Bob8 wrote:Bob8 wrote:GeorgeMarcus wrote:Looking at on-court and on-off numbers in tandem can be very informative, especially with context about lineups/rotations. The first number takes a player's total +/- and averages it out over 100 possessions; the 2nd number compares his team's production with him versus without him and averages it out over 100 possessions.
These data points generate 4 general categories that players can fall into:
Positive-Positive
When the first number is positive, we know the player's team outscored their opponents while he was on the court. Maybe the player was vital to that success or maybe he was riding the coattails of his teammates, but at a minimum the positive number demonstrates the player can be part of a winning recipe.
When the second number is positive, we know that the team's average +/- was better with the player than without him. "But that depends on lineups/rotations" you might say, and you would be right. It's possible the player is benefitting from a more favorable position in the rotation. That said, if a player is regarded as being the clear best on his team, there really isn't an excuse for that player to have a negative on/off. Especially knowing that every NBA team staggers their rotations.
In sum, 2 positives show us the player (a) won more possessions than he lost while on the court and (b) his team produced more efficiently when he was on the court than off the court. Maybe he lucked his way into the best lineups on a great team, but it's much more likely the player provides positive impact of his own. It's no coincidence the top 16 players from RealGM's top 25 project all fall into this category. As for the 17th player, he lands in our next category...
Negative-Positive
A negative-positive tells us the player's team was outscored while he was on the court, but that the team played better with him than without him. Enter #17 ranked Bradley Beal. The Wizards were really bad last year, so it should come as no surprise that Beal had a negative +/- on the season. It's still not the highest praise given that only 1 other player in the top 25 had a negative +/-, but at least he demonstrated positive impact on his own team. The same can't be said for our next example...
Positive-Negative
A positive-negative tells us the player's team outscored their opponents while he was on the court, but that they produced even more efficiently without him. The highest ranked player in this group would be Philly's own, Ben Simmons at #22. As the 3rd best player on last year's Sixers, many of Ben's minutes were staggered against 2 better players in Embiid and Butler. It makes sense that Simmons/Embiid or Simmons/Butler combos might not perform as well as Butler/Embiid. It's not an inditement on Ben, but it does provide further evidence that he's not the best player on the team (nor the 2nd best in last year's case). Only 1 other player in the top 25 had a negative on/off, who also happens to be Beal's lone companion with a negative +/-. That player falls into our final category...
Negative-Negative
Opposite to the positive-positive distinction, the negative-negative shows us the player (a) lost more possessions than he won while on the court and (b) his team performed more efficiently without him than with him. Because there was only 1 negative-negative in the top 25 list, I wanted to see how far I could make it before finding another. Using SI's top 100, I made it all the way to #62 before I found another negative-negative. That's kind of telling, I think. In the rare cases that a star puts up a negative +/-, we should be able to say "yeah but the rest of the team was that bad." With a negative on/off however, that argument doesn't fly. If a true star can't demonstrably improve the production of a bad team, then maybe the player isn't as impactful as we believe him to be.
By now you might realize who I'm referring to. The only player in SI's top 61 players to sport a negative-negative last season was the European wonder: Luka Doncic. By rookie standards, it's really not that worrisome. In fact, the negative-negative at #62 was another rookie: Trae Young. Fast forward to this year though, and Trae has a firmly positive on/off, while Doncic is still a negative-negative 10 games into the season. So what's the deal? His box numbers have been incredible, but his impact just isn't there. Not yet at least. Is there any context Mavs fans can provide to explain what we're seeing? You can laugh off +/- and on/off numbers, but when he's the only player in the top 61 with 2 negatives, it's fair to assume his impact doesn't match his reputation. It's one thing to call him a future superstar, but a current superstar? I'm not convinced. His defense is only one part of the equation because the Mavs' offense produces -4.6 less points per 100 possessions when Luka is on the court.
(Dirk and TexasChuck, please don't hold this thread against me)
I have simple question. Do +/- and on/off numbers measure players impact?
It’s late for me, I cannot wait for your answer. So I will give another question. You’re talking about last season too, implicating that he had negative impact for the team in the last season too. What about last years RPM? We don’t have it for this year, but we have it for last year. Why didn’t you use RPM rather than +/-? Please answer both questions.
Mavs fans, before you defend Luka, lets we hear, why are we defending him in the first place. I don’t believe that anybody would debating about flat earth theory?
It's more of a loaded/heavily-nuanced question than a "simple" question, but I'll break it down for you.
A player's "impact" = the effect he has on team success. The most granular and profoundly accurate measure of team success on a possession-by-possession basis is +/-. But of course, we can't attribute +/- to a single player when he's sharing the court with 4 other teammates at all times. This is why it makes sense to introduce on/off into the equation. The same teammates affecting a player's +/- can be evaluated in a context without that player. As I mentioned earlier, the fact that every team staggers substitutions makes on/off all the more valuable.
If a direct measurement of "impact" existed, player evaluation would be easy/boring. There would be nothing to discuss. No such measurement exists, so we approximate impact using the data available to us. When I rank players, I don't rank them solely on the basis of the data points in this thread, BUT we can still extrapolate significant conclusions. Is Doncic necessarily a negative impact player because of these numbers? No, he's not. Is it possible for Doncic to be a superstar given these findings? No, I don't believe it is. That's why no other superstar- or even regular star- falls into this category. It's grounded in objective reality and not some fluke or conspiracy against Luka.
Regarding RPM: you know what that stat is designed to do? It's designed to approximate RAPM, so it can be used in smaller sample sizes. If we direct our attention to the number it's trying to emulate, Luka's RAPM last season was -0.3068. Even if we took RPM as scripture, Luka ranked 86th last season at +1.29. Significantly worse than any other player regarded as a star. I repeat: I'm not saying Luka is a negative impact player. I'm saying he doesn't provide star impact (yet).
