

The NBA in its early days was defined by those who could dominate near the basket. It's no mistake that the league's early stars looked more like Wilt and Russell and their bruising interior game. But like many products, the NBA evolved through a combination of changes to its rules and decades of competition.

The resulting league in 2020 is almost unrecognisable. Today, point guards like Steph Curry are arguably more critical to a team’s success than the bruising big men who dominated the past. Even the best big men like Joel Embiid or Giannis Antetokounmpo are encouraged to develop their range and roam the perimeter to "stretch the floor".

I would say that the changes have been immeasurable, but that would be an extremely poor metaphor to choose in describing modern basketball. The game has in fact never been more measurable than it is today, where the analytics movement and advances in AI have fed off each other to generate mountainous amounts of data and analysis. Here, I take a look at a few ways in which the league has changed, focussing on the end results: the shots.

The story starts here, back in 2005/2006.

Shot chart — '05/'06 season (Image: Author)

In case you haven't seen charts like these, the size of each hexagon shows show often players shot from this zone, and the colours denote the typical points per shot (accuracy * 2 or 3 points, based on the zone).

The highest-efficiency shots in 2005 were from near the basket, and at the corners. After that, the shots from above the arc had the most value, followed by the deep, dark, depressing oceans of the long twos flooding the space between the rim and the three-point line with its unending void of inefficiency.

For each shot at the rim, you'd expect 1.2 points per shot (60%), and three-pointer would net you between 1.05 points (35%) to 1.2 points (40%) on average, depending on the spot. A mid-range two-pointer would net you a measly 0.8 points (40%) on average, meaning that you're losing somewhere between 0.25 to 0.4 points per shot. Over the course of a game, that might lead to something like 20 to 35 fewer points for the team.

Since then, the recognition of this has changed the face of league and the inefficiencies have been significantly culled from the playbook.

Shot charts — '05/'06 vs '19/'20 seasons (Image: Author)

These comparisons tell more than a few stories.

The inefficient two-pointers that were previously filling the middle of the court in blue have all but disappeared, and the yellow/orange markers now flood the outside of the three-point line like Outriders pushing against the energy barrier at the outskirts of Wakanda.

This change has been hardly the result of a revolution, but something that has happened steadily over this last decade. Take a look at this chart, simply breaking down the percentage of shots taken to twos and threes:

Three-point frequency evolution (Image: Author)

Since 2011-12, the percentage of threes taken by players have increased by 2%. This might not sound like much, but from a baseline of about 20% that's a relative 10% increase every year. In 2020, the number has almost doubled such that almost 40% of all shots taken are three-pointers. And the numbers don't show any signs of slowing down, or plateauing.

Regardless of your feelings about how that changes the game (and there is definitely a lot of that going around), there is no doubt that this is the winning formula for most. The advantages conferred by stepping outside the arc on a more regular basis is significant. The advantages comes not only from the simple maths of the extra 50% of points scored per shot, but the fact that forcing defences to pay more attention beyond the arc opens up the rim.

In the side-by-side comparison of shot charts above, you might also have noticed that the area around the rim look darker. This is not an optical illusion. Dividing the court into zones and plotting the accuracies for each zone as below tells a clear story:

Shot accuracies, '15/'16 vs '19/'20 seasons (Image: Author)

The accuracies at the rim and those on long three-pointers have dramatically increased while the others stayed near constant.

Why? I would hazard a guess that these increases are for two different reasons: one, players are simply practicing shooting from deep (27+ feet) more, which was once viewed as, as Paul George would say, a bad shot. That's just not true. In the modern NBA, the expected value per shot is just a shade under 1.05 points. So, if you can make a three-pointer at 35%, that's slightly better than an average shot. For the elite outside shooters like Steph Curry or Lillard, it's a good shot.

And the increase to the efficiency at the rim is a designed byproduct, and a recognition that this is the other high-efficiency area. The dramatic increases in number of three-pointers attempted means that simply, defences have to cover more ground or at least be prepared to cover ground, which prevents them from crowding the hoop. As a result, better shots are able to be taken at the rim by the cutters & drivers, and scoring is made easier for elite post offensive players like Embiid.

The result is an NBA where the area within 4 feet of the basket, and a strip of the court about 3 feet wide just outside of the three-point line are where approximately three quarters of points are generated. The next chart shows this evolution graphically.

Evolution of scoring by zone (Image: Author)

Once again, the increase in the reliance of threes show no signs of abating. In fact, the chart indicates that we may be at a cusp of another evolution where the deep three-pointers really become a regular part of teams' arsenal. As more emphasis is placed on finding great shooters and existing players spend more time expanding their range, it has become not too unusual to see players take shots from well beyond the three-point arc.

The next chart plots all of this evolution on the court as an animation. You'll notice that far more markers appear outside of the three-point line as well as becoming larger over time, while the reverse happens in the mid-range just inside of the three-point line.

The NBA shot (r)evolution (Image: Author)

At this rate, it's not inconceivable that in a few years, another row of the yellow/orange markers will have appeared behind the three-point arc.

We often talk about the “revolutionary” changes to NBA teams’ offence, shooting more threes than ever before. The data above makes it clear that in fact, it would be more accurate to call it an evolution.

It's not quite clear how far this trend will continue, and where an equilibrium will be reached. There may reach a point where defensives simply ignore players unless they're outside the arc, or near the rim, especially for teams who steadfastly refuse to shoot the ball outside of these areas like the 2016-2020 Houston Rockets.

Houston Rockets' shots between 2016-2020 seasons (Image: Author)

It follows that the players must be able to make enough "bad" shots to force the defence to adjust to them. But we are not quite at that point yet. Also, not all teams have committed to this philosophy. Just take a look at these shots taken by the Knicks in the same span.

New York Knicks' shots between 2016-2020 seasons (Image: Author)

Not everyone has bought into, or possibly doesn't have the personnel to execute, the new game.

And small divergences to the extreme strategy such as those adopted by the Rockets' are also being seen by some teams. Take a look at the Spurs' shot chart here.

San Antonio Spurs' shots between 2016-2020 seasons (Image: Author)

The Spurs take far, far more midrange shots than others in the modern game. But, they also make them at higher percentages than the rest of the league - both from beyond the arc, and in the two-point range. It's arguable what the right mix of twos and threes are, but the Spurs probably benefit from not being as predictable as teams like the Rockets, and also from not turning down open two-pointers which are reasonable shots.

In other words, we are a long way from an "optimal" strategy and convergence, and there's a lot more to come as strategies and players evolve. It's going to be exciting to see what comes next as the game continues to evolve.


If you liked this - follow me on Twitter! I tweet about data, visualisations & sports - and occasionally also do freelance work on those.

JP Hwang

JP Hwang

Data, numbers and analytics geek. Moreyball enthusiast.