Predicting the NBA’s ‘Empty Stats All-Stars’ of 2019-20

In a shock to no one, not all stats are created equal.

One player’s points, rebounds or assists are not the same as another’s. Yet, there are a few reasons some guys’ stats are more empty than others.

The first is that some stats are fungible.  An average game is about 100 possessions on offense and 100 possessions on defense. That might vary from year to year and team to team, but it’s going to be in that neighborhood. In those possessions, there are about 90 shots, 45 rebounds and 25 assists.

Whether you’re on a stacked team or a bad team, someone is going to have to use those possessions, grab those boards and pass the ball, so some of the stats generated just happen because Player X happens to be on Team Y.

Swap him out with someone else, and you get the same thing. Those are fungible stats.

Then, some stats are empty by design: A team might want its point guard to take uncontested defensive rebounds because he’s a great outlet passer and quick with the ball in his hands, so he thusly generates lots of fast-break points. That has the impact of driving up his rebound numbers.

Finally, sometimes players (gasp) pad their numbers. And as long as contracts have incentives for reaching some numbers, it’s hard to blame them if they do.

With those three things in mind, let’s consider who this year’s “Empty Stats All-Stars” might be for some of the top box score categories:

assists: Tomas Satoransky, Chicago Bulls

Mar 31, 2019; Denver, CO, USA; Denver Nuggets guard Jamal Murray (27) and center Nikola Jokic (15) guard Washington Wizards guard Tomas Satoransky (31) in the third quarter at the Pepsi Center. Mandatory Credit: Isaiah J. Downing-USA TODAY Sports

There are two types of point guards: Race cars and race car drivers. (Sorry for the mixed metaphors, but bear with me for a moment.)

A race car is a guy like Russell Westbrook, James Harden or Stephen Curry. They are the guys who can not only run the offense, they are a large part of it. They can both score and pass.

Then there are guys who drive the car, and though that’s not necessarily worse, they are dependent on who they have around them. Prime Rajon Rondo is a great example of a driver. Give him a talented team, and he could facilitate it and make it better. But he’s not going to be scoring huge points in the process.

Tomas Satoransky is such a guy. He’s a great pickup for a Chicago Bulls team that has a lot of shot makers on the roster. Satoransky averaged 6.6 dimes per 36 minutes for the Washington Wizards last year, according to Basketball-Reference, but he should do even better than that in Chicago. He’s a precise passer with a very solid assist-to-turnover ratio of 3.33. He’ll get the ball to his shooters.

In 15 games with the Bulls last year, Otto Porter had an 83.3 effective field-goal percentage on catch-and-shoots, per Ryan Arcidiacono’s was 62.8. Zach LaVine (54.8) and Lauri Markkanen (54.2) were quite solid as well.

Look for the new starting point to rack up big assist numbers, but a lot of that will be more about the car than the driver.

Points: Devin Booker, Phoenix Suns

Mar 25, 2019; Salt Lake City, UT, USA; Phoenix Suns guard Devin Booker (1) shoots the ball against Utah Jazz guard Donovan Mitchell (45) during the second half at Vivint Smart Home Arena. Mandatory Credit: Russ Isabella-USA TODAY Sports

Devin Booker is a bonafide scorer, and he’s getting better. Look how well he compares through the first four years of his career with one of the greatest scorers in history.

Per 100 Possessions
Rk Player From To G FG FGA FG% 3P 3PA 3P% 2P 2PA 2P% FT FTA FT% TRB AST PTS
1 Devin Booker 2016 2019 272 10.8 24.7 .437 2.8 8.0 .354 8.0 16.7 .477 6.8 7.9 .854 5.1 6.2 31.3
2 Kobe Bryant 1997 2000 266 10.1 22.5 .450 1.4 4.1 .331 8.8 18.4 .476 7.3 9.0 .816 7.2 5.5 29.0
Provided by View Original Table
Generated 9/28/2019.

I say this to point out that his inclusion on this list isn’t meant to be a slight. But similar to the Black Mamba during the immediate post-Shaquille O’Neal years, Booker is in a position where he’s the only one on his team that can put the ball on the floor and in the hole from deep (no disrespect to DeAndre Ayton and Kelly Oubre, both of whom can do one or the other and are improving.)

Booker is the only realistic perimeter scorer on the entire roster. Last season, he made 375 unassisted buckets—almost 40 percent of what the current group posted. And a good chunk of the others were from inside by Ayton and Oubre.

That kind of near solo act tends to drive up the points.

Race car driver Ricky Rubio will be the starting point guard, but while he might be great for helping create shots for others, he’s not so great at making them himself. As skewed as the offense was to Booker last year, it’s likely to be the same or even more so this year.

At least Booker theoretically won’t have to work so hard to get his points, which should mean there are even more on his nightly box score.

rebounds: Mason Plumlee, Denver Nuggets

Apr 27, 2019; Denver, CO, USA; Denver Nuggets forward Mason Plumlee (24) pulls in a rebound away from San Antonio Spurs center Jakob Poeltl (25) in the first half in game seven of the first round of the 2019 NBA Playoffs at the Pepsi Center. Mandatory Credit: Ron Chenoy-USA TODAY Sports

Rebounds are a stat that can be incredibly misleading—perhaps more than any other. Seeing someone grabbed a board doesn’t tell you anything about the context of it.

Was someone else going for it? Did they have a teammate blocking out an opponent? Did a teammate defer the rebound to them?

The tracking stats at help with this, but even those numbers need to be taken with some salt. Just having a lot of deferred rebounds doesn’t mean you’re stat-padding. Some of that can be attributed to the system.

However, there is a work-around to get a semblance of honesty to that. By taking the deferred rebounds and uncontested rebounds out of the total rebound chances, we can get the number of times the player actually fought opponents for the rebound.  Then we can use  contested rebounds/contested rebound chances to come up with a kind of “rebound winning percentage.”

Those at the top of the ladder are no surprise. Here’s a look at the top 10 winning percentages among the top 50 in chances:

Player Team Contested Rebounds Rebound Chances Uncontested Rebounds Contested Rebound Chances Percent of Battles won
Andre Drummond DET 597 1919 635 1284 46.5%
Anthony Davis NOP 307 1057 365 692 44.4%
Hassan Whiteside MIA 394 1318 423 895 44.0%
Nikola Vucevic ORL 402 1467 529 938 42.9%
Clint Capela HOU 376 1391 472 919 40.9%
Nikola Jokic DEN 403 1463 462 1001 40.3%
Karl-Anthony Towns MIN 405 1562 549 1013 40.0%
Deandre Ayton PHX 318 1213 411 802 39.7%
Derrick Favors UTA 270 968 282 686 39.4%
Aaron Gordon ORL 190 862 377 485 39.2%

And then, there is the other side of the equation. Here are the bottom 10.

Player Team Contested Rebounds Rebound Chances Uncontested Rebounds Contested Rebound Chances Percent of Battles won
Russell Westbrook OKC 131 1198 666 478 27.4%
James Harden HOU 123 903 395 443 27.8%
Paul George OKC 147 1070 475 510 28.8%
Tobias Harris PHI 149 1092 496 502 29.7%
Mason Plumlee DEN 205 1057 319 664 30.9%
PJ Tucker HOU 182 954 297 580 31.4%
Dario Saric MIN 193 922 264 594 32.5%
Thaddeus Young IND 212 1020 311 647 32.8%
Taj Gibson MIN 200 853 258 595 33.6%
Montrezl Harrell LAC 268 1060 267 793 33.8%

It’s not shocking that Harden and Westbrook are at the bottom of the list, but they’ve swapped teams now. And like most of the names on this list, a good chunk of their boards are “system” boards where they are the ones whom the ball is deferred to because they initiate the offense.

The name that jumps out on this list is Mason Plumlee, the one real center. He averaged 10.9 boards last year, but this list suggests that his are much “cheaper” than guys like Andre Drummond or Anthony Davis who are bigs snatching the misses they vie for.

He’s 6’11” and 250 pounds, so he’s not easily out-muscled. He has a 36″ vertical which is impressive for a man his size. And his numbers haven’t varied much in spite of having played with three different teams and a plethora of teammates and systems. So at a certain point, you have to think the trend is definitely due to Plumlee.

There are two things which might help explain it. First, his 6’11” wingspan is a bit on the T-rex size when you compare him with some of the other bigs. Drummond and Davis both measure out at  7’6″, for example. A lot of rebounding isn’t just about height, it’s about reach and being able to extend to where the ball careens.

The other thing is positioning. Plumlee doesn’t always anticipate well where the ball is going to go, and he’s a bit passive in trying to position himself. Already at a disadvantage reach-wise, that compounds the problem.

So, yes, by the box score, he’s grabbing a respectable number of boards. It’s just that they happen to be easier ones than some of his similar-stat counterparts.

Steals: Chris paul, oklahoma city thunder

Apr 7, 2019; Houston, TX, USA; Houston Rockets guard Chris Paul (3) dribbles the ball during the first quarter against the Phoenix Suns at Toyota Center. Mandatory Credit: Troy Taormina-USA TODAY Sports

The steal is one of the most devastating plays in basketball because it not only ends one possession, it makes it much easier for an opponent to score.

Some people will decree that the cost of a missed steal negates the benefits of it, but that’s overstating the cost of a failed steal attempt. Let’s say a player only gets the steal half the times he shoots passing lane, and that the opponents get a layup every single time that happens. Their offensive rating would still be only 100.0, well below the league average. Ergo, the cost of the failed steal isn’t that great.

That’s also not taking into consideration the relative impact of the increased chance of scoring after the steal. The league-wide offensive rating after a team gets a steal was 128.3 last season based on data from PBP Stats. The Sacramento Kings led the NBA at 137.1.  By comparison, the average points on a possession on plays where there was no steal was 1.09—meaning that the steal increases a team’s points on the following possession by an average of  0.19 points.

Therefore, to investigate the relative impact of steals, I looked at what happened on the possession following the steal to get a bead on how many points the player added to the team through said steals. Using PBP Stats, I gathered the possession and points info on each player, used that to determine the offensive rating when that player gets a steal compared to the team’s normal offensive rating. Then I used the net gain (or net loss) to evaluate the relative value of a player’s steals:

Player Steals/Game Steals Points STL ORtg Team Off RTG Point Differnetial Points Added
James Harden 2 155 225 145.2 115.5 29.7 46.0
Mikal Bridges 1.6 128 178 139.1 105.9 33.2 42.4
Marcus Smart 1.8 139 197 141.7 112.2 29.5 41.0
De’Aaron Fox 1.6 131 184 140.5 110.4 30.1 39.4
Jrue Holiday 1.6 106 157 148.1 111.4 36.7 38.9
Russell Westbrook 1.9 139 190 136.7 110.3 26.4 36.7
Kawhi Leonard 1.8 105 154 146.7 113.1 33.6 35.2
PJ Tucker 1.6 131 176 134.4 115.5 18.9 24.7
Andre Drummond 1.7 131 157 119.8 109.0 10.8 14.2
Jimmy Butler 1.9 121 148 122.3 112.6 9.7 11.8
Paul George 2.2 162 188 116.0 110.3 5.7 9.3
Chris Paul 2 112 129 115.2 115.5 -0.3 -0.4

Based on this, it’s not hard to determine whose steals have the least impact. In fact, Chris Paul’s had a negative impact on the ensuing offense.

The reason steals are so impactful on the next possession is that you’re out in front of the defense and, because of that, you’re able to get uncontested shots at the rim or from the corner 3. However, for that to happen, you have to push the ball up the court. Paul’s tendency to slow down is probably what negates the advantage.

While the effect on the defense is still there, a Chris Paul steal isn’t really any different than any other stop. So while he might be racking up more steals this year, they won’t help the team as much as other leaders on this list.

Blocks: Jaren Jackson Jr., Memphis Grizzlies

Jan 23, 2019; Memphis, TN, USA; Charlotte Hornets guard Jeremy Lamb (3) shoots against Memphis Grizzlies forward Jaren Jackson Jr. (13) during the first half at FedExForum. Mandatory Credit: Justin Ford-USA TODAY Sports

I won’t be the first person to make the observation that the benefit of a block that goes out of bounds or to the opponent isn’t that great since the other team still has the ball and they still have a chance to score.

So how do you determine the relative value of a block?

To start with, the block of a 3-point shot is greater than the value of a 2-point shot for the same reasons that a 3-point make is greater than that of a 2-point make. But not every block is a shot that was going to go in.

So I looked at the top-30 players in blocks from PBP Stats and credited the blocker with the appropriate amount of points based on what type of block it was. However since not every shot was going in, I multiplied by that the league-wide effective field-goal percentage (52) to estimate the gross number of points a player saved with his blocks. Myles Turner led the league with 208 points.

Then we have to determine the points that were “lost” because the blocker’s team did not recover the ball. For that, I used the average points per possession (the defensive rating when the player was on the court/100) of blocks recovered by the opponent. Turner also led here with 72.

Then I subtracted that number from the  gross number of points saved by a block to determine the net points saved. Again, Turner led here with 134. Rudy Gobert (129), Brook Lopez (118) and Mitchell Robinson (109) were the only other players to finish over 100.

Finally, I took net points saved/blocks to get net points saved per block to determine the actual value of a player’s blocks. Kevin Durant sat on top of that with 0.73 , with Gobert (0.69) second and Turner third (0.68).

Here are those among the top 30 whose blocks had the least value.

Name Team DRTG Blocks 2pt Blocks 3pt Blocks Blk Rec Blk Rec Opp Pts Saved Pts Lost Net Pts Saved Net Pts Saved/Block
Jaren Jackson Jr. MEM 108.8 82 79 3 37 45 87 48 38.7 0.47
Richaun Holmes PHX 115.1 79 78 1 41 38 83 43 39.6 0.50
Marc Gasol TOR 113.1 86 81 5 44 42 92 45 46.9 0.55
Montrezl Harrell LAC 111.5 110 108 2 61 49 115 55 60.3 0.55
Jusuf Nurkic POR 110.5 103 102 1 56 47 108 51 56.6 0.55
Steven Adams OKC 107 76 75 1 41 35 80 38 41.8 0.55
Maxi Kleber DAL 110.7 78 76 2 42 36 82 39 43.0 0.55
Joel Embiid PHI 112.6 122 119 3 66 56 128 60 68.1 0.56
LaMarcus Aldridge SAS 111.2 107 107 0 63 44 111 50 61.7 0.58
Clint Capela HOU 110.7 102 102 0 60 42 106 47 59.5 0.58

Jaren Jackson was the run-away leader here. To be fair, he was a rookie last year, and nuances like “blocking for your team rather than just the highlight” take time to learn.

The good news is that it already seems like Jackson doesn’t try to send the ball into the fifth row a lot. But he doesn’t really try to tip it to a teammate either, so it ends up being a 50/50 ball most of the time, which is why it’s not a surprise that his team gets the edge less than 50 percent of the time.

I expect him to improve in this area somewhat this year, but I think he’ll also slip into old habits during his early seasons. Yet, I also think his total block numbers will go up as he has great instincts. It’s just that all those blocks won’t have the same value as some of the other stat leaders until he masters a few more tricks of the trade.