（2级）Nylon Calculus: Does scoring 20 points per game mean anything anymore?

Last week, The Step Back released the third annual edition of their 25-under-25 list, ranking the best players in the NBA under the age of 25. This kind of project is designed to inspire debate and, as expected, we heard plenty of criticism for our evaluations. Of all the feedback we heard, the most frequent was that we, collectively, were underrating D’Angelo Russell (No. 16) and Zach LaVine (No. 33).

Both players are talented young guards who averaged 20 points per game for the first time last season, and both are players who derive the vast majority of their overall impact from scoring. If points are a statistic that still resonates strongly with you, then it’s understandable to be confused about players like Jaren Jackson Jr. (13.8 points per game) and Myles Turner (13.3 points per game) ranked ahead of them.

The fascination of sports fans with round numbers is well documented and averaging 20 points per game is a threshold that continues to equate, for many fans, with star-level play. However, this past year there were nearly twice as many 20-point scorers as five years ago and the number of players at that threshold has been increasing steadily for the better part of a decade.

So, in today’s NBA, what does it mean for a player to average 20 points per game for a season? And when we see a player over that threshold, what can we infer about their overall impact?

For the purposes of answering this question, we’re going to use ESPN’s Real Plus-Minus (RPM) as a measure of overall impact. RPM is an estimate of a player’s impact per 100 possessions that uses both box score statistics and on-off data. An RPM of 0.0 represents an average level of performance and it is, as Seth Partnow is fond of saying, the worst measure of overall value…except for all the others.

The chart below shows every player season since 2013-14 (the first season for which RPM is publicly available) with enough minutes to qualify for the minutes per game leaderboard. Each player season is marked by its Offensive RPM. On the left are player seasons with an average of at least 20 points per game, on the right are player seasons with a scoring average under 20.

Clearly scoring at least 20 points per game is not a prerequisite for being a positive offensive player although it is a fairly good indicator. DeMarcus Cousins in 2013-14 is the lone negative ORPM in our sample of players who scored at least 20 points per game.

However, if we parse those scoring averages by player season with a true shooting percentage above and below 54.7 percent (overall league average for the six seasons in our sample) we can see a lot more separation.

Here we can see that there is a fairly hard cap around an ORPM of plus-4.0 points per 100 possessions for players who don’t have at least an average true shooting percentage. The one exception, hovering up there in elite territory, is Russell Westbrook in 2014-15.

It’s clear that scoring at least 20 points per game correlates strongly with offensive value, however, that’s only half the game. Of the 140 player seasons with a scoring average of at least 20 points per game, exactly half had a negative Defensive Real Plus-Minus, in many cases erasing or nearly erasing the positive impact of their offense.

The graph below is the same as the ones we’ve seen above, but here, each player season is marked by overall RPM, instead of just offensive value.

Adding an estimate of defensive impact into the mix takes us from just one 20-point scorer in the negative to 12, including Zach LaVine last season (-0.49 RPM), and Andrew Wiggins (-1.6) and Devin Booker (-1.3) in 2016-17.

Another interesting point of comparison here is D’Angelo Russell and Dewayne Dedmon last season. Russell averaged 21.1 points per game, Dedmon averaged 10.8. However, by the estimation of RPM, their overall impact was almost identical because the gap between each player offensively (plus-2.1 for Russell, minus-0.45 for Dedmon) was the opposite defensively (minus-0.57 for Russell, plus-2.0 for Dedmon).

So, if we’re to continue using 20 points per game as a mental benchmark it should probably be qualified. Someone who scores at least 20 points per game is almost certainly a positive offensive performer. But if they aren’t efficient, they’re probably not in the upper bounds of positive territory. And if their defensive impact is lousy, it doesn’t really matter what they’re doing on the offensive end.