Sunday Statistical Test: How Much Better Can NBA Offense Get?

The NBA continues to set records for offensive efficiency even as pace continues to quicken.

Offensive Rating has risen steadily since the 2012 lockout season, from 104.6 in that ill-fated 66-game crammed session that produced the worst offense since 2004, setting all-time highs in 2017 (108.8), 2019 (110.4), and 2020 (110.6).

Likewise, league pace has gone from 91.3 in that lockout season eight years ago to 100.3 in 2020, the fastest the league has played since 1989 (100.6, the last time before the 2018-19 season that league pace exceeded 100 possessions per 48 minutes.)

The fastest the league has ever played since pace as we know it was able to be measured was the very first year such a thing was possible; league pace was 107.8 in 1973-74.

The fastest it’s been in the 3-point era was 103.1, first in the first 3-point year of 1979-80 and then again in 1982-83.

All of this has contributed to a raw offensive explosion of 111.8 points per team per NBA game in 2020, which is the most points seen in American professional basketball in decades; the last time the NBA averaged 111.8 or more a game was the 1970-71 season (112.4), while the record for most points per game as league average happened way back in 1961-62, when Wilt Chamberlain scored 100 points in a game by himself and the league averaged a still-standing record of 118.8 points per game.

One can only wonder just how fast they were playing back then, because you have to take an awful lot of 2-point shots to score 119 when league FG% is just 42.6; indeed, there were 107.7 FGA per game and 37.1 FTA, marks that are third and second, respectively, in NBA history, just one year after the league set records in those two categories (1961 featured 109.4 shot attempts and 37.4 charity tosses per team per game for a scoring mark of 118.1 points as teams were less efficient than they would be the following year.)

All of this seems to suggest that we are approaching a sort of theoretical maximum absent a rule change (consider that in 1961, you had “three-to-make-two” from the line, explaining all those extra free throws, for example.)

But can modern, 21st-century, pace-and-space-style basketball reach and even exceed the lofty heights of a scoring explosion that would have seemed like a pipe dream and unbreakable record even five years ago (when the NBA averaged an even 100 points per game in 2014-15)? Will threes-and-layups ball ever combine with ball movement and spacing to produce a game so efficient that even as it never reaches the insane levels of pace achieved in those heady days of Wilt and Bill Russell and Oscar Robertson, it makes up for it on sheer ability to put the ball in the basket often enough to make up for it?

And if we do break those records for scoring, just how high can this lawn chair go if we tie enough balloons to it?

Let’s do some back of the envelope math and try to achieve the limits of the plausible in 2020 and beyond and see, indeed, just what we can get out of this old horse.

Let’s start with a fairly obvious trend that has been evident to anyone with a functioning pair of eyes and the ability to watch a Houston Rockets game any time in the past seven years or so.

In 2018, the Rockets became the first team in NBA history to shoot more three-point shots than two point shots, posting a .502 3PAR. They followed it up with .519 in 2019 and .501 this year.

At the same time, the Rockets have been practically allergic to long twos; only 3.6 percent of their attempts from the field came between 16 feet and the arc in 2020, and that’s up from 2.5 percent in 2019, a team that had Chris Paul on it. CP3 has always loved the midrange shot and even he only shot 6.8 percent of his attempts from that “worst shot in basketball” distance as Mike D’Antoni reined him in.

Even crazier is the fact that Paul made 55.1 percent of his shots from that range, seemingly statistically impossible and suggesting that Houston’s spacing and the defense’s tendency to leave the midrange as no-man’s land made most of those shots wide open looks.

It’s kind of a moot point insofar as 55.1 percent on long twos is the same eFG% as 36.7 percent from three, so if we take the next step and assume we can call Houston’s .519 3PAR that year the theoretical limit of just how many shots a team can take from long range before hitting the wall of (a) some shots are layups and dunks by design and (b) the defense won’t let you take every shot from the perimeter if they know what’s good for them, let’s not only assume that this is representative of what a team can do now but take it a step further and say that with more and more players coming out of college and the G-League with the 3-pointer trained into them during their formative years, the NBA as a whole will soon be able to sustain that level of 3-point shooting. There won’t be cases of guys like Russell Westbrook or DeMar DeRozan, who can’t hit a 3 for beans, or CP3, who is a decent 3-point shooter but whose game developed in a 2-point era. Those players won’t make it into the draft.

Next, we look at layups and dunks, or their proxy, shots between 0-3 feet in Basketball Reference’s database.

The Chicago Bulls led the league in percent of their shots taken from that close, at 34.0 percent, with the Nets and Lakers tied behind them all the way down at 31.5.

Meanwhile, shots from 3-10 feet, a proxy for aborted drives, good rim protection, and shots with the shot clock running out, comprised 16.5 percent of shots leaguewide with the Rockets at the bottom at 12.6, giving us a theoretical floor for such inefficient shooting close to the basket.

And it is inefficient; shooting percentages back up the notion that nobody shoots from 3 to 10 feet if they have a better shot, as field goal percentage league-wide was lower from that distance (39.6 percent) than from any other range (0-3, 10-16, 16-3pt, and 3-pointers themselves) on the floor.

So far, we’ve seen that we can minimize inefficient shots from the short midrange and the long midrange, but what about the, well, mid midrange?

A total of 9.3 percent of shots league-wide in 2020 came between 10 and 16 feet, the staple shot of the Michael Jordan era and anathema to good offense today, because it’s a sign of neither an aborted drive nor a poor offensive design that can’t space the floor or doesn’t have the shooters from long range to post a high 3PAR (not for nothing were the Spurs, Warriors, Knicks, Pacers, and Wizards the top five teams in percent of shots that were long twos; four of them didn’t make the playoffs and Indiana got swept in the first round.)

Of course, the Rockets were dead last in percent of shots taken from 10-16, at just 4.0 percent. When you have a team whose two options are “attack the rim” or “shoot a three”, it should come as no surprise that the team will take almost none of its shots from other ranges if it can help it.

If we assume that Houston’s 20 percent of shots between 3 feet and the arc constitutes the minimum amount that an NBA defense is willing to allow, we now have all the ingredients to calculate our maximum score.

After all, we have shot distribution (50 percent 3-pointers, 30 percent layups and dunks, and 20 percent “other”), we have a rough expectation of how much those shots are worth (36 percent for 3-pointers, two-thirds of the close shots, and about 40 percent from the midrange), and we can then insert the final piece of the puzzle, which is free throw shooting.

The NBA as a whole shot 23.1 free throws per team per game in 2020, and adjusted for pace, free throws per 100 possessions have been trending downward for years (more 3-pointers means fewer fouls because nobody wants to foul a 3-point shooter and send him to the line for three free throws.)

But we’ve seen that stat settle a bit in recent years; the real variable here is how impressive free throw shooting has become. In 2020, teams made 77.3 percent of their free throws, a record, and the past four seasons have seen the second, sixth, seventh, and first-place showings in league-wide FT% in the 74 years that stat has been kept going back to the genesis of the BAA way back in 1946.

The final piece in this statistical breakdown is to look at shot attempts and free throw attempts per 100 possessions, since we’re going to have to adjust this for pace.

2018-19 was the high-water mark, because that was the year the NBA had its lowest-ever turnover rate; a possession, after all, can have just three true outcomes (to borrow a term from baseball), a field goal attempt (with or without an and-one), a trip to the line without a shot attempt, and a turnover.

Happily, league pace was exactly 100.0 that year, so 88.6 field goals and 22.9 free throws can serve as our benchmark.

So that’s 44.3 3-pointers at a 36 percent clip. 26.6 shots inside the restricted area, at a make rate of two-thirds. 17.7 midrange jump shots/floaters/bad ideas at a 40 percent clip. And 22.9 free throws, of which teams will (let’s be charitable and assume a bit more room for improvement) make 78 percent of them.

Per 100 possessions, that’s 47.8 points from three, 35.5 points from in close, and 14.2 points from the midrange, plus 17.9 points from the line.

Add it all up and our theoretical maximum league-average Offensive Rating?

It’s 115.4.

And since that’s an average, it’s not like teams with exceptionally good offenses (like, say, the Dallas Mavericks, who posted 116.7 this season) can’t get even better since they’re already exceeding that theoretical average maximum.

Which in turn means that to break the record for raw points per game in a season league-wide, our pace will have to increase to 102.9.

Considering that it’s not out of the question to imagine a league playing that fast—we did it as recently as 1983 and the league’s been trending in that direction for a decade—we could see an NBA with scoring marks on a nightly basis not seen since Oscar Robertson and Wilt Chamberlain were in their primes.

But maybe more impressively than that, we may see the truly impressive offenses—the Luka Doncic-led Mavs or whoever Mike D’Antoni ends up coaching next—posting Offensive Ratings up over 120.

Hold on to your butts, folks. This is going to be fun.