Bob8 wrote:I'm first to admit that Luka doesn't deserve MVP votes yet. But based on +/- he doesn't deserve to be in any of All Nba teams either. Strangely enough he was in first All Nba team 4x in a row, which is a fantastic achievement for someone being under 25 years old, which just few others in Nba history achieved. How is this possible, if +/- clearly shows that he's not impactful player? Can non impactful player made 4x in a row first All Nba team?
If someone truly believes in +/-, conclusion should be simple, Luka is not even top 3 Mavs' player this season and not even top 100 Nba player. That's what +/- model is saying. Something works or something doesn't work.
I'm addressing these points because they are relevant to the broader discussion, but I will try to leave you alone after this Bob.
+/- stats in general are better served to identify outliers than they are the middle of the pack. The players who are so impactful, rather for positive or negative, that their presence seems to overwhelm the rest of the players in context can be identified with relatively small sample.
This is relevant in an MVP discussion because by definition, these are the players who should be overwhelming their teammates in a positive sense, and generally, that's what we see. Take a look at the top MVP candidates on the whole, you're going to see huge +/- numbers. And so where we see an MVP candidate where this is not the case, and has never been the case, there's something that needs to be identified that's keeping the player in question from showing up in these stats like most of the players seen as his peers are.
It should also be made clear that guys who have big +/- numbers in more limited roles are guys we need to be cautious of using that data for. If a coach is purposefully limiting a guy's minutes, for example, so that he gets to play with lineups that play to his strengths and cover his weaknesses, then +/- data for that player is biased in a positive direction. Doesn't mean everyone who plays in limited minutes is being helped in this way, but caution is advised. What I'd generally advise to a coach is that if a player is putting up huge +/- numbers in limited minutes, and the coach doesn't have a reason why he is avoiding playing the player in other contexts, he should give it a try. But if the coach knows why he's playing the player in one context but not another, then the player's +/- data is only telling us that the coach is probably being quite shrewd.
A lineup example of this is the classic Warrior death lineup with Draymond Green at the 5. Super-effective when it was used but Kerr used it in limited minutes. Maybe he was wrong to do so, but Kerr was also thinking about how effective it was as a change of pace. An opposing team that didn't scheme with the expectation of the Warriors playing a super-small lineup all the time could be exploited by a small lineup.
This also relates to approaches along the lines of what Nick Nurse and some other newer coaches do where what they want is to be able to continually change their on-court strategy over the course of a game so that the opponent can't get comfortable. It works in general so long as you've trained your players to be able to "play multiple" - change schemes and change roles in schemes - more adroitly than opponents.
Re: how could he make All-NBA if not impactful? 2 big things here:
1. All-NBA is just the opinions of people, and those people can be wrong. It's important as an analyst to understand how these people are coming to the conclusions they are, but once you understand that, if you're looking at some facet of the results that they don't understand as well, there's nothing all that concerning about disagreeing with them.
And yeah, there should be no hesitation in thinking that experienced posters in the analytics space might understand the data better than the official voters. That's just a known thing. The voters are absolutely getting better than they used to be, but historically you're largely talking about guys who are in the job they are in because they like sports and like to write, not because they have STEM expertise.
This also relates to why it took decades for the 3-point shot to be properly exploited. The people making the decisions historically, while they are knowledgeable about basketball culture norms, aren't the people you should expect to make data-based leaps.
2. As I've tried to be clear, I'm not saying Luka is not impactful, only that relative to the outliers who make up MVP ranks, he's doesn't measure up. It's certainly very impressive that he can handle as much of the playmaking burden as he can, and it's possible that having him play like this is the best way forward in terms of playoff resilience, but in terms of it resulting in huge on-court impact as a matter of course, we're 6 years in and we just haven't seen it.
Does that mean I wouldn't vote for him for All-NBA? No. I don't typically focus on stuff like All-NBA that much and would naturally see him as a bit weaker of a candidate compared to most because 1st team typically pits MVP candidates again each other, but it's not the case that there are typically 6 guards that look like MVP-level impact guys. Right now, for example, there's really just Shai and then a huge gap, so some non-MVP candidate guy would need to be placed on the 1st team, and it's possible Luka's as good of a candidate as any.
