LukaTheGOAT wrote:Doctor MJ wrote:LukaTheGOAT wrote:
I mean I don't think it would be just 76ers fans who are championing Embiid over Maxey. I believe the general NBA consensus would always side with Embiid because he's several times the player Maxey. Missing more games only dents his impact compared to other MVP level candidates, but he is still well ahead of Maxey.
By the general cumulative value metrics:
Embiid
VORP-3.9
Estimated RAPTOR WAR-6.3
LEBRON WAR-4.6
EPM Estimated Wins-9
Maxey
VORP-2.2
Estimated RAPTOR WAR-5.5
LEBRON WAR-3.9
EPM Estimated Wins-7.1
Just outside of that Embiid being a top 5-10 defender in the sport while Maxey being a neutral, and also Embiid also being the best scorer during this RS seems like a straightforward enough argument for why Embiid is still more valuable than Maxey.
All of these stats are determined by the box score, but when we just look at +/- data, it sides firmly with Maxey.
I'm not saying Maxey is the better player here, but what I would say is that any notion of Embiid "several times the player" of Maxey isn't based on rigorous thought.
I think people clearly are still in love with the idea of a guy scoring tons of points, and thinking about how that other guy couldn't do the same thing, but I'd say the general answer is that teams generally shouldn't be having a guy scoring 35 PPG, and if they let a guy do that, the team's offense won't be hurt like people expect without that player.
But look, here's my question for you:
Right now, Embiid has missed about 1/3rd of the minutes he should have played.
I assume that if he missed 3/3rds of those minutes - as in Embiid playing 0 minutes - you wouldn't consider him the MVP of the 76ers.
So, would you say the threshold that would say you right now is missing about 2/3rd of the season? If Embiid played 500 to Maxey's 1500, is that about even value?
If not, name your fraction. How many hundreds of minutes would Embiid need to have missed this year in order for you to factor it in when asking whether he was the MVP of his team? Does it start hurting in the 700+ range? 800?
First off, 3/4 of the stats I mentioned are not determined by JUST THE BOX-SCORE. They incorporate the box-score, but include plus-minus data to try and make things more stable.
Most people agree on/off data is highly volatile over smaller sample sizes. The box-score as a prior helps to stabilize results over a shorter period. To act as if the box-score can't be a bit informative when we lack solid information is problematic to me. At least the box-score is typically more stable, and we can look at past performances with similar box-score numbers to get an idea of how much that impact should be worth.
Also, yeah Embiid is several times the player of Maxey? I think most people would agree. Maxey is at best a 3rd team all-nba level player when he plays. Embiid at his best is having an all-time level season when he plays and is at worst an MVP-level guy this year. That's a notable gap.
For example, using BT's CORP values, an all-star could be 5% random probability of winning a title. An all-nba guy could be at 8%. A solid MVP guy is a 17%, and an all-time season could approximate to 23%. So yes, based on RS performance, I think it could be multiple/several times the player of Maxey in terms of CORP.
In terms of plus points per game added:
All-Time Season (23-30 percent or +6)
MVP Season (17-23 percent or +5)
Weak MVP Season (12-17 percent or +4)
All-NBA Season (8-12 percent or +2.5)
All-Star Season (5-8 percent or +1)
You could make the argument there is a 5x difference in their per game impact. That is huge.
We went through this on/off stuff last month with Minnesota.
You wrote "DPOY is weird right now because it feels like it should be Gobert given Minny's success, but the on/off data appears to be pointing to Edwards instead."
You implied because Edwards had better on/off, we should question whether Gobert is the DPOY, and possibly if it is Edwards instead. Now Gobert, is your DPOY leader as his net on/off looks better than Edwards now. When if you just watched the games, it was clear Gobert was more valuable on that end, and the now numbers reflect it (counter argument, if you are going to try to say Gobert has played better on defense, then explain how). On/off is volatile and writing off other pieces of information given to you with the on/off excuse seem problematic.
Around the time Chris Paul went down with an injury, he had a +8.5 Net on/off. Steph had a -6.1 Net on/off? Do you really believe Chris Paul drove GSW's offense more than Steph did?
Continuing on, chucking up what Embiid is doing as a scorer, as just scoring a lot of points is underestimating what Embiid is doing. If you want to chuck this up to being a 64 Wilt situation where he scores a lot of points on great efficiency, but believe the team offense could be better with more playmaking, so be it. That still wouldn't prove Maxey is a better offensive player or overall player (both Wilt and Embiid are lifting their teams' offenses a lot). Also, if Embiid is taking less shots, exactly who on his team should be getting more shots?
Also, let's just break down their games.
Embiid is currently averaging 35 pts per game on 64.7 TS%. Maxey is currently at about 26 pts per game on 58.6 TS%, meaning he is scoring on slightly above average efficiency relative to the league.
Sure, you could try and make an argument that Maxey's scoring is more impactful...I would be interested in knowing what in the world your justification is though. Embiid has massively higher volume, is treated as the #1 threat on offense and has more gravity because of it, still is much more efficient. Considering how teams have to tilt their defense to try stop Embiid, I am not sure Maxey's playmaking value is notably higher...I would ask for evidence that Maxey's playmaking moves the needle that much, when his pure passing isn't amazing.
Finally, Embiid is a top 10 defender in the league at worst this year, while Maxey is merely around average. There's around a +2 pts per game difference between them on defense I would imagine. Once again, I don't see how they are particularly close, or how Maxey is in Embiid's stratosphere when they both play. Like what could Maxey possibly be doing to make up the gap in impact. And don't tell me raw on/off that isn't even adjusted for teammates, what exactly in Maxey's game is helping him make up ground to Embiid?
Single Year-RAPM can be highly suspect, but Embiid is higher when you adjust for the players on the floor for them.
Embiid-2.03
Maxey-1.88
https://psteve.shinyapps.io/RAPM/To answer your question, if Embiid only played 1/3 of the season (instead of 2/3), by the value metrics I posted, Maxey would be ahead. I do believe these value metrics are more linear in their nature, and probably don't capture the exponential value of having a star playing, but Maxey would certainly have a better argument.
Well, I appreciate the thorough way you responded to me. I'll warn you up front that my response is pretty pithy.
- Never said they were based just on the box score, and I fully understand why people want to add box score in to...everything they look at. But when we're having a debate about this stuff and one guy scores 35 PPG efficiently, he's going to clobber his teammates by pretty much any box score stat, so it's superfluous to bring one of them up, let alone listing out a whole list of things that all agree because they use a lot of the same stuff.
- Re: Most would agree Embiid is "several times" Maxey. I mean, what you're saying is a meaningless statement. You can't multiply a player by a number and get another player.
- Re: Maxey's at most an All-NBA 3rd guy so 8% and 5x. You're making assumptions about the players in the debate you're in, trying to connect those assumptions to a sense of rigor, and then trying to use that as proof you're winning the debate.
- Re: went through on/off last month with Minny. Me saying we should ponder something without making an assertion that something must definitely be the case is just that. You didn't see me argue that Edwards was the DPOY, right? So obviously I'm not taking the data in question as gospel.
- Re: do I really believe Paul/Curry...Ponder.
- I fully understand the variance concerns with +/-, and understanding this isn't a contradiction of any of the things I have said.
- Embiid does more than score points. Certainly.
- If no Embiid, who? In general, schemes get altered and shots get re-allocated when the primary option is off the floor. When Embiid is on the floor Philly plays aiming to get the ball in to their center. When he isn't, they don't force the ball into whoever is in there in his place like mindless robots. The 76ers have played almost 1000 minutes without Embiid and have a 118.7 ORtg in that time period, why don't you go look at what's happening in that significant amount of time?
- Re: Maxey's scoring more impactful on worse numbers? First let me emphasize again that I'm not looking to take a concrete stance that Maxey > Embiid in any bold way. Now forgetting the holistic comparison, we're not talking about scorer vs scorer. We're talking about an offense built around getting the ball to a center, or an offense built around a playmaking guard. This fundamentally different structure is the central thing to understand about how a team can actually look a lot better than you'd expect when a massive volume/efficiency scorer is out.
You might call this the Dantley Delusion. If you just look at volume/efficiency, you'd expect Dantley was a more valuable offensive player than Jordan. In reality his offensive impact was generally pretty modest, and the general rule was that if you realized you had other offensive talent on your roster, you should probably trade Dantley.
I'm not looking to say Embiid's Dantley, but I am saying that you cannot simply look at a scorer's volume/efficiency, compare it to the rest of his roster, and assume that the team will get that much worse without him.
- Re: Ask for evidence that Maxey's playmaking moves the needle. In nearly 1000 minutes without Embiid the team has a 118.7 ORtg. If you don't think Maxey is playing a major part of that, I would ask you to show evidence pointing to others on the team who are making this happen. Though that's perhaps an unreasonable request given that it seems likely that you believe that noise is an appropriate explanation for the entirety of 1000 minutes of team play.
I'll emphasize that I'm not saying that 1000 minutes of play is proof that the team will always have an 118.7 ORtg without Embiid, only that if 1000 minutes with Embiid is enough for us to believe we can learn something, then the same should be true for 1000 minutes without him.
That aside, I think the key thing here is not so much that Maxey's an amazing passer, but that there's a cost to running your offense focused on a big man's scoring. Doesn't mean that this cost isn't worth paying, but any time the defense knows who you're trying to get the ball to, they can make it harder. When an offense is built around a guard, they don't have this option. If they really want the ball out of the guard's hands, they have to commit multiple defenders out at the perimeter, which opens up major gaps closer to the basket. This gets to the heart of why I would say that the most valuable offensive players in the league have generally been perimeter guys ever since they widened the key in the early '50s. Some bigs are so good they make it worthwhile, and I'd put Embiid in that category...but there's still a tax in any situation where a guy wants the ball in specific positions so that he can score from there.
- Re: what about Embiid's defensive edge? Well of course, I started my original post here drawing a distinction between OPOY & POY with regards to Maxey. I then confused the matter by acknowledging that what's true for ORtg is also true for Net, but yes I do certainly understand an argument for Embiid overall over Maxey on the grounds of defense alone.
- Re: RAPM. First, thank you for the link to the site. I was looking for a new one to look at that data for. I'm not surprised Embiid has a slight edge there given that Maxey's edge in on/off is only slight. But that slight advantage set in comparison to Embiid missing so, so much time only gets us back to the original concern.
- Re: If Embiid played only 1/3rd the minutes of his team's other core players, that would be enough for you to acknowledge he wasn't his team's MVP. Glad to hear it. I still think you really underestimate the damage of a player being gone as much as Embiid has been. If we split the difference between 2/3rds & 1/3rd to go to a half, then you're literally talking about in practice a 50% chance that at a time when you'd expect to have your mega-max franchise player, you get nothing instead.
You talk about stars having "exponential" impact. I understand that you probably meant that hyperbolically, but it seems to agree with your assessment that having Embiid a fraction of the time, and a gaping hole about as often in his place, is still drastically more valuable than having an All-NBA level player who plays All-NBA level minutes.
Let me end like this:
I'm getting pretty chippy with you here and I apologize for that. Not the most mature of me. I perhaps shouldn't try to attribute it to anything other than my mood and imperfect moral fiber...but you're also doing stuff here that you may not be realizing. There's this combination of a) you asserting analytical rigor complete with attempting to correct me on basic things, next to b) you making meaningless mathematical statements while completely abandoning an analytical mindset when it comes to huge minute concerns, that's really hard to swallow.
